Siglund Bruhn - Bach Analysis

J. S.
Bachs
Well-Tempered Clavier
In-depth Analysis and Interpretation
Siglind Bruhn
EDITION
Fachverlag fr
ORZ
eisteswissenschaften
Bruhn, Siglind.
J. S. Bachs Well-Tempered Clavier: In-depth Analysis and
Interpretation
Waldkirch: Edition Gorz, 2014.
kontakt@edition-gorz.de http://edition-gorz.de
Second, completely revised edition
(First edition Hong Kong: Mainer International Ltd., 1993)
ISBN 978-3-938095-19-5
Siglind Bruhn 2014. All rights reserved
No part of this book may be reproduced or transmitted in any form
or by any means, electronic or mechanical, including photocopying,
recording, or by any information storage and retrieval system without
the written permission of the author, except where permitted by law.
Bibliographic information, including detailed bibliographic data,
are obtainable from the Deutsche Bibliothek under http://dnb.ddb.de
Printed in Germany by rombach digitale manufaktur, Freiburg
Table of Contents
How to use this book
Introduction
a) The Well-Tempered Scale
b) The Clavier
c) The Preludes in Bachs WTC
d) The Fugues in Bachs WTC
Questions Regarding a Prelude
Questions Regarding a Fugue
Additional Information
7
9
9
11
13
15
17
19
21
Considerations and Model Answers for Book I

WTC I/1 in C major
Prelude 49
WTC I/2 in C minor
Prelude 61
WTC I/3 in C major
Prelude 73
WTC I/4 in C minor
Prelude 87
WTC I/5 in D major
Prelude 101
WTC I/6 in D minor
Prelude 113
WTC I/7 in E major
Prelude 123
WTC I/8 in E /D minor
Prelude 135
WTC I/9 in E major
Prelude 147
WTC I/10 in E minor
Prelude 157
WTC I/11 in F major
Prelude 167
WTC I/12 in F minor
Prelude 175
WTC I/13 in F major
Prelude 189
WTC I/14 in F minor
Prelude 201
WTC I/15 in G major
Prelude 211
WTC I/16 in G minor
Prelude 223
WTC I/17 in A major
Prelude 233
WTC I/18 in G minor
Prelude 243
WTC I/19 in A major
Prelude 253
WTC I/20 in A minor
Prelude 267
WTC I/21 in B major
Prelude 281
WTC I/22 in B minor
Prelude 289
WTC I/23 in B major
Prelude 301
WTC I/24 in B minor
Prelude 311
5
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
47
52
66
77
91
105
116
129
139
150
161
169
180
192
204
214
226
235
246
256
270
283
292
303
315
6
Considerations and Model Answers for
WTC II/1 in C major
Prelude
WTC II/2 in C minor
Prelude
WTC II/3 in C major
Prelude
WTC II/4 in C minor
Prelude
WTC II/5 in D major
Prelude
WTC II/6 in D minor
Prelude
WTC II/7 in E major
Prelude
WTC II/8 in D minor
Prelude
WTC II/9 in E major
Prelude
WTC II/10 in E minor
Prelude
WTC II/11 in F major
Prelude
WTC II/12 in F minor
Prelude
WTC II/13 in F major
Prelude
WTC II/14 in F minor
Prelude
WTC II/15 in G major
Prelude
WTC II/16 in G minor
Prelude
WTC II/17 in A major
Prelude
WTC II/18 in G minor
Prelude
WTC II/19 in A major
Prelude
WTC II/20 in A minor
Prelude
WTC II/21 in B major
Prelude
WTC II/22 in B minor
Prelude
WTC II/23 in B major
Prelude
WTC II/24 in B minor
Prelude
Further Reading
Table of Contents
327
337
347
357
371
381
393
403
413
425
437
445
453
465
477
485
495
507
519
527
537
547
557
569
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
Fugue
325
330
340
351
363
375
385
397
407
417
429
440
448
458
468
479
488
498
510
521
530
540
549
560
573
579
How to Use This Book

This book is not meant to be read through in one go. It aims to help
you, its reader, achieve a creative understanding and interpretation of
Bachs preludes and fugues by encouraging you to think for yourself.
After a brief introduction to the nature and origin of the Well-Tempered
Clavier, you will find three parts of very different length and emphasis:
I
II
III
questions to be pondered
additional information
discussions of possible answers to the questions
with regard to each of the 48 preludes and fugues
Before you approach the catalogue of questions, decide which prelude

or fugue you wish to embark on and open your score. This is because the
questions will only make sense if you have a particular piece in mind. The
best edition to use for a task like this and for any serious interpretation, for
that matter, is an Urtext, a score that is free of any particular editors
personal markings and contains only Bachs unaltered writing.
Now start working on the first question. Can you answer it right
away? If you have any doubts as to what exactly the question implies, or
what a particular term means, refer to the pages providing additional
information. They will give you a more general background to each of the
specific issues to be considered.
Proceed in the same way through all the questions. Be sure to take
notes on your own answers: there will be far too many details to keep in
your head. You can mark most of your answers directly into your score but
will need a few separate sheets of paper for some drawings.
Finally, when you have found all the answers, or when you feel that
you are stuck with a particular one, look up your piece in the section of
this book that gives an in-depth investigation into the work, and compare
the model answers with your own results. It is important to understand that
for quite a few questions, there are several possible solutions: while some
attitudes may be unwarranted, others are a matter of personal preference
and thus not a matter of right or wrong.
7
How to Use This Book
This book seeks to list alternative solutions and views wherever

possible, discussing the implications of each solution for the character and
interpretation of the piece. In some cases, however, it may not have been
possible or advisable to follow up on all of the alternatives. Using your
own mind you will surely have no problems drawing any additional
conclusions not mentioned here.
If, by the time you work on your third or fourth prelude and fugue,
you have become so independent as an analyst and interpreter that you
need only glance at the questions, if you can skip the information and
peruse the model answers merely to consolidate your ideasthen this
book has proved successful.
Abbreviations used:
SATB
UML
v1, v2, etc.
voices in four-part polyphony: soprano, alto, tenor, bass

voices in three-part polyphony: upper, middle, lower voice
voices in polyphonic works of five or more parts
S
CS
dux
comes
M
CM
rh
var
inv
E
subject: the principal thematic component in a fugue

counter-subject: its recurring companion
the subjects original form, launched from the tonic
the subjects answer, launched from the dominant
motif: a small, distinct, recurring thematic unit
(artificial term: the contrapuntal companion to a motif)
rhythm, rhythmic pattern, or rhythmic figure
varied, variation
inverted, inversion
episode
m. 51
mm. 13-31
TSD
T relative
t relative
I IV V
i iv V
the downbeat of measure 5

from the third beat of measure 1 to the downbeat of m. 3
tonic, subdominant, dominant in the major mode
the minor triad on the 6th scale degree of a major key
the major triad on the 3rd scale degree of a minor key
the main cadential steps in a major key
the main cadential steps in a minor key
r.h., l.h.
right hand (righ-hand part), left hand (left-hand part)

1.1 Can you place your prelude in one of the following categories?
a) Harmonically determined
b) Motivically determined
c) Adhering to the principles of invention or fugue
d) Rhythmically determined
e) Metrically determined
1.2 What is the overall design of your prelude?
a) Where does the first harmonic progression conclude?
Has Bach composed this cadence as a structural break,
or does the flow of the piece continue uninterrupted?
b) Where does the next harmonic progression conclude?
Does this seem to coincide with a structural caesura?
c) How many such structural sections can you distinguish?
d) Does the prelude feature any structural analogies?
Does any passage recur, either repeated or varied?
Does any passage recur in transposition?
Does any passage show a structural progression that
corresponds to an earlier passage?
1.3 Practical considerations for performers:
a) What is the basic character of this prelude, and what tempo
and articulation are therefore most appropriate for the
material?
b) If the score shows any ornaments, how should they be
executed?
1.4 Answer those questions from the following selection that apply to
your prelude:
17
18
1.4a
What is happening in your harmonically determined prelude?

What kind of dynamic development does the harmonic
progression represent?
Are there any secondary features?
How do they relate to the harmonic progression?
1.4b
What is happening in your motivically determined prelude?

What are the relevant motifs?
What is their basic character and their dynamic shape?
How are they developed?
What does the way in which they are developed express?
What is the overall dynamic outline in the prelude?
1.4c
What is happening in your fugue- or invention-type prelude?

What is the material on which Bach built this composition?
What is the design of the piece?
1.4d
What is happening in your rhythmically determined prelude?

What are the rhythmic patterns?
Are some of them interrelated?
In which melodic guises does each rhythmic pattern appear?
What means of motivic development does the composer
deploy when playing with these patterns?
What kind of dynamic build-ups and releases can you detect
in the single parts as well as in the whole piece?
1.4e
What is happening in your metrically determined prelude?

What is the predominant pulse?
What are the secondary features beyond this pulse?
Are there any dynamic build-ups?
If so, by what means are they created?

2.1 What characterizes the subject of this fugue?
a) How long is it? Where does it end?
How does it begin: with a strong beat or an upbeat?
b) Is it made up of one indivisible phrase, or does it consist of
several subphrases?
c) What kind of pitch outline does it show?
Does it contain mainly small intervals, or many leaps?
Does it feature any intervals of particularly high tension?
d) What is the rhythm within the subject?
Does it contain many different note values?
Are there dotted notes, tied notes, and/or syncopations?
Are these rhythmic features also prevalent
throughout the entire fugue?
e) What is the subjects harmonic background?
f) Considering harmony, melody, and rhythm, where is the
climax within the subject, and what kind of dynamic outline
is expressed?
2.2 What is the importance of the subject in this fugue?
a) How many subject statements does the fugue comprise
altogether?
In which part (voice) and in which measures do they appear?
b) Does the subject remain basically unchanged, or are there
modifications of detail, of shape, or of length?
c) Does the subject ever appear in stretto or parallel?
2.3 How many counter-subjects does Bach invent for this fugue?
Answer for each of them:
a) How long is it?
b) What is its phrase structure?
c) Considering the melodic and rhythmic features, where is the
climax and what kind of dynamic outline is expressed in the
counter-subject?
d) Can you draw a sketch showing phrase structure and dynamic
outlines in the interplay of subject and counter-subject(s)?
19
20
2.4 What happens in the episodes of this fugue?

a) How many subject-free passages are there? Where?
b) What material is used? Is any of the episodes related to the
subject? Are there any significant episode motifs? Which are
these? What is their character and dynamic outline?
Are any of the episodes no more than cadential closes?
c) What is the relationship between the episodes? Is any of them
a varied, transposed, or otherwise altered version of another?
d) What role in the development of the composition does each
episode play?
2.5 For interpreters: Consider the main aspects of performance practice:
a) What is the character expressed by the material of this fugue?
Consider the details of the subject, its counter subject(s), and
the episode motifs: Would you conclude that this fugue
constitutes a rather lively or a rather calm basic character?
b) What tempo and articulation are advisable for this material?
c) What tempo proportion for the prelude and fugue seems best?
d) If there are ornaments, how are they to be executed?
Should any of them be transferred to analogous notes that
appear unornamented in the score?
2.6 What is the design of the fugue?
a) Are there any indicators that reveal the structure of this piece?
Does the entering order of the voices show obvious grouping?
Does any subject statement appear in reduced ensemble?
Which episodes seem to conclude a section?
Do any two stretches of the piece show an analogous design?
b) What is the harmonic outline of the whole piece?
In what progression of keys does the subject appear?
Which subject entries belong harmonically together?
Are there any significant cadential closes in the fugue?
c) Can you draw a sketch of structure and texture?
It should show all the subject statements (mark them S),
the counter-subjects (CS1, CS2 etc.), the rests in any voice,
the episodes (mark them E1, E2, etc.)
2.7 What is the overall dynamic outline in this fugue?
a) What kind of dynamic development can you detect in each of
the sections you have found?
b) How do the sections relate in this respect?
1.1 What features in your piece help you
determine the prelude type?
Music is often defined as a sequence of sounds arranged chiefly on
three levels: the horizontal, the up and down in a tune, called pitch pattern;
the vertical, where several pitches coincide thus forming harmonies; and
the temporal, with longer or shorter, regular or irregular time spans from
the beginning of one tone to that of the next, called rhythm.
Each of these levels is usually carefully organized. The horizontal level
contains melodic units, phrases, and sections; this is what we are after
when analyzing structure. The vertical level distinguishes the ways in which
simultaneous tones relate to one another, defining harmony and texture.
Finally, on the level of time, the multitude of rhythmic values is packed
into measures of usually equal length in an orderly way, with a definite
hierarchy between their beats; this is called metric organization.
Although any piece of music will feature most if not all of these components, composers usually seek particular expression of their artistic idea
by molding these components so that a certain design is recognizable. For
such a design there is one basic choice to make: does he want to compose
along the lines of a standard model that can be recognized as such, or does
he prefer his piece to sound more like a spontaneous improvisation?
In the case of established structural models there are certain rules by
which composers must abide: dances are defined by certain metric and
rhythmic features, canons by a specific texture, sonata form by a particular
sequence of material, etc. In the case of a composition put together more
like an improvisation, they may choose one or two of the six components
mentioned above and shape their piece primarily according to these. In a
prelude, i.e., in a composition carrying a title that does not reveal anything
about the content of the piece, discovering this determining force is
therefore of vital importance for both listeners and interpreters.
a) Preludes determined mainly by harmonic processes
Whenever melody and rhythm are negligible or neutral, harmony is the
ruling factor. This rule of the thumb sounds simple enough. But are
melody and rhythm not always present?
21
22
Of course that is true. Yet, while rhythm is never actually absent, it

can be described as neutral if all rhythmic values are of the same length, or
if different note values in different voices are arranged in such a way that
their combined pattern amounts to completely regular motion. (This is
called complementary rhythm.)
Similarly, pitch will of course never be absent, but it does not necessarily form melodies. A broken-chord pattern, whether created by a single
line or jointly by complementary voices, is not primarily a melodic feature.
(It could turn out to be melodicas it often does in themes of the Viennese
Classical styleif it gives way soon enough to a more lyrical continuation.
However, if the same broken-chord pattern continues for many measures,
these will be judged as lacking genuine melodic features.)
b) Preludes determined mainly by motif or figure development
Motif is the term for a small melodic unit that is individual enough to
be immediately recognizable when it recurs. (By common consent, such a
unit is only called a motif if it does appear more than just once.)
Figure is the term preferred whenever such a unit is unsingable, i.e.,
with a rhythm so fast or a pitch pattern so unmelodious that, while we
might recognize it as a whole, we may not necessarily become aware of
each single detail contained in it.
We speak of development whenever such a unit is taken up repeatedly
and under varying circumstances, prompting it to adjust some of its features.
A composition should be described as determined mainly by motivic
development if its material relies predominantly on a (usually small)
number of motifs, i.e., if, in a major portion of a piece, there is always at
least one voice that derives from those melodic units.
c) Preludes adhering to the principles of invention or fugue
Both invention and fugue are standard structural models that commonly
appear bearing the corresponding headings. Yet a prelude, whose title
leaves it undeterminable in form and texture, can be composed in such a
way that it abides by the rules of either of these models.
How can we distinguish between motivic development and these
established forms?
Whenever a prelude is conceived as a strictly contrapuntal composition (i.e., in a texture of two or three voices all of which are independent from each other and all of which are continued consistently
throughout the piece); and whenever it relies mainly on one melodic
idea, then it is probably either an invention or a fugue.
23
If we wish to investigate further, how can we distinguish between

the two models? In both cases, the main idea may occur accompanied by a companion or counter idea; in both forms, the main
idea may retreat for a while and then return. We are therefore
looking for other features that might aid distinction.
If the main idea enters in all voices successively, and if these
imitations begin alternately from the tonic and the dominant, we can
speak of a fugue. This is particularly the case whenever there are
counter-subjects that, like the subject, wander through all voices.
(Where a prelude in the style of a fugue usually differs from a real
fugue is at the beginning: instead of setting out with an unaccompanied subject statement it will introduce its main idea surrounded
by a few harmonizing notes or chords.)
If none of the above-stated conditions is given, we can speak of an
invention. This is particularly the case whenever the main idea
appears not only in imitation but also in sequence.
d) Preludes determined mainly by rhythmic patterns
There are preludes that seem determined by a few small units of a
particular kind. These units retain their rhythm but constantly change pitch,
so that one may define the composition as being based on rhythmic
patterns. Consequently, it is this little repertoire of rhythmic models which
should find its way into the listeners memory.
e) Preludes determined mainly by metric organization
Whenever a piece appears determined mainly by its metric features, we
can be sure that it is meant to convey a certain undefinable atmosphere
rather than the impression of an intricate play with structures or material
development. In such a piece there will be a regular pulse (or any very
simple, fairly monotonous variation on this pulse) dominating everything
else.
Such music was traditionally used (and is still being used in this way in
many non-Western cultures) to instill in the audience a meditative mood,
if not to create a mesmerizing effect in the listeners mind. In the context
of western music, such a meditative piece often precedes a composition
that its author regards as too profound to be grasped by an unprepared
mind. The metrically-determined prelude thus serves to relax listeners and
empty their minds of too much activity.
24
1.2 What factors determine the overall design of a prelude?

Every piece of classical music is horizontally structured, i.e., it consists
of several sections, each of them determined by a completed harmonic
progression. Simple forms, like some dances and songs, are often made up
of two halves that are more or less analogous (binary). Others feature a
kind of frame around a contrasting center: A B A (ternary). Such pieces
characteristically show repeat marks indicating the boundary between A
and what follows. Conversely, preludes are more typically designed as a
chain of phrases. Their length tends to be more irregular than in simple
forms where patterns with an even number of measures are preferred.
a) Where does the first section end?
When trying to determine the first structurally relevant cadential close,
one is looking for a dominant (V) leading into a tonic (I). This combination
should be preceded somewhere by a chord on either IV or II. While the
dominant may sound with various additions to its triad (e.g., with a seventh
or a ninth) and may even sometimes appear without its root (vii7), the tonic
is less volatile: it never lacks its basis and is not normally spiced with
additional notes.
Whether the first cadence closes a separate structural section or serves
only as an introduction to a larger portion depends on the melodic design
up to this point. This can be easily determined by considering the following:
If the first cadential progression supports a single melodic unit that
is complemented by an answer in the following measures, a more
encompassing structural portion must be assumed.
If melodic design is either absent or arbitrary, the first cadence
qualifies for the closure of an independent structural section.
(Exception: In cases where one of the main voicesthe top line or
the bassfails to take part in this initial phrase, either by pausing or
by sounding a pedal note, and thus suspends the real beginning, a
larger context is also implied.)
b) The second cadential progression
The second cadential progression in a piece usually departs from the
works home key toward a new tonal center that will again be confirmed
by a perfect cadence. This modulating passage is nearly always more
prolonged than the initial harmonic progression. Its end usually marks the
conclusion of a structural section, irrespective of the function of the
previous cadence.
25
c) The sections
The number of structural sections in a prelude is theoretically not
restricted, although it is rare to find more than six. Their boundaries are
always delineated as described above: a conclusion in the harmonic progression coinciding with a conclusion in the melodic line if there is one.
d) What exactly constitutes a structural analogy?
Structural analogies recall the simpler binary and ternary designs and
are therefore a very popular building principle in preludes. Analogies may
consist of as little as two or three measures but may extend to the length of
a complete section that is in some way recapitulated.
The simplest, but also least frequent, manner of creating an analogy is
by literally repeating several measures. More ingenious and more widely
used are variation, transposition, and correspondence.
In variations, several measures recur with modified features. Various
details may be altered: One or several of the voices may be written
in a more elaborate pitch and/or rhythmic pattern, the leading line
may be given to another voice, or an accompanying voice may
change its pattern. The features that will always remain the same in
literal variations are the key, the overall harmonic progression, and
the length of the segments.
Transpositions in their simplest form keep all the original details
intact but merely transfer them to a new tonal environment. If there
are, in addition, more than arbitrary changes in the details, we would
speak of a varied transposition.
In correspondences, the structure recurs in principle rather than
literally. The surface changes may be so considerable as to make it
difficult to spot the analogy at first sight. For a good understanding
of the structure, however, correspondences are just as important as
other analogies, not least because they are much more frequent.
Imagine the following structural pattern in an early section: two-bar
model with sequence + modulation to a new key + cadential close with twobar closing formula. Let us assume that the segment recurs analogously at
a later stage of the piece, initiated by a different two-bar model with its
sequence, that the modulation takes an extra measure to get to where it is
heading, and that the closing formula, although of equal length, features a
different melodic design in the treble. This is a subtle way of creating correspondenceand very rewarding to detect. Understandably, such analogies
only make sense if the surface designpitch pattern, rhythm, and texture
is neutral enough to draw sufficient attention to structural processes.
26
1.3 How do you express the character of the prelude?

a) What features help in determining the preludes basic character?
The character of a composition (often called the basic character) is
conveyed through its material. Determining this basic character may at first
seem a difficult task since Bachs scores contain almost no indications of
dynamics, touch, articulation, or tempo. There are few exceptions in the
Well-Tempered Clavier. Single wedges and slurs occur occasionally (see
e.g., the subject of the D-minor fugue in vol. 1), and tempo is indicated
three times (both the B-minor fugue in vol. 1 and the G-minor prelude in
vol. II are marked Largo, while the B-minor prelude in vol. II is Allegro.)
In the absence of verbal hints we must assume that musicians of Bachs
era were capable of deducing the character from information contained in
the music itselfi.e., from rhythmic patterns and pitch design. If we try to
put ourselves in their shoes, we must therefore ask: what kind of rhythmic
and intervallic structure distinguishes a rather calm character from a rather
lively one? The answer is simple enough and worth remembering: In pitch,
a predominance of steps or leaps distinguishes the contrasting characters:
leaps are regarded as more energetic, stepwise motion as more emotional.
In rhythm, the distinction is between simple and complex features: a
complex rhythm requires time to unfold whereas a simple rhythm invites
virtuosity and might even sound less interesting if played too calmly. The
rule of thumb is therefore:
A predominance of stepwise motion and a complex rhythm
indicate a rather calm basic character.
Frequent leaps and a fairly simple rhythmic pattern
indicate a rather lively basic character.
Once you have established the basic character, you need to translate
this fairly abstract concept into practical performance features. Here are
some hints for conclusions to be drawn regarding tempo and articulation:
Whenever the material of a composition represents the rather calm
basic character, the tempo can be anywhere between moderate and
very slow. The general articulation is legato. Exception: cadentialbass patterns and consecutive leaps (i.e., more than one leap in a
row) require non legato.
Whenever the material of a composition represents the rather lively
basic character, the tempo can be anywhere between moderately
flowing and very fast. The appropriate articulation is a crisp legato
(also called quasi legato) for the shorter note values and non legato
27
for the longer values. The main exception in the longer values: an
appoggiatura is always inseparably linked to its resolution, and the
keynote / leading note / keynote (dosido) formula is always legato.
Longer notes in a rather lively piece need further specification
whenever we are dealing with notes prolonged by a dot or tie. While
these are certainly longer, they are detached only where the prolongation itself (i.e., the value of the dot or tie) is of longer
duration. If the time value of the prolongation equals one of the
shorter notes in this composition, the note will be linked to the
following one in quasi legato. Thus in a composition where non
legato quarter-notes and quasi legato eighth-notes are a rule, a
half-note prolonged by a dot or tied into a quarter-note would be
detached from the following note; but a note of any value prolonged
by an eighth-note tie would not be detached.
b) How do you make decisions regarding ornament realization?
Ornaments in polyphonic compositions of the Baroque era are, more
often than not, decisive features in the material they embellish. Their
convincing execution depends on a smooth beginning, a speed that relates
to the rest of the piece, and an appropriate ending.
The most common ornament in Bachs fugues is the long trill, indicated
by one of several symbols: the abbreviations or t, the mordent symbol,
or one of the signs for compound trills. Here is how the long trill is played:
In keeping with the basic rule for ornaments in this epoch, it begins on the
upper neighboring note. However, there is a very frequent exception:
Whenever the trilled note is approached stepwise, i.e., whenever the main
note is preceded by the interval of a second from above or below, the trill
begins on the main note. Another exception, albeit much less frequent,
applies whenever an ornament decorates the first note of a phrase (see e.g.,
the initial note of the G-minor prelude from vol. I). In this case, too, the
ornament begins on the main note. The speed of the shakes is related to the
general motion in the composition in such a way that the value of each
ornamental note is preferably twice as fast as that of the fastest note values
that do not qualify as written-out suffixes.
While trills beginning on the upper neighboring note enjoy regular
motion throughout, those launched from the main note and therefore comprising an uneven number of notes hold the first trill note for the duration
of each of the following two-note shakes, in such a way that all further
upper auxiliaries fall on the stronger pulses within the trill, thus evoking
the effect of a long appoggiatura.
28
Whenever the note following the trill is its harmonic resolution and
falls on a strong beat, the end of the trill should be prepared (announced)
with a suffix, i.e., a turn to the lower neighboring note and back to the
main note. Conversely, no suffix is desirable if the resolution appears too
late, too early, or not at all. The first case happens if the trill is prolonged
with a tie; in this case the shakes must stop short on the last main note
before the bar line, so that the tie can be fully appreciated. The second case
occurs in dotted note patterns; here the shake ceases preferably shortly
before the dot. Finally, if the trill is succeeded by a rest or a leap not
providing harmonic resolution, the shake should continue up to the very
end of the ornamented note value, ending again on the last main note.
Any ornament appearing in the subject or one of the counter-subjects
must be regarded as integral to this component of the thematic material.
Regardless of whether or not the composer repeats the ornament symbol in
each recurrence of this component, the performer should retain such a trill
throughout the work (provided the subject or counter-subject does not
appear in drastically modified shape).
1.4a How do you deal with preludes

determined mainly by harmonic processes?
Harmonic progressions and the development of dynamic tension
When trying to find a valid interpretive concept in pieces determined
primarily by harmonic processes, the basic facts to be considered are the
tension generated by each of the chords and the dynamic process in consecutive chords. There is a simple rule for defining such relations:
The tension increases with every active step, i.e., with every step
moving away from the relevant tonic.
The tension decreases with every passive step, i.e., with every step
moving toward a resolution.
Thus, in the simple harmonic progression with a perfect cadence, the
subdominant chord (or its substitute) as the active step represents the
highest harmonic tension, the dominant chord with its clearly determined
tendency to resolve follows with less tension, and the tonic is most relaxed.
This natural design is enhanced whenever a function appears as a seventh
or ninth chord: a IV7 carries even more tension than the simple IV; a V7
tends even more strongly toward its resolution. Last but not least, the
degree by which the harmonic tension (and with it, the dynamic level that
depicts this tension) increases is related to the harmonic steps audacity.
29
Secondary features
Beyond these basic considerations of chordal relationships there are
two secondary features that must be given special attention since they will
regularly repeal the rule mentioned above: the sequence and the pedal note.
The term sequence describes the process by which a modelthis can
be a melodic unit or, in the case of a harmonically-determined work, a
group of chordsis repeated once or several times on different pitch
levels. In such a case, the primary dynamic process is always established in
the model. The sequential groups must follow the same pattern, or else the
relationship would not be comprehensible. The model thus sets an example
that will be observed in each sequence, regardless of the actual harmonic
relationships there.
The pedal note also creates a law of its own. In the majority of cases,
it comes as a bass note, usually entering on the dominant of the home key,
which is sustained (or, more common on instruments other than the organ,
reiterated) for several measures while the harmonies made up by the other
voices sound reluctant to surrender to the truth that the piece is soon to
come to an end. Pedal notes have their roots in Baroque organ music. Yet
while a sustained or reiterated note on the organ will objectively maintain
a constant dynamic level, subjectively or psychologically it will work its
way ever deeper into the listeners consciousness. This is the reason why
a basically simple eventthat of a sustained or reiterated noteinsinuates
a gradual, smooth but persistent increase of tension.
Finally, any truly melodic event overrules the processes of chordal
tension in pieces determined by harmonic relationships. If a melodic unit,
however small, has enough character of its own, it may create motions that
are momentarily independent from the underlying harmonic process.
1.4b How do you deal with preludes

determined mainly by motivic elements?
The relevant motifs, their character and dynamic design
Finding the motifs on which a piece is built would seem fairly easy.
However, their scope should be determined carefully to guarantee a correct
idea of structure and phrasing.
Remember that the beginning of a motif may occasionally be varied in
subsequent entries. The ending may also include some pitfalls: on the one
hand, the final note may feature various versions that all belong, as a
rhythmically necessary element, to the motif; on the other hand, notes
30
repeatedly succeeding the motif may be no more than an extensionand

thus not generate anything of their own. They may not belong to the motif
at all, and should then be clearly distinguished in color and/or phrasing.
The basic character of a motif results from the same characteristics as
in all other polyphonic compositions in this style: A predominance of
stepwise motion (interrupted, if at all, only by a high tension interval)
evokes a rather calm basic character; so does a complex rhythmic structure
with a variety of note values, including syncopations and/or dotted and tied
notes. Conversely, frequent leaps of a fourth or more and broken chords as
part of the motif indicate a rather lively basic character; so does a simple
rhythmic structure. Spelled-out ornaments are another indicator for lively
basic character.
Within each motif or figure, any of the following details can trigger a
rise in tension: an active harmonic step, an appoggiatura, a leading-note, a
high-tension interval, and a syncopation or other outstanding rhythmic
feature.
As far as the development of motifs or figures is concerned, the most
frequent processes are
the sequence (a unit is repeated on a different step of the scale,
usually with some of its intervals adjusted),
the imitation (a unit is repeated in another voice),
the inversion (a unit sounds upside down), and
the partial sequence, varied sequence, or extended sequence (the
original idea or a recognizable fragment of it recurs, changed in
shape and/or length).
The overall expressive value created by each of these processes can be
described by another rule of the thumb: Rising sequences usually cause an
increase of tension while falling sequences cause a decrease. Imitations,
variations and inversions have in themselves no influence on the increase or
decrease of intensity. Stretto formations and parallels represent heightened
emotional intensity. Abridged and extended sequences tend to represent a
loosening of the grip of tightly organized material, thus a relaxation.
The development of tension
For the overall dynamic design, you would consider similar factors as
in a fugue. The development of tension is determined above all by the
density of prominent material. This density can be achieved horizontally or
vertically. (For more details refer to the discussion under point 2.7 below.)
Another factor is the change of mode or, rarely, the change of character
owing to inversion.
31
1.4c How do you deal with preludes

that follow the outlines of an invention or a fugue?
If you find that your prelude shows the determining features of a fugue,
please refer to the information provided below (paragraphs 2.1-2.7). Most
of the information found there also applies if your prelude is an invention.
However, the following differences between the two standard models exist,
in terminology and in design.
The main idea of an invention is commonly not called subject but
motifalthough any term will of course serve its purpose in furthering
your understanding. The rules concerning an expected order and grouping
of entries in a fugue do not apply to an invention. After the initial motif has
been stated at the outset of the composition (with or without any accompaniment), there will be a number of sequences and imitations. The motif
may subsequently be shortened to a less individual figure or modified to
engender new combinations of its two halves. It may or may not give way
to secondary motifs. In the course of this neutralizing process, it will
modulate to either its dominant or, if the home key is in the minor mode, to
its relative major key, and come to a transitory halt in a perfect cadence.
The second section of a prelude in the style of an invention will begin in
this new key, often re-exposing the material in a way similar to that heard
at the opening of the piece, but frequently by inverting the voices. After
this, almost any continuation seems acceptable in terms of structure if we
take Bachs own two- and three-part inventions as a guide. With regard to
the material accompanying the main motif, too, everything is possible
from notes or chords representing nothing but a harmonic pattern through
simple quasi-independent yet not very individual lines and up to countermotifs with all the characteristics of what would be the counter-subject in
a fugue. Even stretto imitations, although infrequent, may occur.
1.4d How do you deal with preludes

determined mainly by rhythmic patterns?
Rhythmic patterns can be thematic in a way similar to motifs: whenever a particular sequence of note values is significant enough to be easily
recognized, independent from other sequences, and repeatedly taken up,
we speak of a rhythmic pattern. It should usually come in varying pitch
arrangements or we might just as well identify it as a motif. Furthermore,
the sequence of note values combined to a rhythmic pattern is typically
32
determined by two factors: by its metric position (i.e., it may begin on a

downbeat or be conceived as from a weak to a strong beat), and by its
length. If both the metric position and the length of two patterns within a
composition are identical, we call them interrelated.
The means of development that composers may use when playing with
these patterns are basically the same as those used for melodically determined motifs: variation of detail, inversion, abbreviation and extension,
augmentation and diminution, stretto and parallel. Tension build-ups and
relaxations are initiated mostly by either the pitch level of such sequences
(rising sequences increase tension, falling sequences diminish it) or by the
density of the texture. Moreover, dynamic climaxes can also be created by
features beyond the main rhythmic patterns, especially by an unusual
harmonic step.
1.4e How do you deal with preludes

determined mainly by metric organization?
Compositions perceived as determined by their metric organization
feature a dominant pulse. The dominant pulse can be any note value other
than the fastest one appearing in the piece. You can usually find it by
checking whether any of the other rhythmic values is omnipresent or at
least strikingly constant in one of the secondary voices, or whether it is
given additional emphasis by double stems. (Thus if the first note in each
group of four sixteenth-notes is stemmed in two directions indicating the
participation in a second voice, the pulse of the piece is most likely one of
quarter-notes.)
The secondary features beyond this pulse include, above all, melodic
lines, harmonic progressions, and texture. These are also the only elements
that can bring about tension buildups. Climaxes, however, tend to be fewer
and more gradually prepared than in other pieces, which it not surprising in
view of the meditative mood these compositions generally seek to convey.
Particularly in pieces with a quarter-note beyond 16th-note structure, as
described above, the ornamental nature often allows the piece to flow
without too many ups and downs in intensity.
33
2.1 What exactly is a subject?

Subject is the term used for the leading idea of a fugue. This idea is
always introduced at the very beginning of the piece. At this opening, it
regularly appears unaccompanied, i.e., there are rests in the other voices so
that the listener can gain a distinct understanding of this most important
component of the thematic material. Throughout a fugue, the subject
sounds many times. These appearances are called statements, subject
entries, or entrances.
a) How long is the subject?
If we compare the musical language with the verbal idiom, as we do
when we talk about statements, the structure of a musical phrase would
have to meet a corresponding set of requirements as does a verbal phrase.
In any complete sentence, we expect a certain number of components
without which the message would appear incomplete. And while there are
many ways in which a clause may begin, we usually expect it to end with
a full stop. Corresponding processes determine a musical phrasehere: the
subject. A subject can begin at any point in a measure. When interpreting
a fugue it is worth establishing whether the beginning falls on a weak or
strong beat (or between beats) because this will influence the character of
the entire piece. While the melodic details in a subject abide by no rule
(just as the choice of words in a sentence is not prescribed), the functions
these melodic steps represent in terms of the harmonic progression follow
a certain order (as would the grammatical components of a clause). The
equivalent in music to a full stop in language is the perfect cadence. The
conclusion of a complete musical phrasesuch as the subject in a fugue
is therefore represented by the harmonic resolution from dominant to tonic
(from V to I). In addition, the metric position may play a certain role.
While the conclusion of a phrase could theoretically fall on any beat, subject
endings on a strong beat are strikingly more frequent than extensions into
weak beats.
If you have always determined the end of a subject by comparing its
first entry to two or three later ones (how many notes remain the same?),
you should know that this method is not entirely safe. The final note may
be varied in later entries, as happens frequently when an original ending on
the third degree is eventually replaced by a conclusion on the keynote.
Another pitfall: the notes immediately following the end of the subject
without actually belonging to it may appear similarly after some of the
later statements and thus mislead you.
34
b) What is a phrase, and what are sub-phrases?

Continuing our comparison of a phrase in music with a clause in verbal
language, we can state: Just as a sentence may consist of a single clause or
contain sub-clauses of different order, so can a musical statement be of
simple or more complex structure.
The existence of sub-phrases can most often be detected by looking for
one of the following three features: sequences, changes in pitch level, and
changes in rhythmic pattern. It also helps to think of singers or wind
players. If they would breathe somewhere during the course of the subject,
then this is most likely the point of phrasing.
Whenever a subject consists of a single indivisible phrase, the musical
tension unfolds in a single, unbroken rise and fall. Whenever a subject
contains sub-phrases, its message is only correctly conveyed if the musical
line is structured.
Generally speaking, a fugal subject is always conceived as a unity: a
oneness. Thus even if it does consist of several subsidiary units, these
should not appear as equal in importance or as rivaling segments. Instead,
a subject will always have one predominant center: the focusing point or
climax. In a structured subject, this climax may either be reached in several
consecutive sweeps each of which brings about higher dynamic tension, or
the resolution of tension after the climax may occur in several gradually
descending curves. In other cases, a main clause is preceded or followed
by a prefacing thought or an afterthought.
c) When you report the subjects pitch outline, what are you looking for?
The pitch outline in Baroque polyphonic pieces usually fits in one of
two categories. There may be a predominance of small intervals combined
with only an occasional, single high tension leap. High-tension intervals
include the minor sixth, the minor seventh, the tritone (augmented fourth
/ diminished fifth), and the diminished fourth. Alternatively, the pitch
contour may feature a variety of intervals including several larger leaps
and occasional broken-chord patterns. In this context, groups of shorter
note values often represent written-out ornamental figures, particularly
mordents, turns, and inverted mordents.
d) When you analyze the rhythmic pattern, what are you looking for?
The rhythm of polyphonic compositions can be roughly grouped into
two categories: Its pattern may be simple, featuring two predominant note
values in the relevant material, or it may be more complex, including a
variety of note values, dotted and tied notes, and syncopations.
35
e) What is most important in a harmonic progression?

When asked to examine a subjects harmonic background you can
proceed in two ways: you can either limit yourself to the essential details
you know to be relevant to the buildup and decline of dynamic tension, or
you can thoroughly analyze each harmonic step underlying the phrase. As
the second option is fairly complicated and needs some experience, let us
begin with the first. There are actually only two essential details in a
harmonic progression. Ask yourself: Where is the active harmonic step
that is the step from the tonic chord to the subdominant or its substitute,
i.e., IIV or Iii? Look out for any conspicuous fourth or sixth degrees of
the scale. (In C major, e.g., you would try to find any F or A falling on a
strong beat or on a syncopation.) Ask yourself further: Is this active step
followed directly by the passive step VI (dominanttonic), or is there a
modulation to another key? In the case of a modulating subject, you will
find a raising accidental suddenly cropping up, most often before the
fourth scale degree. If the initial, unaccompanied statement of your fugues
subject shows a sharpening accidental (i.e., an additional sharp in any key
signature with sharps, or a natural neutralizing the final flat in any key with
flats) then this is where the shift is taking place. In this situation, when
determining the end of the subject you are not looking for your original
tonic chord but for the tonic in the new key. (This is almost always the
dominant.)
In case you choose the second option and decide to undertake a more
detailed analysis, here is some help: Most subjects are built on the steps of
a simple progression, i.e., on tonic / subdominant / dominant / tonic in the
setting I IV (I) V I or I ii (I) V I; you would therefore do best to take note
of those chords first. (In C major, these would be: C-E-G, F-A-C or D-F-A,
G-B-D, C-E-G.) Now identify in your subject all those notes that fall
directly on a beat and determine their harmonic background. You will
realize that, apart from the first and fifth degrees of the scale (i.e., C and G
in C major), all notes can be clearly attributed to one of the harmonic steps.
The following peculiarities should be noted: The dominant often
appears as a seventh chord (e.g., G B D F). The basic progression may
appear harmonically ornamented; in this case, a function builds something
like a harmonic inverted mordent with the chord normally preceding it,
e.g.: [I-V-I]-IV-V-I or [I-IV-I]-IV-V-I. And in a minor key cadence, the
tonic and the subdominant (i + iv) are in the minor mode. The dominant,
however, uses the notes of the harmonic minor scale, particularly the
leading note. It is thus regularly a major chord and will therefore feature an
accidental (e.g., in C minor: C-E-G, F-A-C, G-B-D, C-E-G).
36
f) Which are the features likely to increase tension within a phrase?

Several features in the three areas of harmony, melody, and rhythm
may contribute to heightened tension.
Active harmonic steps: these are above all the steps from the tonic to
the subdominant (I-IV) and from the tonic to the supertonic or the
relative minor of the subdominant (I-II or I-ii).
Appoggiaturas: these are notes that withhold or delay the
melodic resolution into the main harmonic step. Whenever the unaccompanied subject seems to imply a change of harmony on a weak
beat, it is wise to double check the harmonic background compared
with that in later subject statements. You will probably find that the
harmonic change occurs on the strong beat, but that a melodic note
is purposefully lagging behind, thus creating extra tension.
Leading-notes, particularly artificial ones: A natural leading-note is
a degree of the scale that is a semitone neighbor to a tonic-chord
note; it thus has a tendency to lead into it. (In a major scale the
leading-notes are: the seventh degree, leading up to the octave, and
the fourth degree, leading down to the third, e.g., B-C and F-E. In a
minor scale they are: the harmonically raised seventh, leading up to
the octave and the sixth degree leading down to the fifth, e.g., B-C
and A-G.) Artificial leading notes are semitones created with the
help of an accidental (e.g., in C major: F-G).
High-tension intervals: these include above all the minor sixth,
minor seventh, tritone, and diminished fourth.
Syncopations or other rhythmic prolongations: these are deliberate
distortions of the metric order in which a weak beat is prolonged (by
a dot, a longer note value, or a tie) in such a way that it ingests the
following stronger beat and with it, its accent.)
2.2 What is the importance of the subject in the fugue?

It has often been observed that it seems wrong to say a fugue has a
subject; one should rather state that there is a subject that has generated
a fugue. The perfect little musical entity we call subject is in fact at the
origin of the fugue. Its companions are dependent on it to the largest
imaginable extent: were the subject any different, they too would not be
what they are. The subject is responsible for the feelings of density and
relaxation in the fugue, and it is the main force in creating structure.
Whenever it rests for a while, its absence is distinctly felt.
37
Although this basic truth is valid in all fugues, the degree of impact
exercised by the subject on its surroundings may vary slightly in each case.
Some of the common constellations are: the subject can be spread regularly
across the fugue and always be accompanied by other prominent musical
ideas. Thus, although leading, it will appear as one in a group. As the
subject retreats momentarily, other characteristic motifs may develop.
Thus, the subject may appear as the leader of one team that is contrasted
with another (admittedly less important) team representing a different
color. Last but not least, the subject may not have the same importance at
all times but, e.g., demand more and more attention as the fugue develops.
It may do so by presenting its statements in more powerful variations, or
by appearing in several voices almost at once.
a) When you discuss the subject statements, what are you looking for?
A fugue may contain any number of subject statements. The twentyfour entries in the opening fugue of the Well-Tempered Clavier represent
by no means the largest number. These statements are usually named after
the voice in which they sound. The following conventions are practical:
In a 4-part fugue, the established names for the voices are those of
the vocal ensemble: soprano, alto, tenor, bass (abbreviated S A T B).
In a 3-part fugue, using a selection from this 4-part combination
would seem arbitrary. A good solution is to refer to the voices as
upper voice, middle voice, and lower voice (abbreviated U M L).
In the few 5-part fugues, the least problematic solution is to count
them from v1 to v5 (i.e., from first or uppermost voice to the fifth or
lowest voice. The attempt made by some analysts to retain the vocal
terms by introducing alto I and alto II or tenor I and tenor II has led
to lengthy arguments among scholars. Such a quarrel seems somehow quite beside the point, and the v1-to-v5 option may prevent
further disputes.
b) What kind of changes can you expect to occur in the subject?
In the course of the fugue, the subject may appear in various guises.
This is important to remember when trying to locate all the statements.
One of its intervals may be modified to adjust to a different
harmonic background. In many fugues, this is the case in the second
and fourth entries, i.e., in the answer, the entries beginning on the
dominant. An entry with such an interval adjustment is called a
tonal answeras opposed to the real answer in which all
intervals remain unchanged.
38
The end or the beginning of the subject (or both) may feature
variations in pitch and rhythm. The most frequent modifications are
prolonged or shortened first notes and delayed final resolutions.
The subject may appear partly or entirely in rhythmic variation (e.g.,
showing dotted-note figures where the original was in regular motion)
or in metric variation (e.g., beginning on a weak instead of a strong
beator vice versaand continuing slightly off beat).
The subject may appear upside down. This is called an inversion.
Finally, the speed may appear changed in such a way that all note
values are doubled and the subject is twice as long. This modification is called the augmentation. Also possible, although less
frequent, are entries in diminution where the time value of all
notes is decreased, usually by 50%, and the subject therefore takes
only half the time it did originally.
c) What are stretto and parallel?

There are certain rules dictating the way in which subject statements
follow one another. These are stricter at the beginning of the piece, perhaps
to guarantee the opportunity for the listener to distinguish all aspects of the
material. Thus in the first section of a fugue each voice is expected to wait
for the conclusion of the previous statement. Only then is it allowed to
launch a new entry. Later in the fugue, however, it may occur that a voice
becomes impatient and embarks on a subject statement while another is
still in the middle of it. This is called a stretto; it always indicates
heightened tension. Strettos may occur between two or more voices, and
between the original and varied statements of the subject.
As polyphonic compositions stress the independence of all voices, a
parallel where the subject sounds simultaneously in two voices is a very
special feature. If it does occur, a parallel statement indicates a very high
state of exaltation. Like the stretto, it can also involve different versions of
the subject.
Talking of grouped entries, a final means of intensification is the
repeated entry in which one voice, after having stated the subject, does
not give way to another voice but immediately launches the subject all
over again. It usually does so either on a different harmonic step or by
using any of the above-mentioned variationsas if to add a new side or
aspect to its argument.
39
2.3 What exactly are counter-subjects?

A counter-subject is a more or less consistent companion to the
subject. It may enhance the character of its leader or rival it. But, like a
faithful companion, it is bound to the subject in such a way that it will not
usually appear in its entirety while the subject is resting. Note that there are
fugues in which a prominent musical idea is introduced only after one or
more sections have passed, but will then sound both against the subject and
on its own. In this case we are dealing with a second subject. Occasionally
you may even find a third subject. Fugues featuring more than one
subject are accordingly described as double fugue or triple fugue.
In a regularly built fugue with only one subject, the first countersubject will be introduced in the voice that launched the piece, accompany
ing the subsequent subject entry. Similarly, the second counter-subject (if
the fugue features such an additional companion) can be expected to enter
for the first time against the third subject entry. One possible beginning of
a four-part fugue is therefore (S = subject, CS = counter-subject):
soprano:
S
alto:
S
CS1
tenor:
S
CS1
CS2
bass:
S
CS1
CS2
CS3
There are some, though not many, fugues with no counter-subject at all.
This happens if the subject is so dominant that it does not accommodate
any rivaling ideas. In other cases, a counter-subject may accompany only
a few subject statements and then disappear, never to be heard again.
When trying to find out whether a line qualifies as a counter-subject,
you should ask yourself the following questions:
Does it recur? A melodic detail that only sounds once will not be
given a formal name.
Is it independent? A line that duplicates part of the subject in
either rhythm or pitch pattern is not regarded as independent.
Is it melodic? A neutral scale segment or a melodically meaningless group of notes would not be recognized even if it recurs.
a) How can you determine the length of a counter-subject?
The length of a counter-subject will always roughly correspond to that
of the subject. Where exactly it begins and ends can best be decided when
comparing it with later appearances.
40
b) What phrase structure can you expect to find?

The phrase structure of a counter-subject can vary considerably. We
might find anything from an indivisible unit to multiple segments with
sequences or other combinations of sub-phrases. Note particularly that the
phrase structure of a counter-subject will often not coincide with that of the
subject in answer to the demand for independence.
c) When determining dynamic tension, what features are you looking for?
The development of tension represented in a particular counter-subject
can be initially determined along the same lines as that in the subject.
However, one important aspect to be kept in mind is that, given the request
for independence, the climax of one will, in most cases, not coincide with
that of the other. For performers, this is very important to remember. It is
vital to play each element with dynamic independence, despite the seeming
difficulty of the task. Parallel dynamic movements, although easier to
perform, often make it hard if not impossible for listeners to distinguish the
components in a polyphonic texture. This is a pity since helping the
audience understand is the essential task of an interpreter.
d) What should a sketch featuring the phrase structure and the dynamic
tension in the primary material of the fugue show?
Sketches, though certainly rather tedious to draw, are an invaluable
help for two reasons: designing them makes things clearer in your head,
and the visual image is a much better guide during the performance than
some intellectual concept.
Write out the subject juxtaposed with all its counter-subjects, using
a separate staff for each.
If any part of the material does not consist of a single unit, add a
little hook or tick after the last note of each sub-phrase.
Now draw underneath each staff the opening and closing hairpins
that in music represent the increase and decrease of tension.
2.4 How do you find the episodes in a fugue?

The measures in a fugue where the subject is temporarily resting are
called subject-free passages or episodes. A good way of marking them
is by numbering them throughout the fugue (E1, E2, etc.). Where their
range is to be remembered on paper, mm. 5-7 may sometimes be clear
enough, whereas in other cases you may wish to specify mm. 53-71.
41
a) How many episodes can you expect?

The number of episodes in a fugue is not restricted. Theoretically,
there could be an episode after each subject entry, although this is not often
the case. Also, there are quite a few episodes that, as can be gleaned from
their material, consist of two or more distinct segments. In such cases
composers will often use episode patterns that have been established
earlier in the fugue. Whenever this occurs, special attention should be
given to these episodes as such subdivisions are frequently important in the
structural design of the piece.
b) When ascertaining the material of an episode, what are you looking for?
There are basically three types of episodes. The first type is directly
related to the subject in terms of the material on which it draws. It uses the
head, the tail, or another fragment of the subject, playing with it in
patterns of sequence (same voice, different pitch) or imitation (different
voice, any pitch). Such episodes are often closely linked to the preceding
or following subject entry.
The second episode type is characterized by one or more unique motifs
that set it distinctly apart from the main material. Such an episode motif is
usually a short melodic figure that recurs several times and is characteristic
enough to be easily recognizable. As with all material in a polyphonic
composition, it can move through all the voices and may even be varied or
inverted. The character of an episode motif is determined by the same
features as that of the subject and its counter-subjects: by its pitch contour
and rhythmic features. Motifs that are distinct in themselves and independent of the fugues primary thematic material should sound in a
coloring as different as possible from the main portions of the fugue.
The third episode type fulfills a different function. In some instances,
the stretch of music between the end of one subject entry and the beginning of the next merely extends the subjects cadential close or adds one in
another key. This type is often recognizable either by a typical cadentialbass pattern or a closing formula. Typical patterns for such formulas are:
treble
bass
42
c) In what way can one episode be related to another?

Episodes in analogous sections are often closely related. An episode
may take up the material of an earlier episode literally, varied, transposed,
or with inverted voices. It may add a voice to a thin texture, extend or
shorten the material of an earlier episode, or develop it freely.
d) What are the possible roles of an episode in the development of tension?
An episode can establish three crucial relationships to its surroundings:
it can link two subject statements by leading from one toward the next, it
can be conclusive by resolving tension that was built up during a preceding
subject statement, or it can represent a different register and serve as a color
contrast. The contrasting type may contain independent episode motifs or
motifs related to one of the counter-subjects.
2.5 Performance practice in Baroque polyphonic style

a) The basic character
The basic character of a fugue, just like that of a prelude, is conveyed
through its material and follows the same rule of thumb: a predominance
of stepwise motion and a complex rhythmic pattern indicate a rather calm
basic character, while frequent leaps and/or broken-chord patterns, possibly
combined with ornamental figures in the faster note values, and a fairly
simple rhythmic pattern indicate a rather lively basic character. (You have
already collected all the information needed under 2.1c and 2.1 d above.)
Let us assume that your performance already portrays most aspects of
structure. How do you express abstract aspects like character, mood, and
intensity in a way that every listener will be able to grasp? The predominant means by which a performer conveys character are: tempo, touch, and
articulation. All three work together in shaping and shading the character.
They may do so in different combinations but should obviously not set out
to contradict each other. While touch may enhance slightly different
nuances in each composition, tempo and articulation are directly connected
with the basic character of a piece and can thus be determined more
objectively.
b), d) Tempo, articulation, and ornament realization
For details regarding the adequate choice of articulation and the
execution of ornaments, please refer back to the information on these
aspects given in paragraph 1.3 on pp. 26-27 above.
43
c) How do you decide what relation to choose between the tempo of the
prelude and that of the fugue?
All through the Baroque and Classical era, pieces of music that jointly
form a larger work (i.e., the movements of a sonata or suite, or prelude and
fugue, prelude and toccata, etc.) were conceived as being related with
regard to their pulse. This concept is comparable to the much more familiar
demand of a rhythmic relationship in architecture. Imagine a group of
buildings constructed in the same style but to the order of different patrons
and therefore with unrelated measures (height and width, angle of roof,
pattern of windows and doors, etc.). Such buildings will appeal to the eye
as individual houses, each of them a unit in its own right. A complex of
related buildings, howevera castle, a temple district, or a similarly
structured ensemblewill display a subtle yet very definite relationship
between all its measurements.
As music unfolds in time, which is measured in pulsations, the
equivalent factor of a rhythmic relationship is created by proportions of
the pulses. For this purpose, any of the metric values in a piece of music
can be used as a point of reference: the actual metric value indicated in the
time signature, the larger unit of a half or whole measure, or the smallest
note value appearing in the piece.
A tempo proportion is then created in such a way that one note value of
the first piece relates to any value of the subsequent piece by being either
equally, half, or twice as fast (1:1 /1:2 / 2:1), or three times as fast or slow.
Even an implied but not actually sounding value, like the triplet fraction in
a piece not featuring triplets, can be related to a pulse in another movement. This kind of pulse translation may be preferable particularly in the
case of successive pieces of the same time signature and rhythmic organization, which might otherwise present too little distinction. To give a few
examples (out of many possible ones):
The quarter-note beat of a fugue may equal
a quarter-note in the prelude
an eighth-note in the prelude
a half-note in the prelude, etc.
an entire measure in the prelude
three eighth-notes / three quarter-notes / three sixteenth-notes in the
prelude
a virtual (imagined, not materializing) triplet eighth-note or triplet
quarter-note in the prelude.
44
2.6 What determines the design of a fugue?

As you will have already found out, the subject statements of a fugue
can appear immediately one after another, or they can be separated by
episodes. Whatever the pattern, obviously these elements do not simply
pass by as a shapeless, disorganized chain of events. On the contrary,
Bachs fugues are all constructed with carefully balanced sections. Since
this design or structural layout is slightly different in each fugue and of
vital importance for an understanding of the composition, it is useful to
develop some secure method of analysis.
a) Which of the structural features are basic traits of a fugue?
Which help to determine the particular design?
Remember the nursery explanation of the structure of a fugue?
Its like a conversation:
One speaker brings up a topic.
Friends pick it up one after another.
The first round ends once each person has added his or her
view. At most, one of them may add a summary remark.
(In a fugue, this is called a redundant entry.)
After a relaxation, with or without some contrasting talk,
the discussion of the main topic resumes.
In each consecutive round, these are the rules:
The minimum number of speakers is two. (A monologue is
no conversation!)
Nobody should speak twice unless summing up a round or
supporting another persons statement. (In intensifications,
an additional voice may reinforce another without making a
statement of its own. Thus stretto, parallel, or repeated
statement are counted as one group entry.)
The maximum number of speakers is again all plus one
(e.g., with four friends it is 4 + 1 = 5, one person being
allowed to comment twice).
This rule of thumb will almost always help to determine the length of
the first section (or exposition). It often even enlightens the entire plan
of the composition, with all its consecutive rounds. Concerning the number
of rounds, you can expect at least two. As for the maximum, there is no
established rule. (There was a time when people, influenced by theories of
the late nineteenth century, tried to detect a three-section structure in every
fugue. Despite the fact that this has long since been unmasked as a
misconception originating in sonata and ternary forms, these theories still
seem to linger in some minds.)
45
As soon as each voice in your fugue has entered with the subject, the
full ensemble is reached. However, while the musical debate relaxes (as
in an episode) or embarks on a fresh start (as at the beginning of a new
section), one of the voices often takes a rest. A subject statement that
sounds accompanied by less than the full ensemble may therefore indicate
the beginning of a new section.
Analogies in a fugue are not necessarily obvious at first sight, although
some are. What is required is that a sequence of events (like a number of
subject entries, accompanied by certain material, perhaps including an
episode) recurs later in the piece in the same or a very similar order. This
similarity does not refer to the position of the material in the ensemble; so
the order may be the same with different voices being used to present the
materialwhich is why some analogies are at first hard to find.
b) What do the harmonic features tell us about the design?
While a larger part of the piece will always appear closely related to
the main key, there are usually passages that show a harmonic digression
to either a different tonal area and/or to the opposite mode (to the minormode area in a major-mode fugue or vice versa).
To determine the harmonic background of a subject entry, it is best to
look at both its beginning (i.e., the first two or three notes, to avoid
misinterpretations in tonal answers) and its end. Subject entries within one
section often appear in tonic / dominant / tonic / dominant progression. A
change in this pattern is worth noting. Significant cadences are usually
those that appear as obvious formulas or cadential-bass patterns outside the
main material of the fugue.
c) How can you sketch the design of a fugue?
Such a design is usually sketched using a horizontal column for the
measure numbers, corresponding horizontal lines showing the material
sounded in each of the voices, and vertical columns for the juxtapositions.
In the colored graphs included in this book, bricks of different shade and
intensity represent subject and counter-subject entries, roughly like this:
mm. 1
S
A
T
B
S
S
CS1
5
S
CS1
CS2
7
8
S
CS1
CS2
CS3
10
episode
(rest)
etc.
46
2.7 Which factors determine the development of tension?

The development of tension is the way in which the intensity, density,
and dramatic vigor develop, between the first and the last note of the
composition. Some fugues begin by presenting their material in a very
condensed manner. Later, however, the subject takes frequent rests,
appears in a weaker mode or in weaker surroundings (e.g., with the other
voices resting or sounding neutral lines, instead of counter-subjects). In the
initial sections of other fugues, the material is introduced in more of a
matter-of-fact way. It is only after the exposition that more and more
intensity starts building up. Yet other fugues are conceived along symmetrical lines, etc.
a) What do you take into consideration when trying to decide whether or
not (and how much) the tension rises within a section?
The development of tension within any segment of the fugue is determined above all by the density of prominent material. This density can be
achieved horizontally or vertically. Horizontally, a group of subject statements in which the entries follow each other without delay will create
more urgency than scattered statements interrupted by episodes. Even
more so, both the rare but very effective repeated entry and the subject
augmentation invite special attention since the statement is doubled in
length. Vertically, a subject entry that comes accompanied by characteristic
counter-subjects in several voices creates more intensity than one that has
only one serious contestant, or one that is surrounded merely by neutral
lines. Particularly effective in the buildup of tension are strettos and
parallel entries; they create an effect of breathlessness or of forces joined
in battle.
b) When discussing how the sections of your fugue relate to each other in
terms of tension, what are the relevant factors?
The way in which the sections within the whole composition relate to
each other depends above all on the factors mentioned above: the density
of prominent material and the harmonic progression. Note, however, that
the minor mode does not always soften the character of the material, just as
the major mode does not necessarily convert an introverted atmosphere
into one more extroverted.
WTC I/1 in C major Prelude

The first prelude in Bachs Well-Tempered Clavier consists of nothing
but broken chords. It must therefore be described as determined purely by
processes of increasing and decreasing harmonic tension. Effects such as
creating an echo in every repeated chord or emphasizing the chords peaknotes and thus drawing attention to an imaginary melodic line divert from
the composers intention. They would misguide listeners and should thus
be avoided.
The sections are clearly discernible by their simple cadential patterns.
The first cadential closure is reached
in m. 4, the steps leading to it being:
6
m. 1 = I, m. 2 = ii2, m. 3 = V5 , m. 4 = I. Since all structural breaks in this
prelude are harmonically defined, this cadence must be interpreted as the
end of a first (short) section. The next harmonic progression concludes in
m. 11. As the sharps from m. 6 onward demonstrate, Bach now modulates
to G major. The final steps of this progression are: m. 9 = ii7 of G, m. 10 =
V7 of G, m. 11 = I of G. Once again, this harmonic close marks a structural
caesura.
There are altogether four sections in this prelude:
I
mm. 1-4
complete progression in C major
II
mm. 5-11
modulation to G major
III
mm. 12-19
modulation back to C major
IV
mm. 20-35
complex, extended cadence in C major
While no portion ever recurseither note for note or in variationthere is
a transposed passage that forms part of a structural analogy: mm. 15-19 are
an exact transposition of mm. 7-11 (see the progression I6, IV2, ii7, V7, I
in mm. 7-11 in G major, in mm. 15-19 in C major). In connection with this
transposition we also discover that mm. 7-8 are a sequence of mm. 5-6 and
mm. 14-15, a sequence of mm. 12-13. In both cases, a two-bar model is
repeated one diatonic step lower. The only structural divergence between
these two patterns results from the different amount of overlapping with
the transposed portion. We can thus confirm the following analogy:
section II (mm. 5-11) corresponds with section III (mm. 12-19);
both consist of
a model of two measures, a descending sequence,
and an analogous cadential close.
49
50
WTC I/1
All aspects of performance in this prelude reveal themselves by means

of this basic analysis. As the composition contains no melodic elements,
there are no open questions about articulation or single-voice phrasing; all
notes should be sustained for the exact length of their written value.
Phrasing between structural sections is best made transparent by varying
the degrees of tension. A performer who works on a modern piano will
achieve this through dynamic means. (The phrase-ending ritardando, so
essential on early keyboard instruments, can be restricted to a minimum on
instruments capable of expressing phrasing through dynamic shading.)
The dynamic development between consecutive notes poses the biggest problem. What should be heard in a piece determined by harmonic
processes is the relationship between chords, not that between 16th-notes;
the notes forming each chord must therefore sound as equal as is humanly
possible, and greatest care should be taken to avoid any emphasis on the
uppermost pitches.
A very important factor for any sensible performance of the C-major
prelude is tempo. This should again be chosen with a view to best convey
the harmonic processes. Too fast a rendition easily diverts the listeners
attention toward an apparent display of virtuosity; too slow a tempo makes
it difficult to hear more than just one chord at a time.
The only ornament in this prelude derives from an early copy. It appears
in m. 34 and serves to emphasize the final cadence by embellishing it. The
sign indicates a simple mordent that, since it is approached stepwise,
begins on the main note: E-F-E.
As was stated above, the C-major prelude derives its expression from
its harmonic progressions. There are, however, two instances in which
these are overshadowed by secondary processes. One case is owed to the
above-mentioned sequences, the other to a protracted pedal note. The
development of tension as represented by all features is as follows:
In the course of the first simple cadence (mm. 1-4), the subdominant
function commands the greatest tension. This tension subsequently finds a
stepwise resolution via the dominant toward the tonic. The dynamic analog
to this process may be described as p - mp + - mp - p.
The following section (mm. 5-11) brings a modulation to G major,
which is reached in m. 11. As the section opening contains a sequence, the
priority of mere harmonic relationships between consecutive chords is
temporarily suspended. Within the first two measures, the harmonic relationship between the inverted A minor and D7 chords (mm. 5, 6) constitutes
a rather strong decrease in tension. (This can easily be experienced when
listening to the two harmonies, preferably when played as block chords).
C major
51
In keeping with the law defining dynamics in sequences, the same relationship must apply to the following two chords, on a softer level because the
sequence descends. From the last chord of the sequential pattern onward
there follows a further gradual release of tension. The dynamic analog to
the process in this portion of the prelude may be expressed in these terms:
mf + - mp + - mf - mp - mp - p+ - p.
A similar development occurs in the third section. Like the preceding
one it starts with a sequence, and again, the relationship between the first
and the second chords in the model is one of relaxation. The release of
tension is even stronger here than in the earlier sequence as mm. 12 and 14
each consist of a diminished seventh chord resolving onto an inversion of
the supertonic (m. 13) and of the tonic itself (m. 15). At the end of this
sequence the harmonic tension therefore appears already abated to something very close to the softest shade used in this piece. The following four
measures, a transposition of mm. 8-11, should portray a dynamic outline
exactly like the one in the corresponding measures so as to assist listeners in
grasping this analogy. The concept of mm. 12-19 is thus approximately:
poco f - mp + - mf + - mp - mp - mp - p + - p.
The fourth section is almost as long as the first three sections together.
The emergence of the dominant pedal in m. 24 serves to divide it into two
subsections. The first of these subsections, from m. 20 to the downbeat of
m. 24, ends in an imperfect cadence. The harmonic development sets out
from the C-major seventh chord (m. 20) but then leads away from the tonic
in bold steps. The process includes two diminished-seventh chords; in
addition, there is a hint of an independent bass line, which enhances the
sense of urgency. Upon closer inspection, the end of this bass line reveals
a circular movement preparing the beginning of the pedal note by sounding
both the natural leading-note (from the semitone below) and the artificial
one (from the semitone above) to the dominant keynote G. The audacity of
these harmonic steps can best be conveyed in the following dynamic plan:
mp - mp + - mf + - poco f +.
After this forceful increase, the next subsection begins with a sudden
hush to a level hardly above piano. From here the pedal note takes effect. Its
typical gradual and smooth dynamic growth continues not only while the
bass remains on G but all the way to the end of the piece, thus closing the
prelude on something like a triumphant forte chord. The dynamic balance
in this section may therefore be represented quite convincingly as follows:
p + - mp - mp - mp + - mf - mf - mf +- poco f - poco f - poco f + - f - f.
52
WTC I/1
The graph attempts to show the processes determining this prelude.1
WTC I/1 in C major Fugue

The subject of this fugue is one and a half measures long. It begins
unaccented, after what can be called an implied breath on beat 1. This
creates the impression of a very long upbeat, in which the tension toward
the following downbeat is smooth and not very strong.
In its original statement the subject ends at m. 23 on E. That this E has
to be regarded as the final note of the subject becomes obvious when we
compare the first entry with later ones: they either end after this E or its
equivalent in another key (see, e.g., mm. 9-10, 10-12), or they continue in
a different way each time. The closure is further supported by the subjects
harmonic background: the dominant is reached at the beginning of m. 1
1
Carl Czerny added an extra measure. He argued that first, it seemed highly improbable, for
a composer as conscious of subtle numeric balancing as Bach was, to have written a piece
consisting of the uneven number of 35 measures; second, a bass line with a diminished third
such as the one in mm. 22-23 seemed melodically incorrect and would therefore require a
chromatic link. Czerny therefore inserted a tonic six-four chord to correct Bachs error,
and by doing so destroyed the forceful tension built up here with so much ingenuity.
C major
53
(on either D or G; the dominants point of arrival varies throughout the

fugue) and resolves onto the keynote-chord on this E. The fact that the first
statement is followed by an extension, made up of two sequences of the
last four notes, does not change its basic confines but serves to grant a
smooth transition from this first entry to the next.
Ten of the subjects thirteen intervals are seconds, and stepwise motion
also dominates the fugue as a whole. The few larger intervals, however,
confound the picture somewhat because of their particular nature. They do
not, as might normally be expected in the context of overall stepwise motion,
represent high-tension intervalsintervals whose expressive gesture is
regarded as blending particularly well with the character expressed by
stepwise motion. Instead, the three leaps interrupting the subjects smooth
line are two perfect fourths and a perfect fifth, i.e., leaps without particular
emotional content.
The rhythmic structure is complex. The subject features four different
note values: eighth-notes, dotted eighth-notes (or, in m. 2, the same value
written with tie prolongation), 16th-notes, and 32nd-notes. Syncopations
are a regular feature throughout the fugue; in fact the fugue comprises only
six measures not displaying at least one syncopation.
The question of phrasing in the subject allows for two competing
answers, both of which can be supported with evidence from within the
composition itself. The choice between them is thus one of individual conviction or interpretation: (a) The overall pitch pattern, showing ascending
steps at the beginning, descending steps at the end, and leaps in the middle,
supports the option that this subject is conceived as an indivisible phrase.
The absence of obvious melodic sequences confirms this view. (b) For
interpreters who feel that rhythm plays a crucial role in this subject, there
is a pattern that is repeated in slight variationa sequence-like process
revealing the subject as made up of two halves. The first half consists of
four eighth-notes moving upward in stepwise motion. They are followed,
after a rhythmic prolongation, by shorter note values leading downward.
The second half starts similarly with four eighth-notes that, this time, move
in jumps up and down but are also followed, after the same rhythmic
prolongation, by a group of shorter note values leading downward. The
phrased subject would thus be defined by the analogous rhythmic structure
of its two segments. This option finds support in a harmonic progression
that allowsalthough Bach does not
always compose thusfor two complete cadences coinciding with the two
V
rhythmically determined halves (see
V V I vi ii
54
WTC I/1
the harmonic progression underlying the subject statements in mm. 2-4 and
5-7). As far as the subtler harmonic progressions from one eighth-note to
the next are concerned, there are so many variations throughout this fugue
that it is not possible to state one unequivocal solution.
Ludwig Czaczkes in his analytical
book on Bachs WTC made his own
choice when he reconstructed in fourpart texture what he regarded as this
subjects harmonic basis.
When deciding on subtle dynamic
processes in the subject, the aspects to
be taken into particular consideration
are the two rhythmic phenomena and their respective harmonic functions:
the syncopation is rhythmically stronger, but it falls on the harmonically
rather weak tonic-to-dominant movement; the dotted eighth-note is rhythmically weaker but supported by the stronger harmony, the subdominant.
The preference of one climax over the other therefore rests on whether
interpreters feel the harmonic process more strongly than the rhythmical
one, or vice versa.
If a performer chooses option (a) for the phrase structure and reads the
subject as a unit that allows for no further subdivision, only one climax has
to be determined, which will come smoothly prepared by an increase in
tension and be followed by a relaxation. Performers perceiving the subject
as made up of two halves (b) have to decide which of them is stronger and,
consequently, which of the two shorter tension build-ups is more powerful.
In the latter case the subject would contain two releases, with the E at the
end of the first subphrase as a transitory solution, i.e., not quite as soft as
the final E.
The design of the C-major fugue has often been called programmatic:
as the first of a collection of twenty-four fugues it contains twenty-four
subject statements. These appear as follows:
1.
2.
3.
4.
5.
6.
7.
8.
mm. 1- 2
mm. 2- 4
mm. 4- 5
mm. 5- 7
mm. 7- 8
mm. 7- 8
mm. 9-10
mm. 10-12
A
S
T
B
S
T
A
B
9. mm. 10-12
10. mm. 12-13
11. mm. 14-15
12. mm. 14-15
13. mm. 15-16
14. mm. 15-16
15. mm. 16-17
16. mm. 16-18
A
T
A
T
B
S
S
A
17.
18.
19.
20.
21.
22.
23.
24.
mm. 17-18
mm. 17-18
mm. 19-20
mm. 19-20
mm. 20-22
mm. 21-23
mm. 24-25
mm. 24-26
T
B
T
A
S
T
T
A
C major
55
Three of these subject statements are varied: The bass entry in m. 17

begins with a rhythmically doubled first note, annihilating the subjects
characteristic up-beat character. The tenor statement in mm.14-15 shows a
variation at the end, after the syncopation. The soprano entry in mm. 15-16
states only the subjects first half before giving way to a new beginning,
which now unfolds as a complete statement in this voice.
While parallels of the subject are not used in this fugue, stretto is a
prominent feature. After the first four entrances, i.e., as soon as each of the
voices participating in the fugue has entered, subject statements overlap
more often than not. The nine stretto combinations that Bach uses are listed
below. (The normal-sized capital distinguishes the respective group leader,
i.e., the voice that counts as building the round. The leader in a stretto
can be detected by asking the following three questions: which voice
presents the version that is more faithful to the original?, which serves to
establish a new key?, or simplyin the absence of other distinctions
which voice enters first?)
Strettos begin in measures
involving voices
7
10
ST, BA
14
15
16
17
AT, BS, SA, TB
19
20-21
TA, ST
24
TA
Interestingly, the first six stretto combinations give a complete account
of the six alliances possible between four voices: the soprano is paired with
the alto (entries 15-16), with the tenor (5-6), and with the bass (13-14); the
alto is further paired with the tenor (11-12) and with the bass (8-9); finally,
tenor and bass are also paired (17-18).
The C-major fugue does not feature any counter-subject. In view of the
immense density of material created by the twenty-four subject entries, this
will hardly come as a surprise. In the absence of counter-subjects, this
fugue claims full attention for its strettos. A sketch of the phrase structure
and the dynamic design should therefore depict the pattern created in such
group statements.
In the two examples below, the entries in mm. 14-16 are rendered in
alternative interpretations: the first suggestion shows what happens if the
subject is seen as an indivisible phrase climaxing on the syncopation; the
second excerpt interprets the subject as consisting of two subphrases, with
the climax of the first subphrase stronger than that of the second.
56
WTC I/1
Only twice is the density of the material in this fugue briefly interrupted. As the subject is absent in these measures, they qualify as episodes:
E 1 = m. 13 (last three eighth-notes) to m. 14 (first eighth-note),
E 2 = m. 23 (after the first 16th-note) to m. 24 (first eighth-note).
Furthermore, the fugue ends with two subject-free measures: the last
statement concludes on the first 16th-note of m. 26, thus: E3 mm. 26-27.
None of these subject-free measures is even remotely related to the subject;
neither are there any episode motifs. The two subject-free passages within
the fugue are both clearly identifiable as cadential closes. The second half
of m. 13 features typical closing formulas in all three sounding voices,
leaving no doubt that something is about to come to an end here. In mm.
23-24, soprano and bass once again present those typical patterns while the
inner voices join in neutral, quasi-chordal style.
Surprisingly, the ending of the composition avoids all these established
formulas. Two measures earlier Bach had already started a tonic pedal,
thus making it clear that the piece is approaching its end. Now he allows
C major
57
the other three voices to continue freely, with harmonic and melodic developments independent of both material bounds and pattern restrictions. One
after the other the voices then come to a halt: the tenor withdraws first,
upon which the alto becomes less eloquent. The soprano is the last to
remain active by running up to the high C.
The final note in the soprano, the high C, deserves a mention. This note
has not been previously reached in this composition. As is widely known,
it marks the upper limit of the keyboard instruments in Bachs time, just as
the preludes final low C marks the confines at the other extreme. These
two tiny details are another hint that Bach, in writing his twenty-four
preludes and fugues, aimed to demonstrate in full all that was possible on
this instrument.
The dynamic role each of the three subject-free passages plays in the
development of this fugue will have become apparent from what has been
said above. The two inner episodes serve as cadential closes and thus have
a relaxing tendency. The two final measures, on the contrary, succeed in
creating somethinghowever smallof their own, thus contributing to a
build-up toward a final climax.
Both the overall stepwise motion and the rhythmic structure with its
four different note values and frequent syncopations suggest that the basic
character of this composition is rather calm. Within the overall frame of
this character, the chain of three leaps in the subject adds an element of
contrast that should be feltand playedas such.
The most appropriate tempo is a calmly flowing one: flowing enough
so that the quarter-notes, rather than the eighth-notes, are felt as a pulse;
calm enough so that the 16th-notes still sound serene. Regarding the
relative tempo of the fugue to its prelude, retaining the beat would lead to
a somewhat dull result since both the time signature and the 16th-note
motion are the same in both pieces. Therefore, a proportion of 3:2 or 2:3
seems better suited. [Approximate metronome setting: prelude beat = 80 or
90, fugue beat = 60.]
The articulation in the fugue is mainly legato. One conspicuous (and
regular) exception occurs with the three leaps in the subject, which, being
consecutive jumps, must be detached. Similarly, leaps and cadential-bass
patterns in non-subject portions of any voice are also exempt from legato
articulation.
The fugue contains three ornaments, namely in mm. 13, 18, and 19.
The one in m. 18 is indicated in the Urtext as deriving from Bachs own
manuscript. It is a mordent that begins on the upper auxiliary note and
touches down twice on the main note, thus resulting in four regular notes
58
WTC I/1
A-G-A-G. The two other ornaments both embellish typical soprano closing
formulas. As the Urtext printing in brackets indicates they do not appear in
the autograph but were added in various early copies. Ornamenting this
characteristic dotted note in a cadential formula was common practice, and
performers should include some kind of embellishment in any case. The
simple mordent (as suggested in m. 19 by the symbol in brackets) is a more
likely solution in both cases than the compound ornament recommended
by the small-print version in m. 13 since an anticipation of the final note in
the soprano formula (see A-A mm. 13-14 and D-D m. 19) normally replaces
and thus precludesa trill with a suffix ending.
The design of the fugue is clearly laid out. Bach orders the voices
presenting the subject, and the sequence of single and stretto statements, in
a consistent pattern:
mm. 1-7
entries in all four voices, no stretto
mm. 7-14
entries in all four voices, two with stretto partners
mm. 14-19
entries in all four voices, all with stretto partners
mm. 19-27
two stretto pairs including three of the voices, followed
by one repeated (redundant) stretto pair in the coda.
Within these four groups, the coupling of voices in stretto also seems
to follow a plan: the second round joins the higher and the lower voices
in each register (S + T, A + B), the third round groups the adjacent and the
outer voices (S + A, A + T, T + B, B + S), the fourth round leaves out the
bass completely but uses the next lowest voice, the tenor, in all three
combinations.
The complete ensemble is built up four times: The first four entries
produce the regular assembling of all voices involved in the fugue (see
mm. 1-7: from one to four voices). The ensemble is consequently reduced
to three voices in the first stretto statement. In both the following alto entry
and the next stretto, one of the voices is resting, so that the full ensemble is
again reached at the fourth statement of this group (see mm. 7-13, from
three to four voices). After the cadential close in A minor, the ensemble is
momentarily reduced to only two voices. This marks the return to C major
as a genuine new beginning. The following three strettos then sound in full
ensemble (see mm. 14-19, two to four voices). Finally, the stretto in mm.
19-20 that overlaps with the closing formula has the soprano pausing. The
full ensemble is restored with the ensuing statement (see mm. 19-27, from
three to four voices).
The harmonic progression within this fugue leads first from C major
to its relative A minor, confirmed by the cadential close of mm. 13-14. A
new start in C major is diverted after four entry pairs to D minor. This key
C major
59
is again confirmed by a closing formula (m. 19), although the two outer
voices alone hold on to this tonal anchor while the two inner parts continue
with stretto statements of the subject. As can be seen from both the keys of
the stretto group-leaders and the prolonged bass notes (see mm. 19-20: D,
21-22: G, and 24-27: C), the fugue then reverts to the home key, which is
confirmed by a cadential close in C major in mm. 23-24. In terms of the
harmonic development, the last four measures thus appear as a coda, since
the final cadence has already taken place and the tonic bass note sounds as
an extended pedal note.
There are several interesting analogies within the four sections. Both
the first and the second sections feature a harmonic progression launched
from the tonic and followed by entries on the unusual repeated dominant.
mm. 1-5: A (I), S (V), A (V) mm. 7-12: S + T (I), A (V), B + A (V)
The first sections two initial entries relate to their counterparts in the
second section (counting again the strettos group leader as the relevant
voice) in inversion, as do the third and fourth entries of these two rounds.
mm. 1-7: A, S T, B
mm. 7-13: S, A B, T
The first stretto of the third section (mm. 14-15) is analogous to the
first stretto of the second section (mm. 7-8) insofar as both the pitch level
and the distance between the second entry and its leader (two eighth-notes)
are identical. The second stretto in the third round (mm. 15-16) appears as
an intensified variation of the corresponding second entry in the second
round (mm. 9-10), again with the same pitch level. Hence:
mm. 7-10:
mm. 14-16:
stretto (S + T, on C + G)
stretto (T + A, on C + G)
distance 2 eighth-notes
distance 2 eighth-notes
single entry (A) on G
stretto (B + S) leader on G
60
WTC I/1
Finally, the fourth stretto of the third section (mm. 17-19) brings a
modulation into a new key, as does the fourth entry of the second section
(mm. 12-13); in addition, both the group leader here and the single entry
there sound in the dominant of the key of their destination.
mm. 12-13
mm. 17-19
fourth entry of section II
fourth stretto of section III
on V of A minor
leader on V of D minor
The first section shows a gradual, constant build-up of the ensemble
without any interruption by episodes. The tension rises steadily but, in the
absence of additional intensifying factors, reaches only medium level. The
second section, consisting of a twofold stretto-plus-single-entry sequence,
indicates a slight twofold relaxation. The first is supported at its end by the
closing formula (see soprano mm. 9-10: GF-G), while the second is
enhanced by the modulation to the minor key and the ensuing resolution
within the cadential close. The third section builds up tension from two to
four voices in the densest imaginable stretto setting. The dynamic processes are similar to those in the first section but considerably intensified.
Tension reaches its height at the end of this round; here the modulation
with its final turn to D major, the Picardy-third version of the expected D
minor, with the elaborate cadential formulas and the impatiently overlapping first stretto of the ensuing section create a supreme climax. The
fourth section resembles the second insofar as it appears made up of two
halves. After two entries that modulate back from D through G to C, thus
initiating a harmonic relaxation, the cadential measure 23 represents an
obvious caesura. The following coda replaces the expected stretto + single
entry with a stretto over an extended pedal note, but complements this with
two measures of subject-free development.
There are thus striking relationships between the two halves of the
fugue: The first section builds up tension. This build-up extends into the
beginning of the second section but is then not developed further, owing
both to the divided layout of this section and to the regressing harmonic
development. The third section builds up tension toward the overall climax
of the fugue. This elevated tension is continued into the overlapping beginning of the fourth section, after which it subsides gradually, both because
of the divided layout of this section and its softening modulation and
cadential close.
WTC I/2 in C minor Prelude

At first sight most remarkable in this prelude are the striking changes
in tempo, texture, and material. These unusual features are a result of
Bachs decision to upgrade a small C-minor prelude composed for his son
Wilhelm Friedemann. The contrasting sections found in the Well-Tempered
Clavier version represent the insertions the composer added later. In the
original version, the dominant-seventh chord in m. 25 leads directly into
two concluding measures above a tonic pedal.1
In its attitude and expressive aims, the C-minor prelude resembles the
preceding C-major prelude. Any attempt to render this piece as a display of
virtuosity would be counterproductive to the musical idea. Similarly, an
emphasis on the horizontal process in one or both of the voices would fail
to convey the essence of this composition. The main body of this prelude
represents the category of pieces determined by harmonic processes. Just
as in the C-major prelude, each measure stands for a harmonic step whose
relationship to the harmonic surroundings determines the message and
the dynamic representation.
The first harmonic progression ends in m. 4, the steps being:
m. 1 = i, m. 2 = iv64 , m. 3 = vii + C pedal, m. 4 = i.
Since structural breaks in this prelude are defined solely by harmonic
processes, this cadential close must be regarded as indicating the end of a
first (short) section. The next harmonic progressionand with it the next
structural sectionends in m. 14 where Bach concludes a modulation to
the relative key of E major. The final steps of the cadence in this new key
are: m. 10 = V2, m. 11 = I64 , m. 12 = IV6, m. 13 = V65 , m. 14 = I.
There are altogether four such structural sections in this prelude:
I
mm. 1-4
complete progression in C minor
II
mm. 5-14
modulation to E major
III
mm. 15-18
modulation back to C minor
IV
mm. 18-38
complex, extended cadence in C
1
A comparison of the shorter original version of this prelude with the C-major prelude
reveals many parallels: As in the C-major prelude, the first 24 measures of the C-minor
prelude all show a half-measure model followed by its repetition. The extension of the first
section, the chain of descending sequences in the modulating second section, and a
protracted dominant pedal in the final section are all identical in both pieces.
61
62
WTC I/2
As in the preceding prelude, this harmonically-determined piece does

not allow for any melodic articulation or single-voice phrasing since it is
not the melodic element that counts here: all notes are to be sustained for
the exact length of their written value. Phrasing between the structural
sections, i.e., between one closed harmonic progression and the next,
should be conveyed through tension curves. For performers on the modern
piano this is attained through dynamic modifications that correspond with
the harmonic development.
The dynamic relationship between consecutive notes may pose a
problem for many performers. The result one should be aiming for is to
convey a chordal progression behind the single 16th-notes. Thus the notes
forming each chord must sound as equal as possible, and greatest care
should be taken to avoid emphasis of the uppermost pitches.
Choosing the tempo is a complex matter because of the indications
Bach gave for the portions he added to the smaller piece composed for his
son. For the harmonically determined portion alone, the same holds true as
for the C-major prelude: the tempo should be chosen so as to best convey
the idea of harmonic processes. Too fast a performance easily diverts the
listeners attention by what appears as a display of virtuosity; too slow a
tempo makes it impossible to hear more than just one chord at a time.
Next, there is the question of a meaningful tempo balance between the
four sections. A reasonable assumption (and the choice of most musicians)
is to interpret the final Allegro as a return to the original tempo, thus giving
the prelude a rounded form. For the two tempos in between, the following
proportion works well:
Presto = 1 : 2
Adagio = 4 : 1
Allegro = 1 : 2
in mm. 1-27 = in mm. 28-33

in mm. 28-33 = in m. 34
in m. 34 = in mm. 35-38.
This option results in a well-balanced overall structure within the portion

of the prelude that abandons the initial two-part texture. This is what we
will hear:
six measures, on the dominant pedal and between pedals:
mm. 25-27
= 3 measures in the original tempo
+
mm. 28-33 in Presto = 3 measures in the original tempo
six measures, on the (actual or implied) tonic pedal:
m. 34 in Adagio
= 2 measures in the original tempo
+
mm. 35-38 in Allegro = 4 measures in the original tempo.
C minor
63
Another option assumes a 2:3 proportion between the outer sections on

the one hand and the PrestoAdagio on the other hand. While this option
does not maintain the balanced structure mentioned above, it allows for a
simpler transition into the fugue. Approximate metronome settings for both
options are as follows:
(a) beginning beat = 72, Presto = 144, Adagio = 36, Allegro = 72;
(b) beginning beat = 88, Presto = 132, Adagio = 66, Allegro = 88.
The tempo in the fugue is best chosen in complex proportion after a
simple tempo proportion within the prelude or in simple proportion after a
complex proportion within the prelude. This means: (a) 3:4, or prelude
beat = 72, fugue beat = 96, (b) 1:1, or prelude and fugue beats = 88.
Ornamentation in this prelude includes two inverted mordents: one
(contained in the autograph) in the middle of the Adagio measure; the other
(derived from an early copy) on the final note. The former is played with E,
in keeping with the harmony of this measure, while the latter is a whole-tone
ornament. In addition to these embellishments, the two arpeggios in the
Adagio need careful consideration. Is the upper note primarily part of the
(vertical) arpeggio, or is it above all part of a (horizontal) melodic line? If
one plays the note as the score seems to indicate, without any sophisticated
interpretation, then it would sound as the last note in a broken chord. In
Baroque style, such a broken chord would begin on the beat, which would
cause the uppermost note to fall after the beat. The question is therefore
whether melodic continuity would not be destroyed by such a rendition. In
the second arpeggio in particular, the rendition with F as an (after-beat)
target note would interrupt the melodic flow quite awkwardly. In this case
it is therefore preferable to interpret the melodic note as independent of the
chord, i.e., to play the F together with the bass note C on the beat and then
arpeggiate the remaining chord notes. Going back half a measure to the
beginning of the Adagio, the upper note E can probably be perceived in
two ways, either as a note in a chord or as a melodic component. A
rendition similar to that of the second arpeggio is therefore possible (and
perhaps beneficial to the symmetry in this measure) but not necessary for
its understanding.2
2
The practice described above was quite well-known in Baroque times. Performers were
expected to understand, from their immediate grasp of the musical sense, whether or not
an arpeggio included the upper note, and then play accordingly. The fact that there was
obviously no need to specify this in writing may indicate that musicians usually did
understandor, perhaps just as often, that they were keyboard players studying with the
composer and thus could be taught during the lesson.
64
WTC I/2
On two occasions, the harmonic patterns that determine the message

of this prelude are eclipsed by secondary processes: in one instance by
sequences, in the other by a bass pedal. The development of tension in the
entire prelude as created jointly by all these processes can be described as
follows:
In the course of the first simple cadence, the subdominant commands
the greatest tension; this tension subsequently finds a stepwise resolution
through the dominant toward the tonic. The dynamic equivalent to this
process is approximately p - mp + - mp - p.
The following section begins with a string of sequences. A first question
is therefore: what kind of relationship exists between the two measures of
the model. If, e.g., the step from the A-major chord in m. 5 toward the D7
chord in m. 6 is taken as active, this model sets the pattern that the next
two-bar combinations must followon a generally softer level because the
sequence is descending. The section is rounded off with a cadential close
in the relative key, E major. The dynamic outline may be expressed as
mf - mf + - mp + - mf - mp - mf - mp - mp - mp - p +.
After this extensive second section, the third section is surprisingly brief.
No sequences delay the return to the initial C-minor key; instead, the harmonic progression moves directly into the steps of the simple cadence in
the home key (mm. 15-18). The dynamic development of the four
measures should therefore appear more like an appendix to the preceding
section: mp - mp - p + - p.
The fourth section requires a more detailed description. In the WTC
version, it begins in the middle of m. 18, thus making this measure the
only one of those built on the same surface pattern that contains a change
of harmony.3 The bass-note transition from C to B converts the tonic
chord into an inverted V7/iv (dominant-seventh of the subdominant) and
thus triggers a new, active harmonic motion. This leads very soon (m. 21)
to a dominant pedal.
The gradual tension increase inherent in any prolonged pedal note is
enhanced here not only by the rising pitches of the chords (see mm. 21-24
and 25-28) but also by the sudden change, in the latter portion, to a more
3
In this matter it seems difficult to decide which of the two versions reflects the composers
true intention. The text of the initial Little Prelude shows m. 18 with two equal halves and
a repeated bass note, while the version preferred in todays Urtext editions of the WellTempered Clavier has a bass note B on the third beat of this measure. The latter version
certainly sounds harmonically more convincing, but it does so at the expense of continuity:
this measure diverges significantly from the pattern observed throughout the remainder of
this part of the piece.
C minor
65
virtuoso, toccata-style texture. This new surface pattern prepares the

intriguing inner expansion with which Bach enriched his original version
of this prelude when including it in the Well-Tempered Clavier.
The ensuing passages constitute not only a change in tempo but also a
surprise in the surface patterns. While the harmonic basis remains for a
while on the dominant pedal, the one-track, broken-chord texture gives
way to a virtuoso two-bar figure in Presto. This figure is imitated in stretto
by the lower voice, after one measure and thus overlapping for exactly one
measure. Both the model and its imitation are then sequenced, thus creating
a polyphonic structure that appears suddenly quite dense. The leading upper
voice adds a second, related but shorter figure (see mm. 32-331) that is also
imitated by the lower voice in the following measure. Within these six
measures with their two-part imitation, the harmonic change from the
dominant to the tonic pedal is carried out in such a subtle way (somewhere
between mm. 30 and 31, but with an absence of the actual bass note) that
the urge to come to a close seems as if suspended.
When the tonic pedal finally appears (m. 341), it marks the beginning
of yet another section, headed Adagio and featuring a recitative-style upper
voice over scarce, arpeggiated chords. These surface featuresthe changes
in tempo, texture, and characterare so intriguing that the tonic pedal
once again stands little chance of claiming attention for the impending end
of the piece. The third change reestablishes the texture that had determined
the last measures before this adventurous expansion (compare mm. 35-38
with mm. 25-27). Thus the bracket is finally closed.
Regarding the dynamic design in this final and largest section of the Cminor prelude, our observations permit the following conclusions:
The climax falls in the downbeat of m. 28: here, the last and longest
ringing of the dominant pedal coincides with the beginning of the
Presto and the change to the denser polyphonic texture. This climax
has been prepared in the continuous buildup begun at the end of the
last home-key cadence in m. 18.
After the climax, the overall tension diminishes in a twofold downward sweep. A (much smaller) second peak falls on the downbeat of
m. 34: here, the first appearance of the tonic pedal coincides with the
beginning of the Adagio and the first arpeggiated chord. The overall
decline of tension is then continued to the end of the piece.
This overall development of tension can be expressed in the following
dynamic terms. (pf stands for poco f. The spacing of the measures
see mm. 28, 30, 32, 34a, 34bdepicts the actual time of the musical
process in what might be the most desirable tempi.)
66
WTC I/2
mm. 21 22
mp + mf
mm. 28
f
30
pf
32
mf
23
mf
24
25
mf + pf
26 27
pf + f
28
f
34a 34b 35 36 37 38
pf mf + mf mp + mp p
The graph sums up the tension development throughout the entire prelude:
WTC I/2 in C minor Fugue

Launched on the second eighth-note of a 4/4 measure, the subject of
this fugue is two measures long. The initial notes are heard as an upbeat,
not so much to the downbeat of the following measure but to the middle of
m. 1. The E that falls on the first beat of m. 3 marks the end of the subject.
This fact, evident from a comparison with all later subject statements, is
supported by the harmonic background: the dominant in the form of a
ninth chord (G B D F A) is reached in the middle of m. 3 and resolves
onto the tonic on the following downbeat.
One of the remarkable melodic characteristics of this subject is that it
contains three identical note groups: the initial three-note figure C-B-C,
metrically placed as an upbeat, is repeated twice in metrically equivalent
position. These identical groups must be regarded as structurally corresponding; i.e., each of them initiates a separate little subphrase.
C minor
67
The subjects rhythm shows a fairly regular combination of 16th-notes

and eighth-notes, complemented by a single quarter-note shortly before the
end, featured as a syncopation. The pitch pattern comprises a mixture of
steps and leaps. Upon closer inspection, a much clearer picture emerges,
since almost all stepwise motions in 16th-note rhythm can be unmasked as
written-out ornaments: the figure C-B-C is a spelled-out inverted mordent
and the ascent F-G-A , a slide. Furthermore, none of the leaps is a hightension interval. The subjects main body thus represents neither a primarily
linear nor a highly emotional pitch progression. Were one to reduce the
subject to its unornamented line, this is what would remain: C-G-A,
C-D-G, C-D-AGFE or C-G-A, C-D-G, C-D-FGFE.
The harmonic background is that of a simple cadence. Bachs harmonizations in the course of this fugue feature only minor variations; see, e.g.,
the entry mm. 26-28 where iv is replaced by VI.
iv
i6
V9
The sequential structure is a determining factor in the subjects dynamic

outline. In the first subphrase (the head motif), the four initial notes remain
in the range of the tonic while the fifth note moves on to the subdominant.
This harmonic progression indicates an increase in tension within the first
subphrase. In a sequential pattern, the shaping in the subsequent subphrases
must follow the model, thus resulting in three similar dynamic increases.
When attempting to organize the three increases into a more encompassing
pattern, we find that there is more than one option for the overall
development of tension.
Performers choosing to stress the melodic aspect will emphasize the
peak-note descent to the keynote (A-G-F-E) with a consistent
diminuendo. Within this frame, the ascending three-note group in
the middle of m. 2 will sound as an extension of the note Fwhich
makes sense since there is no change of harmony at the emergence
of the syncopation. The tension decrease throughout the subject is
supported by the overall harmonic progression, which also climaxes
on the first peak notes subdominant.
Performers who prefer to focus on the rhythmic structure will draw
attention to the syncopation, interpreting each subphrase as more
intense than the preceding one, with the climaxes followed by a
release of tension only in the last three notes.
68
WTC I/2
There are altogether eight subject statements in this fugue:
1. mm. 1-3
M
5. mm. 15-17
M
2. mm. 3-5
U
6. mm. 20-22
U
3. mm. 7-9
L
7. mm. 26-28
L
4. mm. 11-13 U
8. mm. 29-31
U
Apart from the interval adjustment in the tonal answer, the subject
appears unchanged throughout the fugue. Both the interval variation in the
major-mode statements and the Picardy third at the end of the last entry are
regular features. The fugue does not contain any strettos or parallel entries.
It features two counter-subjects, only the first of which is truly independent.
CS1 makes its first appearance immediately after the initial subject
statement; i.e., it stretches from the second 16th-note of m. 3 to
the downbeat of m. 5. The conspicuous change in the rhythmic
pattern (only 16th-notes in m. 3, only eighth-notes thereafter) and
the pitch level (see the tenth interval C1-E 2 in m. 3) reveal this
counter-subject as structurally conceived with two contrasting
subphrases. The first segment presents a descending scale that
describes a decrease of tension. The longer second segment also
moves essentially in falling direction, with its most likely interpretation being an even more distinct diminuendo. (Climaxes on
either the lower C, the target of the descending scale, or the
admittedly interesting F are inadvisable. As both notes coincide
with tension peaks in the subject, stressing them would counteract polyphonic clarity. The rhythmic evenness in the countersubjects second half already weakens its impact in the context
of simultaneous lines; an additional dynamic parallel would
offset the requirement of independence.)
CS2 is introduced against the next subject entry. It starts belatedly but
concludes together with the other components on m. 91. Rhythmically, CS2 is almost entirely attuned to CS1. Nor is its pitch
pattern, which undergoes several changes in the course of the
fugue, very significant. Thus the subjects second companion
stages little contrast of its own. The overall impression in the
fugue is of a dialogue between the subject and one countersubject with a homophonic accompaniment.
The following sketch shows the phrase structure and dynamic design
created in the combination of the subject and its two counter-subjects.
C minor
69
There are six subject-free passages in this fugue:

E1 mm. 5-7
E4 mm. 17-20
E2 mm. 9-12
E5 mm. 22-26
E3 mm. 13-15
E6 mm. 28-29
The material of E1, E2, E4, and E5 is closely related to the subject; at
least one voice constantly displays a variation of its head motif. In E2 and
E5, this subject-derived motif is extended: what was originally the final
eighth-note now sounds as a quarter-note, followed by an eighth-note rest
and two chromatically ascending eighth-notes. This longer version of the
head motif complements the dynamic design: the slight tension rise in the
subphrase is now rounded off with a relaxation throughout the extension.4
The first episode motif (M1) is introduced in E1. It consists of an
ascending scale that comes to a halt on a syncopated A. The two features,
the scalar motion and the syncopation, reveal M1 as related both to the first
counter-subject (which begins with a scale, though a descending one) and
to the subject (which contains the syncopation group F-G-A). Interpretation of the character and dynamic design in this motif depends on which of
these relationship one regards as more salient; M1 could be taken as a
relaxed scale in diminuendo (like the beginning of CS1) or as an upbeat to
a little climax (on F or on A, whichever your choice was in the subject).
M2 is brought forth in the lower voice of E2. It is a figure of one-bar
length made up of a descending one-octave scale that bends back for two
notes and is then followed by another descent down to the lower fifth. This
motif derives from the beginning of the first counter-subject and picks up
its character and tension decrease. As E3 shows, this motif can appear not
4
Note that in the stretto imitation used in these episodes, the [decreasing] chromatic ascent
of this extension coincides with the [increasing] regular eighth-notes that open the head motif.
This takes many performers unawares: they lose sight of which voice is leading and when.
70
WTC I/2
only in descending but also in ascending direction. The most consistent

(though not often heard) interpretive conclusion would be to retain both
the character and the diminuendo. (The implications of this interpretation
of E3 for the overall dynamic development are considerable and will be
dealt with later.)
The double-note motif that appears in the two lower voices of E3 but
is never again taken up in the fugue is least related to the primary material.
It merely bears a slight resemblance to the first three notes of CS2 that,
together with the same note-group in CS1, also forms descending parallels.
By far the shortest episode of this fugue is E6. Its harmonic progression
and melodic features leave no doubt that it is a typical cadential close. The
lower voice presents a cadential-bass pattern, the middle voice the characteristic closing formula with syncopation, and even the upper voice joins in by
displaying one of the established closing features, the keynote anticipation.
The relationships between the episodes of this fugue are as follows:
The first half of E4 (mm. 17-18) is a variation of E1: compare L in
E4 with U in E1, M in E4 with M in E1. U in E4 is new but in part
parallel to L.
The second half of E4 is a varied repetition of the first half of E4:
compare mm. 17-183 with mm. 183-201 (M + L in inverted voices).
The first half of E5 (mm. 22-24) is a variation of E2: compare U/M
in E5 with U/M in E2 (transposed) and L in E5 with L in E2
(metrically varied transposition).
Both E1 and the two halves of E4 are composed in ascending sequences.
They prepare the following entry by building up tension toward it. In the
case of E1, the tension is slightly abated at the very end of the episode.
By contrast, E2 and E5, both determined by falling sequences, create a
diminuendo. This seems only logical in the case of E2, which follows the
last of the three initial statements in this three-part fugue. With its decreasing tension it thus sets the so far uninterrupted tension rise apart from what
follows. The symmetrically beginning E5 seems intended also to give the
impression that the main statements in this round have already been made,
and lures listeners into expecting a redundant subject entry. But Bach
changes his mind; he extends this episode by adding an ascending, tensionincreasing second half (mm. 25-26) that leads to a statement in the lower
voice, thus completing the round. A similar process recurs in the final
episode: the cadential measures 28-29 seem to close this section in a mood
of relaxation, but Bach adds a further entryredundant this timein
homophonic texture above a tonic pedal.
The interpretation of E3 requires careful pondering:
C minor
71
If conceived as a crescendo, E3 appears as a link between the

preceding and the ensuing subject statements. As a result it strings
togetherby means of its dynamic direction entries that do not
belong to the same group.
If conceived as a diminuendo, E3 suggests an interpretation of the
fugue as composed in two halves. This is corroborated by the fact
that E3 is the only episode not at all related to the subject and
occurs close to the middle of the composition.
The leaps in the pitch pattern and the prevalence of two rhythmic
valueseighth-notes and 16th-notessuggest a lively basic character for
this composition. The ideal tempo is dictated by two essential details: the
dance-like character, created above all by the unusually regular phrase
structure in the subject, and the ornamental character of the initial figure
C-B-C. The most appropriate pace would therefore be one in which the
eighth-notes sound gracefully bouncing and the 16th-notes are swift
enough to be perceived as single notes.
Articulation in the fugue is a light quasi legato for all 16th-notes and a
not too heavy non legato for all eighth-notes and longer notes, with the
exception of syncopations. The Urtext of the fugue in C minor does not
contain any ornaments.
Design indicators are particularly eloquent in this fugue. The entering
order of the voices shows an exact repetition: middle voice, upper voice,
lower voice, and upper voice in both mm. 1-13 and 15-31. Moreover, there
is a striking symmetry between the linking episodes in both halves of the
fugue: E4 is a variation of E1, E5/1st half is a transposition of E2. This
twofold symmetry overlaps artfully insofar as the corresponding episodes
do not link corresponding voice entries. The result is a design that, on one
of the two levels, relates the unaccompanied first subject statement to the
homophonically accompanied last onea truly ingenious plan.
subject in M
subject in U
subject in M
E1
E4
subject in L
subject in U
E2
E5
subject in U
subject in L
E3
E6
subject in U
The harmonic progression in this fugue leads from C minor to the
relative major key, E major (in the fourth subject entry), but returns to the
home key very soon thereafter and never leaves it again.
72
WTC I/2
The last two-and-a-half measures appear as a coda: the final cadence

has already taken place, the bass is limited to an extended tonic pedal note,
and the middle voice does for once not take up the counter-subject or
sound any other polyphonically independent line but accompanies the final
statement in strict homophonic style. Moreover, these measures sound in
voice splitting, i.e., they abandon the original three-part setting and split
into an octave (L pedal) and into chords or double notes (M).
Here is a sketch showing the design of the C-minor fugue:
The dynamic tension curve in the first section shows a gradual increase
through the first three entries, caused by the buildup of the ensemble and
the increasing gesture of the first episode. There follows a slight decline
during the second episode and in the fourth entry, which is not only redundant but also sounds in the major mode, thus appearing more relaxed than
the original minor-mode version. As this fugue is not dramatic but rather
cheerful and dance-like, the rising and falling dynamic developments in
this first section are not particularly strong. The second section repeats this
dynamic pattern: the first three subject statements, including the episodes
linking them, increase very smoothly. (The second of these episodes, E5,
begins by decreasing tension before it realizes that this is premature and
that there is one more entry to come in the buildup.) The cadential close
after the second sections third entry brings the tension release, so that the
redundant entry, coinciding with the coda, sounds like a softer afterthought.
The relationship between the two sections is one of enhanced repetition.
The second section sets out in a more involved way with its first subject
statement accompanied by both counter-subjects. The episodes are longer,
thus creating a higher degree of anticipation for the next subject entry. The
final entry, in its detachment from the fugues polyphonic texture, may
depict more of a retreat than the last entry of the first section, which
appears lighter mainly because of its changed mode.
WTC I/3 in C major Prelude

This prelude is designed in two-part texture. It features two motifs.
Both of them are immediately imitated before they return to their original
register. The principal motif is conceived not as a one-dimensional melodic
line but in hidden two-part structure. Moreover, the secondary voice in this
motif is not polyphonically independent but a disguised parallel. Yet in its
development, the two lines create a brief play of genuine contrapuntal
juxtaposition. The second motif is designed in barely veiled homophony.
The first cadence ends at m. 71. This cadential close lies embedded in
a melodic flow that continues uninterrupted. A change of surface pattern
occurs in m. 8, where two melodic voices lead into the inverted-voice
texture of mm. 9-15. In this sense, this initial cadence is only an indirect
indication of a subordinate structural ending within a larger context. There
is, therefore, no caesura, and no cut after the reappearance of the tonic.
Exactly the same holds true for the following harmonic progression which,
now in the tonal realm of G major, draws to a cadential close at m. 151.
Here again, the melodic pattern continues through another measure before
giving way to a continuation in inverted voices that marks the beginning of
a new harmonic development.
Owing to the 3/8 time, the structural units determined by these cadences
appear short and, with their eight-measure extension, supremely regular.
The phrase in mm. 25-31 is the first not to be followed by such a voice73
74
WTC I/3
swapping link. Instead, the second half of the cadential close serves as the
beginning of a new development. The following diagram shows the
phrases and their tonal areas in the entire prelude. The bridging measures
that link consecutive phrases by prolonging the tonic of a cadential close
before a renewed change of voices do not harmonically belong to either of
the closed progressions; they are thus deliberately omitted here. The
graphic arrangement tries to visualize the harmonic progressions.
1. mm. 1-7 C major
2. mm. 9-15 G major
3. mm. 17-23 D minor
4. mm. 25-31 A minor
5. mm. 31-35 A minor/D minor
6. mm. 35-39 D minor/G major
7. mm. 39-43 G major/C major
8. mm. 43-47 C major/F major
9. mm. 47-53 F major
10. mm. 55-61 C major
11. mm. 63-73 G pedal
12. mm. 75-83 C major
13. mm. 87-104 G pedal, resolving to C only in the final measure
The prelude features both identical and structurally analogous phrases.

There is a stretch of seven measures in the first half of the piece that recurs
untransposed (mm. 1-7 . 55-61); another span of ten measures in the
second half of the prelude is repeated a few measures later an octave lower
(mm. 63-72 . 87-96). The preludes opening phrase is taken up altogether
five times, both transposed and in inverted voices. While the first three
reappearances follow the model phrase and are therefore best named
imitation in inverted voices, a true recapitulation can be found after the
interspersed portion with its different patterns (see mm. 47- 62). These two
phrases in the middle of the piece sound reminiscent of the preludes very
beginning, particularly since they are conceived as mirror images of the
first two phrases.
Moreover, one can detect a large-scale structural analogy: the first half
of the prelude, visually distinct from the second by its uninterrupted flow
of 16th-notes in at least one of the voices, is designed in ternary form. So
is the second half (with the exception of the final line). Here is a scheme of
this preludes structure, drawing only on the architectonic patterns but not
yet on the motivic components:
C major
75
mm. 1-31 two-part pattern, voices interdependent

mm. 31-46 contrapuntal interplay
mm. 47-62 two-part pattern, voices interdependent
mm. 63-74 homophonic pattern
mm. 75-86 contrapuntal interplay
mm. 87-96 homophonic pattern
On a different scale, there is an analogy in the harmonic structure: the
first four phraseseight measures each, in dependent two-part pattern
move from the tonic through the dominant and the supertonic to the
relative minor. The following four phrasesfour measures each, in contrapuntal texturereverse this process and end on the tonic.
The time signature determines the tempo together the indirect melodic
pattern in the non-contrapuntal segments. As we know from other music
by Bach, 3/8 does not so much indicate a pulse in eighth-notes but rather
one in whole-measure beats. With these larger pulses in mind, it should not
be difficult to find the appropriate range for the tempo.
The articulation in a hidden two-part structure is a little more complex
than that in a one-track melodic texture. Here it is the hidden melodic
line that is to be articulated, not the progression from note to note as it
appears on the surface. In the secondary line of the main motifs model, i.e.,
the first seven measures, the left-hand lineC-D-E-F-E-D-C
should sound non legato in the context of this lively character. This non
legato effect is, however, naturally achieved by the interspersed repeated
pedal note. It would be counterproductive to separate the first C from its
octave and so forth, since by doing so one would obtain the opposite effect
and hear all notes as belonging to one line. In other words: In order to
achieve a melodically correct result in the hidden two-part structure,
some surface progressions must be played legato. The distinction in each
hand between melody and background is achieved above all by two
means: touch and intensity. The ideal shading of the texture consists of an
intense, dynamically molded melodic line sounding against an almost
neutral, dynamically more subdued background.
The performance of the motif derivatives in contrapuntal play must be
considered carefully. M1a and M1b develop from the principal motifs
link. After the fourth appearance of M1 in mm. 25-31, this link forms two
new thematic units: In the right hand the syncopation, followed by a
simplified version of what was originally the bridge to the next phrase, is
repeated in sequence and thus builds a four-bar motif (U: mm. 313-351). In
the left hand, another feature pertaining to the link, the turn figure from
U: m. 8, also brings about a little four-measure motif (L: mm. 31-351). Just
76
WTC I/3
as M1 sounded four times in alternating voice inversions, so do these two

derivations. However, besides this similarity there are considerable differences: While in M1 each line is conceived in hidden two-part structure,
both M1a and M1b are designed as one-track developments. While M1
consists of two interdependent lines whose melodic components form
parallels, M1a and M1b are polyphonically independent. While in M1
there is only one rise and fall in tension tracing the parallel curve, M1a and
M1b show independent dynamic outlines: M1b creates a two-measure
crescendo followed by a two-measure diminuendo; the two halves of M1a
each feature and accented syncopation with subsequent relaxation.
M2 is introduced in mm. 63-73. It bears relationships with M1. Like
M1 it is designed in homophonic texture. Like M1 it contains a pedal note
in both hands (G). Like M1 it determines the main part of a ternary form.
Finally, like M1 it is followed by a link that, this time, does not connect
the model to its imitation but joins this motif to the next (see mm. 73-75).
On the surface, the motif may remind listeners of toccata style: the two
parts only meet on the downbeats, after which they move in complementary
rhythm. It is interesting to see that later, after its repetition an octave lower,
M2 develops into a different pattern, also in toccata style and also resting
on the pedal note G (see mm. 97-102).
M2a determines the middle section of this second ternary form in the
prelude. It features polyphonic texture, thus relating to the corresponding
section in the first half of the composition. Here, in mm. 75-83, the right
hand retains its complementary rhythm but ascends in a large sweep up to
B (almost the highest available note on Bachs keyboard). From there
C major
77
it moves in 16th-notes, gradually releasing the tension it previously built

up. The entire upper-voice motif is then repeated a tone lower (and slightly
softer). At the same time, the left hand describes a curve that sounds like a
simplified version of M2a. It begins with the descent, i.e., with the middle
of the motif, which falls over two measures from E to the lower-octave C.
Thereafter it follows the right hand in stretto imitation, rising over another
two measures before sequencing the figure.
The development of tension in this prelude relies mainly on two facts:
the harmonic progression (particularly in the first half) and the difference
of intensity between the homophonic and the polyphonic sections. The
diagram inserted above shows these progressions.
WTC I/3 in C major Fugue

This subject spans a little less than two measures. It begins after a 3/8note rest with what listeners experience as a secondary upbeat. Such an
upbeat, relating to the middle beat in quadruple time, is a fairly strong
impulse-giving feature, not least because we anticipate that it will shift to
a normal upbeat position in the course of the fugue.1 The ending of the first
subject statement falls on the downbeat of m. 3. The dominant harmony is
represented by the last two eighth-notes in m. 2, after which the expected
resolution onto the tonic is reached with the following keynote.
At first glance, the subjects melodic structure seems to comprise two
segments: sequencing leaps unite the last six notes, thus appearing to distinguish the second half of the phrase from the first. This first half would
then end after the written-out inverted mordent on the second beat of m. 2
(i.e. between F and D). However, the straightforward harmonic motion,
which describes a single progression throughout the whole subject, speaks
in favor of an interpretation as an indivisible phrase. This is supported by
the indirect descent throughout the whole phrase: GG-F-E-D-C.
6
4
Baroque polyphony, in contrast both to the contemporary dance types and also to the
music of the ensuing period, knew frequent metric shifts of its thematic material. This is
especially true in quadruple time where a subject or motif first introduced in the middle of
the measure could be placed at the beginning of a measure, and vice versa, in later
statements within the same piece.
78
WTC I/3
There are only two rhythmic values: eighth-notes and 16th-notes. The
pitch pattern is characterized by leaps rather than steps; note the melodic
broken chord at the beginning (E-C-G) and the alternating sixth and
seventh intervals at the end. There are few steps; all can be identified as
written-out ornaments: a turn in m. 1 and an inverted mordent in m. 2. In
the subjects harmonic progression, the active step to the subdominant falls
on the downbeat of m. 2. An analysis of the underlying chord progressions
that Bach uses later in the fugue reveals the G in m. 2 as an appoggiatura
to the following F. These two notes, G and F, thus form a pair that may
under no circumstances be separated by either phrasing or articulation.
The climax in the subject occurs unmistakably on the downbeat of the
second measure. Here, two powerful tension-enhancing features coincide:
the appoggiatura and the active harmonic movement from the tonic to the
subdominant (or, more often in this piece, its relative minor on ii). The
peak note E (which may tempt all those who connect strong feelings with
high pitches) is in reality only part of a broken-chord pattern on the tonic
and therefore melodically and harmonically insignificant. The dynamic
curve in the subject thus begins with an energetic crescendo through the
first segment up to the downbeat G. This crescendo should develop evenly
and not burst out too early, so as to give the E and C enough impetus
toward the appoggiatura G. In the fairly abrupt tension decay that follows
from this appoggiatura to its resolution, approximately half of the tension
is lost. The remainder is then released gradually throughout the series of
leaps.
There are twelve subject statements in this fugue:
1. mm. 1-3 U
5. mm. 14-16 L
9. mm. 42-44 U
2. mm. 3-5 M
6. mm. 19-21 M
10. mm. 44-46 M
3. mm. 5-7 L
7. mm. 24-26 U
11. mm. 46-48 L
4. mm. 10-12 U
8. mm. 26-28 M
12. mm. 51-53 U
In three of the subject entries, nos. 4, 7, and 12, the upbeat eighth-note
is replaced by three 16th-notes, while in statement 10 it appears split into
two 16th-notes. Metric displacementsa beginning on the last eighth-note
of a baroccur in statements 7, 8, and 12. The subject does not appear in
either stretto or parallel but takes three counter-subjects. CS1 is introduced
against the subjects second entry (see mm. 3-5: U) and remains a faithful
companion ever thereafter. It is exactly two measures long, a little longer
than the subject, as it begins slightly earlier in the bar. In its full scope it
C major
79
begins with an inverted-mordent figure on the keynote, followed by a

five-note scalar ascent and a turn figure on the peak. From here to its end
there is a gradual descent in ornamental waves, interrupted only rhythmically by one prolonged note.
CS1 is thus conceived as a unit without subdivisions. If we consider
the dynamic development without taking into account the simultaneous
events in the subject, there are two possible interpretations, one based on
pitch, the other based on rhythmic features. The pitch pattern suggests a
climax on the highest note at the beginning of the turn figure, while the
rhythmic pattern seems to favor the longest (tied) note. As soon as we stop
regarding the counter-subject as an independent body and look at it as a
counterpart to the subject, the choice between these two options becomes
much easier. As the second solution would cause the climaxes of the
subject and its first companion to coincide, this option is contrary to the
polyphonic requirement of greatest possible independence of the voices.
Therefore, the first dynamic design is the more appropriate choice.2 In the
course of the fugue, the first counter-subject undergoes one significant
variation. The turn figure is sometimes written using the leading-note to the
fifth degree; in other instances it retains the context of the natural scale.
This results in inconsistencies in otherwise completely analogous portions
(see, e.g., U in mm. 3 and 44, M in mm. 10 and 52).
CS2 is heard six times in the course of the fugue. It is introduced at its
anticipated place, i.e., against the subjects third entry, and taken up again
in mm. 19-20, 25-26 (with a shortened beginning), 26-28, 44-46 (with a
varied beginning), and 46-48. Its rhythmic and harmonic features create a
distinct contrast to the two other components: Rhythmically, an initial upbeat precedes syncopations the first two of which are eight times (!) as
long as the so far prevalent 16th-notes. Harmonically, its beginning places
the first notes of the subject in a V7 context, while its ending omits the
resolution into the tonic (at least in the original statement; this is later
corrected). The tension layout is quite unequivocal: the first prolonged
syncopation builds the high-tension interval of a minor seventh over the
initial notes of both subject and CS1. It thus represents a natural climax,
with the ensuing descent providing the relaxation.
CS3, an unexpected further companion in a three-part fugue, appears
only twice. In mm. 10-12 it sounds against the fourth subject statement; in
mm. 51-53 against the last. Its characteristic features are the eighth-note
2
While the mind may easily accept this truth, the fingers seem more reluctant and often find
it difficult to resist stressing the longer note.
80
WTC I/3
upbeat followed by a descent in longer note values. These features and the
long release of tension resulting from this melodic shape reveal its
relationship with CS2. This impression, however, is weakened both by the
harmonic progression and by the concluding cadential-bass steps.
The sketch shows the phrase structure and dynamic design in the
primary thematic material of this fugue:
The subject statements are interspersed six times with subject-free

passages; a seventh episode closes the fugue.
E1 mm. 7-102
E3 mm. 16-191
E6 mm. 48-514
E2 mm. 12-142
E4 mm. 21-244
E7 mm. 533-55
E5 mm. 283-422
Two of these episodes are related to the subject: In E4, the upper voice
recalls the first half of the subject twice (mm. 22-24); in E5, the upper
voice quotes the subjects first segment three times (mm. 34-37), after
which the lower voice imitates the threefold quotation of the subjects
initial segment (mm. 38-41).
E1 introduces an independent motif that plays a major role in the fugue
(see mm. 7/8 U: from G-E to tied note A). M1 is imitated in stretto, with
a slight interval adjustment, in the middle voice of the same bar. Both the
model and its imitation are then sequenced. In a second sequence, both
voices show a variation in the second half of the motif; moreover, the upper
voice is completed by means of an extension that provides the harmonic
resolution to its tied-note appoggiatura. In the original version, the dynamic
curve expressed in M1 is obvious: the jump upward creates a rise in tension
followed by a release in the falling broken chord. In the extended version,
the upward motion is enlarged. It would therefore seem logical that the rise
C major
81
in tension be also increased. Also in E1, the lower voice presents a motif
that can be traced back to the first counter-subject. It sets out with the same
inverted-mordent figure, followed by ornamental waves recalling the final
groups of CS1 in inversion. This motif is also frequently used within the
fugue and will be called M2. In terms of tension, M2 contains very little
active power. The short upbeat-like impulse in the inverted-mordent figure
is followed by a long, subdued drop in tension.
E2 is related to E1, but both M1 and M2 recur in considerable transformation. There are some significant changes that create an entirely different
character: In M1, the leader of the imitative pattern is now the middle
voice and the tie prolongations in the motif are replaced by rests, creating
a definite interruption of the tension. Moreover, the imitating voice does not
follow its leader but restates the second, relaxing half of the motif. Toward
the end of the episode, both voices abandon the motivic context altogether
and join in a cadential figure. M2 sounds in the upper voice where it no
longer appears as a sequencing one-bar figure, but is extended to a two-bar
curve. In free inversion, it is pushed upward to a slide-decorated peak; the
effect of this climax is enhanced by the fact that it sounds in a diminishedseventh interval to the C in the lower voice. The ensuing release ends in
a so-called female extension, a melodic tail after the harmonically resolved strong beat, on the fifth 16th-note of m. 14.
None of the subject-free passages serves exclusively as a cadential close.
But the final 2 measures (E7) and the first 1 measures of E4 both
present non-motivic material that leads to perfect cadences with distinct
closing formulas (see the cadential-bass patterns in L: mm. 21-22 and 55 as
well as the two typical melodic formulas in U: mm. 22 and 55). In the case
of E4, the cadential close divides the episode into two segments (E4a: mm.
21-223, E4b: mm. 223 -244). Another episode, E5, is subdivided even
further (E5a: mm. 283-303, E5b: mm. 303-344, E5c: mm. 344-422).
Finally, several episodes or their segments are varied repetitions of
earlier models. E1 recurs in three varied repetitions: E3 uses the tensionextended version of M1 from m. 9, E6 is closest to the model but begins
with a half-bar extension, and E5b appears most remote in its use of the
motivic material: the voices are exchanged, with the lower voice in the
lead, the middle voice reduced to a broken-chord figure, and the upper
voice recalling M2.3 E2 recurs once: E5a recalls it in inverted voices, with
3
Despite the varied beginning, the statements of M1 in the lower voice should nevertheless
retain the tension curve characteristic for this motif (with, e.g., a crescendo in mm. 304-311
followed by a diminuendo up to m. 31m), while the figure in the middle voice is too
removed from the original to take part in any subtle dynamic shaping.
82
WTC I/3
the lower voice now featuring the dramatic ascent (a slide should be added
on the peak note in m. 29) while the upper voice, partly crossing over the
middle voice, recalls the developed version of M1 and the middle voice
just fills the texture. E5c, the last segment of the longest episode, features
an internal correspondence: U: mm. 35-38 (with 3/16 upbeat) are taken up,
voices inverted, in L: mm. 39-42 (with 3/16 upbeat). One may get a clearer
picture of what is happening in these episodes by completing the earlier
diagram as follows:
E1
E2
E3 (. E1)
E4a
E4b
E5 (. E2 + E1 + E4b)
E6 (. E1)
E7
The role these episodes and their segments play in the dynamic development that shapes the fugue as a whole is both relevant in each case and
significant with regard to the understanding of the overall structure. E1, as
it is determined by the introduction of new material, demands a change of
register or color (the same holds true for its three variations). Its sequences
progress downward, thus causing a relaxation and suggesting that a section
is drawing to its close. (In fact, all three voices have already stated the
subject.) The last sequence, however, extends the rise within M1 and reestablishes a higher level of tension, thus preparing the listener for more to
come: a redundant entry. E3, the first variation of E1, shows even more of
this extended rising. The ascending trend within the melodic units counterbalances the relaxation implied in the descending sequences and thus
defines this episode as one linking adjacent subject entries. By contrast, E5b,
the second variation of E1, stresses the decline. Its role in the overall
tension is one of announcing the forthcoming end of a section. E6, the third
variation of E1, returns to the pattern of the original: the smooth tension
decay in the descending sequential pattern is arrested at the last moment,
thus granting the following (redundant) subject entry to be perceived as
still being part of the section.
Both the original E2 and its variation in E5a are self-contained units. In
a color distinctly different from that of the subject-determined passages,
their dynamic outline runs in curves; within each of them, a rise to the
climax and a subsequent relaxation are concluded by a cadential close.
Compared to the relaxing E1 and the self-contained E2, E4b represents the
type of episode that conveys a preparation for a subsequent entry. The
incomplete subject statements serve to suspend the tension before the
ensuing full entry. This impression is further enhanced by three facts: this
episode segment sets off after a cadential close, it is presented in reduced
ensemble, and the secondary voice is confined to non-motivic material.
C major
83
E5c seems to repeat this pattern on a heightened level: It also begins

after a complete decline in tension, it is also presented in reduced ensemble
(this time it is the middle voice that is resting), and its secondary voice is
not only non-motivic but actually displays a barely disguised prelude-style
accompaniment pattern. This last fact especially sets this episode portion
furthest apart from the remainder of the fugue. Most of the active strength
and tension otherwise characterizing this work seem temporarily withheld.
Finally, the two cadential formulas in the episodes together with their
extended preparations form dynamic curves. In E4a, the climax falls on the
downbeat of m. 22. E7, the closing episode of the fugue, is launched from
the interrupted cadence at the end of the final subject statement and
describes an increase toward the dominant bass note G. Whether the very
ending, with its quotation from the subjects tail and its voice splitting, is
interpreted as a relaxation or as a triumphant close, remains at the individual performers discretion.
The basic character of this fugue is rather lively. Both the pitch pattern
with its many written-out ornaments, leaps, and broken chords and the
rhythmic pattern with its predominance of two note values support this
impression. The only component introducing a hint of contrast is the second
counter-subject with its long notes, chain of syncopations, and stepwise
motion. The overall tempo should be fast enough to allow the written-out
embellishments in the primary material to retain a touch of their ornamental characteri.e., the four notes of the turn at the subjects beginning
should be heard as an entity rather than as separate notes. The appropriate
articulation in this rather lively composition consists of non-legato eighthnotes and legato 16th-notes. It is possible and gives the fugue a lovely depth
(while admittedly increasing acrobatic demands) to play the contrasting
CS2 with the characteristics of a rather calm character, i.e., with legato
articulation and a singing touch. But it is obviously also possible to opt for
unity of character in all components of the material and play this countersubject with long but slightly detached notes.
A good tempo balance between the prelude and its fugue is reached by
transforming a triplet (i.e., three eighth-notes in the pulse of the prelude)
into a duplet (i.e., into two quarter-notes in the pace of the fugue). The
proportion thus reads:
1 measure
corresponds with
measure
in the prelude
in the fugue.
(Approximate metronome settings: 72-80 for one prelude bar, 108-120 for
one beat in the fugue.)
84
WTC I/3
The fugue features three ornaments: the slide in E2 (and, correspondingly, in E5a), the cadential mordent in U: m. 22 (not included in the fair
copy but deriving from a copy, presumably because Bach regarded this
conventional ornament as self-evident), and the compound ornament toward
the end of E5 (see U: m. 38). The slide often causes confusion, probably
because it is conventionally printed slightly to the left of the note head it
ornaments. Yet just like other Baroque embellishments, it also begins on
the beat. Thus in m. 13, the right-hand downbeat G falls on the left-hand
C. It is followed, in 32nd-notes or faster, by A and a B that is sustained for
the remainder of the note value. Correspondingly, in m. 29 the left hand
plays C-D-E with the C (not the E) coinciding with the middle-voice F.
The cadential ornament in m. 22 is a simple mordent, beginning on the
upper neighbor note and consisting of a double shake: E-D-E-D. The
symbol for the complex ornament in m. 38 asks for a turn progressing into
a trill. Because of its tied ending and delayed resolution, this trill should
conclude without a suffix. The result is, in 16th-note motion, an initial
A-G-F-G.4 This is followed by six A-G groups. (In order to further
enhance the suspension it is possible to play only five A-G groups and
stop a little earlier before the bar line.)
The most prominent feature of this fugue, in terms of design, is the
striking analogy of mm. 1-12 and mm. 42-53. Here are the details: The
order, position, and keys of the three initial entries recur identically. The
surrounding counter-subjects also correspond (although mm. 42-46 now
feature an additional voice, owing to the fact that the ensemble in a fugue
never drops back to a single voice). As has been shown above, the
subsequent episode, E6 in mm. 48-51, is a variation of E1, with its initial
half-measure extension serving to modulate. The subject entry that follows
also corresponds with the redundant entry in the first section although, as
a result of the modulation in the episode, mm. 51-53 now sound on the
tonic and not on the dominant as did mm. 10-12. However, the order and
position of the voices are the same. The analogy of these two entries is
additionally enhanced by the fact that they are the only ones in the entire
fugue to be accompanied by CS3. This prominent analogy defines the
major structural traits of the fugue. In addition, the episodes play an
important role in determining the design.
4
The pitch of the lower auxiliary needs a comment. The harmony underlying these
measures of retransition is normally interpreted as the dominant, represented by an
alternation of dominant-six-four (C major with G in the bass) and dominant-seventh
chords. If this interpretation is adopted, i.e., if G major acts as a dominant, its seventh is F
(heard repeatedly in these bars), whereas its leading note in ornaments should be F.
C major
85
There are three instances in the C-major fugue where the concluding
force of an episode sheds light on Bachs intention of partitioning the work
into sections. The first is the cadential close that ends E2 at the beginning
of m. 14. The second is the explicit cadence in the middle of m. 22, already
mentioned repeatedly. The third is more complex. In E5a, the variation of
E2 seems to conclude something in the middle of m. 30. However, this
cadential close is followed by the variation of E1 that, as has been shown,
makes no attempt to launch any kind of new development. Neither does the
ensuing segment in which the tension is suspended. To see these three
episode segments follow one another is already unusual enough. To see
them trying to surpass each other in tensionlessness is even stranger.
However, it is this very strangeness, this long retreat from the active
striving in the piece, which gives this portion its particular effect: as a
protracted buffer before the last section of the fugue it succeeds in
highlighting the symmetrical design.
Finally, looking for features that might indicate section beginnings we
find that two subject entries in this fugue appear in reduced ensemble.
They are the first minor mode statement in mm. 14-16 (which is thus triply
justified in being regarded as a section beginning) and the first entry in the
recapitulating final section.
86
WTC I/3
The harmonic outline confirms the other findings: The first four subject
statements remain in the home key of C major. The following two are in
minor modethe relative minor keys of the tonic and the dominant
respectively. E minor, the relative of the dominant G major, is also the
key in which Bach concludes this section with a cadential formula. The
episode segment E4b modulates back to C major, and all remaining
subject entries are presented in the home key, in the conventional
alternation of tonic and dominant.
The analogy of the first and last sections requires correspondences also
on the level of dynamics. Both times, the first three entries sound gaily
bouncing, with a slight increase caused by the growing number of voices.
The redundant fourth subject statements regain this mood after the very
timely pickup of tension at the end of the respective episodes. The second
section also shows a slight tension increase between its two subject statements, mainly because of the growth from two to three voices.
In the third section, however, the first of the two entries most probably
contains more tension than its successor. The main reasons are that it
appears at a point of heightened expectancy (after the two incomplete
subject statements) and that, in unmodified ensemble strength, the subject
appears in the upper voice, making it appear more powerful than the
following statements middle-voice position. This decreasing tendency is
then continued through the long string of episode segments that, as has
been shown above, become ever lighter.
Among the four sections of this fugue, the analogous outer ones take
the lead. The second section sounds softened, due both to its minor mode
and its shorter extension. In the third section, the four measures containing
the two subject statements return to the home (major) key but contain no
special features that would emphasize them in any way. While they may
sound more self-assured than the preceding minor-mode entries, this mood
is overshadowed by the nearly sixteen measures of surrounding episodic
material.
WTC I/4 in C minor Prelude

The C minor prelude is based on motivic development at different
levels, from faithful imitation and sequencing to partial quotation and
rather free adaptation. Its texture is polyphonic without strict part writing.
The beginning and the end are written in four voices, occasionally splitting
to five; however, large portions in between feature only three or even two
parts, without clearly definable exits and re-entries of the voices.
The first cadence is deceptive, with a conclusion on the downbeat of
m. 5: after the D7 harmony in mm. 3-4, the melodic progression in all
voices seems to announce the return to a tonic chord, which would consist
of C/E/G/C. However, the bass does not resolve but remains tied, and
the alto rises to A instead of remaining on G. The resulting chord is thus
a polytonal mixture of the tonic and the chord on step VI.1 The nature of
this chord makes it very unlikely that this cadential close is intended as a
structural caesura. The design of the left-hand part corroborates this by
featuring descending double sixths in mm. 5-7 as an extension of the sixth
in which the two lower voices ended the cadence.
The harmonic progression then modulates to the relative key E major
and draws to a close on the downbeat of m. 8. Although the feeling of a
structural close will be much stronger a few measures later owing to the
pronounced cadential formula there, this perfect resolution into the related
key has to be regarded as structurally relevant, particularly since it coincides with the completion of a melodic idea. The more obvious caesura
after the middle beat of m. 14 marks the close of the following structural
unit with a cadence in G minor. A shorter fourth section consists exclusively of three dominant-seventh chords and their resolutions: mm. 15-20 =
F7 / B major, G7 / C minor, C7 / F minor. This leads to the longest section
in this composition: a harmonic progression with several near-cadences
thwarted by last-minute deviations. As if in symmetry to the beginning
there is another deceptive cadence in m. 35, delaying the perfect cadence,
which materializes only in the preludes final measure.
The analogous harmonic structure of the two-measure subphrases in
mm. 15-20 apart, there are several structural correspondences in this
prelude: U: mm. 5-81 . 204-234 transposed a third lower, thus transforming
1
C E G + A C E. Note that this combination of pitches is exactly the same as that in the
final interrupted cadence of the C minor fuguecompare fugue m. 112 with prelude m. 5.
87
88
WTC I/4
the cadence in E major to one in C minor. This return to the tonic passes
almost unnoticed because of Bachs decision to continue and develop the
same melodic idea throughout the next measures. Moreover, mm. 8-10
recur in mm. 23-25, transposed and with the middle and lower voices
varied. (See also the remote resemblances in the following measures.)
The character of the piece is calm, yet the time signature might present
a pitfall in terms of tempo: the calm character expresses itself by means of
a half-measure pulse. (The harmonic tempo, proceeding exclusively in
half-measure steps, supports this pace.) While these pulses should convey
a generously swinging motion, the actual tempo of the quarter-notes is not
really slow. The appropriate articulation is an overall legato not even
interrupted for cadential-bass formulas since Bach writes all of them
explicitly with tied notes (see e.g. mm. 13, 34, and 38).
There are several ornaments. The opening phrase features arpeggiated
chords, grace-note groups, single grace notes, and a mordent. The arpeggios
and the grace-note pairs represent basically the same musical feature; both
are played on the beat and fast. The notes are sustained whenever they are
essential to fill the chord (as in mm. 1, 3, and 12); they sound equally fast
but unsustained whenever they contain pitches also provided for by other
voices (as in mm. 2, 4, and 8). The single grace-notes (see mm. 2 and 4)
are appoggiaturas since they represent harmonies different from the main
notes, and must be played with due weight. Appearing in the context of
imitating voices, in which case the rules of polyphony require the retreating voice to give way to the imitating one, these appoggiaturas are best
resolved after one quarter-note each. The mordent, approached stepwise by
the appoggiatura, begins on the main note and consists of a single
three-note shake. Both the grace-note (appoggiatura) and the mordent on
the main note (resolution) should be transferred to all recurrences of the
same motif, wherever harmonic and structural progressions allow this.2
The grace-notes in the second phrase are somewhat more problematic.
Harmonically, they represent appoggiaturas and thus require time. Melodically, they interrupt the smooth scalar motion and create note repetition.
Rhythmically, they would have to be played in 16th-notes. Perhaps most
importantly in terms of the impact on the character of the prelude: as this
splitting of an eighth-note falls on the final note of a motif, an impression
of congestion arises together with a blur of the phrasing. These reasons
combined suggest that it may be advisable to do without these ornaments.
2
Left hand m. 3 add grace-note A on beat 1, mordent on G beat 2, right hand m. 4 add
mordent on D, and left hand m. 9 add grace-note C, mordent on B.
C minor
89
The third phrase introduces a number of new ornaments. In m. 11, the

appoggiatura in the middle voice resolves on the second eighth-note; the
inverted mordent uses A-G-A. As the second half of the measure is a
sequence of the first half, one might consider adding the grace-note E
before the D. Whether Bach left it out purposefully to avoid the resulting
pitch progression E-D-E, or whether it is one of the many corresponding
segments in which ornamentation is optional, is hard to decide. In m. 13,
the grace-note is also one eighth-note long, and the turn should sound
B-A-G-A in regular 16th-notes.
Within the fourth phrase, the varied sequence mentioned above appears
without any of the ornaments that decorated the model. Performers may
wish to consider adding the three inverted mordents:
on C (U: m. 166, corresponding with that on B in m. 146),
on G (L: m. 174, corresponding with that on F in M: m. 154),
on C (L: m. 181, corresponding with that on B in L: m. 161).
In the fifth section, the eighth-note graces (U: mm. 21-23) should be
dealt with along the same lines as those in mm. 5-8. In mm. 26-28, two of
the three ascending lines in the bass end with an inverted mordent, and the
same ornament should probably be added on F in m. 28. In m. 29, the
compound symbol above the treble note B indicates a mordent preceded
by an appoggiatura (see the written-out example of the vertical stroke +
mordent that Bach gives in the Table of Ornaments, reprinted in all Urtext
editions). The appoggiatura C should be sustained for one eighth-note, the
ornamental notes B-C-B-C falling on the bass note E and the final B
on the lower-voice D.
There are two relevant motifs. They are introduced in the first and
second phrases of the prelude respectively. M1 is presented by the treble
voice in mm. 1-2. It begins and ends on the fifth degree of the C minor
scale. Its first half measure features the ornamented broken chord G-E-C
in eighth-note motion. Hardly is the keynote reached than it is propelled
suddenly and enhanced by an arpeggioup to the octave. In the motifs
second half, a stepwise descent in dotted rhythm leads to a long final note.
The character of this motif might be described as mild and graceful. As for
the dynamic design, the sudden octave jump offers itself as the obvious
climax toward which the tension in the first half measure rises and from
which it then descends. Within the preludes first phrase, M1 is imitated
(mm. 2-3) and sequenced on the dominant (mm. 3-4). In the imitation of
this sequence (mm. 4-5), the slight abbreviation at the end is due to the
phrasing in the right hand. For the same reason, grace-note and mordent
are not appropriate here.
90
WTC I/4
The only other complete quotation of M1 occurs in mm. 8-9. There are,
however, quite a few developments, deriving mainly from the first half bar.
In mm. 9-10, the right-hands threefold use of this partial motif brings
about a rise in tension. The middle voice answers with a single quotation
that leaves out the octave leap (m. 10-11), and the lower voice continues
with its own little build-up in a twofold motion whose final steps point
downward (m. 11-12). A similar development of M1 can be witnessed in
mm. 24-28: the threefold rise in the right-hand line corresponds with that
in mm. 9-10, and the lower voice also answers with a twofold quotation of
the partial motif, which this time takes up the original octave leap. It is
then extended through two additional sequences with upturned endings,
resulting in a powerful increase toward the peak note F (m. 28m). All the
while the right hand uses the half-measure figure for a gradual descent. A
third development of this kind involves only the right hand: mm. 30-31
feature a fourfold rising motion that prepares a climax on the G at m. 321.
These developments aside, there are short quotations of thematic components that mainly serve to create overall unity. The partial motif frames
the end of the third phrase by appearing in m. 12 (twice) and in m. 14, it
wanders through the voices in the fourth phrase (see the almost continuous
motion from L: m. 15 to M: m. 18), and concludes this section (M: m. 20).
Two further quotations appear toward the end of the piece (L: m. 34 and
M: m. 37). The rhythmic figure from the center of M1 also occurs independently (mm. 9-11, 15, 17, 21-27, 30-31, 35, 36, and 38). Even the cadential
formulas in mm. 13-14 and 34-35 rely in their leading voices on this
rhythmic figure.
M2 is exactly one measure long. First presented in mm. 5-6, where it
begins after the downbeat and ends on the first eighth-note of the following
measure, it consists of an ascending broken chord that brings with it a
natural though soft increase in tension, and a falling scale that resolves this
tension. This motif only appears twice in the prelude: in mm. 5-7 where the
model is followed by two descending sequences, and in mm. 20-23 where
the same progressionmodel plus two descending sequencesis extended
through a fourth sequence whose end melts into the M1 development of
mm. 23-28.
The preludes overall dynamic level remains gentle. As the tension
rises slightly through the first phrase and falls through the descending
sequences of the second, there is an initial furtive climax in the middle of
m. 5. The second overall climax occurs in the third phrase. While it is
somewhat stronger than the first, it is weakened by the fact that the treble
reaches it in the middle of 10, the bass only on the downbeat of m. 12.
C minor
91
The fourth phrase features three dynamic curves in its three subphrases.
They gradually rise in intensity, yet none of them is important enough to
serve as an overall climax. The more significant peak follows at the beginning of the long section where, corresponding with the two initial phrases
of the piece, it falls on the first peak note of M2, at the end of m. 20.
The three most powerful climaxes in the prelude are thus:
right hand,
middle of m. 25,
left hand,
middle of m. 28,
both hands together, downbeat of m. 32.
Thereafter, the tension recedes (no accent on A in m. 33!). The final
passage, after the interrupted cadence in mm. 34-35, features an extremely
soft, subdued build-up through mm. 35-36 to the downbeat of m. 37.
WTC I/4 in C minor Fugue

The subject begins with a whole-note in m. 1 and ends four notes later
with a return to the keynote. The quarter-note C on the downbeat of m. 4
completes a full cadence, since the whole-note D in the third measure,
representing both the subdominant and the following dominant harmony
(for more details see below), finds its resolution onto the tonic on this C.
92
WTC I/4
With its length of a mere five notes,

this compact little unit is one of the
most condensed subjects in Bachs
Well-Tempered Clavier. The subjects
five notes define a single, indivisible
phrase. The pitch range is extremely limited: the keynote is surrounded by
only one note below and two notes above it. Three of the four intervals are
seconds. The remaining interval, however, makes all the difference. This
seemingly inconspicuous little curve features in m. 2 the rarest of hightension intervals, the diminished fourth between B and E.
With regard to the rhythmic pattern, it is necessary to look beyond this
subject at the other components of the thematic material. The subjects
three different note valueswhole-note, half-note, and quarter-noteare
later complemented by much rhythmic variety including eighth-notes and
frequent syncopations, particularly in the form of half-notes tied to a
quarter-note on the following downbeat.
In determining what kind of dynamic design may be expressed in the
brief five-note phrase with regard to its melodic make-up, rhythmic shape,
and implied harmonic progression, one finds that the subject of this fugue
allows for two rather different interpretations. The choice depends on how
the individual performer values the facts mentioned above:
Performers who perceive the specific power emanating from the
diminished fourth in m. 2 will interpret this striking interval as the
subjects center of tension. This concept requires a tempo slow
enough to allow for the melodic step to be fully savored.
Performers who, encouraged by the alla breve time signature, wish
to focus on rhythm, may feel that the long D in m. 3 carries the
climax. Harmonically, this also makes sense as the beginning of
this note represents the subdominant field within the cadence. This
concept, in order to sound consistent, implies a generally faster
tempo for the whole piece. There should also then be less emotional
emphasis on the climax than in the first interpretation.
The choice between these two approaches is so much a matter of
personal inclination that any academic reasoning would seem out of place.
However, one little warning for performers: avoid trying to serve two
masters at one time. An adherence to the second concept will necessarily
have to under-emphasize the interesting interval in favor of the rhythmic
pulse. Equally, an adherence to the first concept cannot, at the same time,
stress the rhythmic element. In either case the subject should convey great
calm together with a peculiarly intense, introverted excitement.
C minor
93
The subject appears twenty-nine times in this fugue. (As this is a

five-part fugue and fruitless discussions about whether there are two altos,
two tenors etc. should be avoided, v1, v2, v3, v4, v5 is here used to refer to
voice one, voice two etc.)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
m. 1-4
m. 4-7
m. 7-10
m. 12-15
m. 14-17
m. 19-22
m. 22-26
m. 25-29
m. 29-33
m. 32-35
v5
v4
v3
v2
v1
v4
v4
v3
v5
v3
11. m. 35-39 v4
12. m. 38-41 v3
13. m. 44-48 v2
14. m. 48-51 v1
15. m. 51-54 v4
16. m. 54-57 v2
17. m. 59-62 v1
18. m. 66-69 v1
19. m. 73-76 v5
20. m. 76-80 v1
21.
22.
23.
24.
25.
26.
27.
28.
29.
m. 81-84
m. 89-92
m. 94-96
m. 95-97
m. 96-98
m. 97-100
m. 100-102
m. 107-109
m. 112-115
v4
v1
v1
v2
v1
v5
v4
v1
v2
Modifications of the subject occur both rhythmically and melodically,

both at the beginning and at the end of the subject: At the beginning, the
interval adjustment characteristic for tonal answers occurs only once, in
mm. 12-13. More frequent are rhythmic alterations: the first note may be
shortenedoften to a half-note (m. 7), once even to an eighth-note (the
upbeat to m. 49). At the end, the resolving note appears in many guises. It
may come early, thus falling on an unaccented beat and sounding harmonically wrong (mm. 14-15). It may also be delayed (as in m. 46) or reached
indirectly (as in m. 41). It may fall on a beat so unlikely for a resolution
that it is not felt as such (as in mm. 50-51), or it may appear integrated into
another melodic unit (as in mm. 57 and 62). Finally, it may even be
completely missing (as in mm. 94-98).
A small but emotionally influential modification occurs in the middle
of the phrase. The diminished fourthan interval essentially characteristic
of the minor scalemay give way to a perfect fourth whenever the subject
is stated in the major mode (as in mm. 29-35 and 54-57). Whenever this
happens, the next interval, originally a semitone, is adjusted correspondingly to a whole-tone descent.3
3
The note group D C F E in v1 mm. 54-57, however, is not a subject entry. The combination of an initial major second with a perfect fourth is too unlikely, and assuming a
stretto at this early stage in the fugue would mean ignoring Bachs design, which seems so
consciously to reserve the stretto intensification for later.
94
WTC I/4
Summing up all these modifications we obtain an amazing result: the

only feature in this subject to remain unaltered throughout the fugue is the
rhythmic pattern of two half-notes in the subjects second measure. And
while parallel statements of the subject are not used, there are a few strettos
toward the end of the fugue (see entries 23 to 26 in mm. 94-99).
The C-minor fugue is widely known to be a triple fugue. This term
indicates that two of the prominent ideas beside the subject do not behave
as mere companions but lead lives of their own. Yet this does not necessarily exclude the existence of a counter-subject. And indeed, at the point
where one expects the first companion to enter, there appears a line that is
characteristic, independent, and recurs later (see v5 mm. 4-8 D to D). The
only detail of this counter-subject that is a bit confusing is its length: CS
surpasses the length of the subject it is supposedly accompanying. (Yet, as
the following will show, a different concept of phrase length is improbable.) CS is taken up, albeit without its first note, in voice 4 (see A in m. 7
to G in m. 11). Thereafter, it never appears again in its full range. There
are, however, several segments that qualify as separate building blocks in
this fugue. These shall be investigated briefly.
One immediately obvious segment to recur separately derives from the
end of CS. It consists of a syncopation (see mm. 6-7: F), sometimes approached with an upbeat (a quarter-note or two eighth-notes), and followed
by a stepwise descent ending in a smooth hook. This CS segment will be
called M1; its frequent appearances can be verified in the five-part sketch.
Note that M1 occurs with greatest density in the first section (mm. 1-221),
which is built in such a way that it generates a powerful dynamic increase,
brought about both by the rising order of statements and by the increasing
number of voices. Owing to its gesture of relaxation, the effect M1 exerts
in this context is one of soothing or balancing. The first segment of CS
also revolves around a syncopation, one that is followed by an inverted
mordent figure (see mm. 4-5). This figure is then dropped and forgotten for
a long time, but resurfaces in mm. 41-42 (v3) and in mm. 45-46 (v1)
before eventually turning out to have been a forerunner of the third subject
(see below). Because of this final relationship with the third subject, it is
here called M3. A last element deserving attention is the rising tetrachord
introduced in mm. 17-19 (see v5: G to C, B to E). It accompanies the
subject particularly between mm. 23 and 35 where, quite in contrast to the
effect of the above-mentioned M1, it reinforces the dynamically increasing
tendency. From m. 35 onward it turns into the upbeat of the fugues second
subject (see below). Because of this later integration, it shall be referred to
as M2.
C minor
95
After having identified the counter-subject as well as the smaller motifs

that, in part, derive from it, the stage is set for the other subjects in this triple
fugue. Their appearance is connected to, though not entirely corresponding
with, the fugues large-scale ternary design. Looking at the composition
without knowing much about it, one would be able to distinguish three
major parts: a first part featuring only occasional eighth-notes (mm. 1-35),
a second part displaying a constant flow of eighth-notes (mm. 36-93), and
a third part where the eighth-notes are again abandoned (mm. 94-115). The
second subject (S2) emerges at the outset of the second part, in the voice
that moves in eighth-notes. Just like the main subject, S2 appears with
various modifications. Trying to pinpoint its main features one finds an
upbeat (which, in its first statement, is the stepwise ascent of M2) followed
by an ornamental line of sequencing one-measure figures hiding a
whole-measure descent (see mm. 36-40: G, F, E, D, C). Both the exact
length, i.e., the number of measures expounding the ornamental figure, and
the exact metrical ending of the second subject remain open to constant
changes. The tension curve in this second subject, however, is quite
unmistakable: a short rise in the upbeat is followed by an extended release
in the descent. In comparison to the first subject, the intensity is considerably reduced. The character of S2 may be described as introverted
and relaxed.
The beginning of the fugues second large part also marks the resurrection of M3. As a pattern consisting of upbeat + syncopation + inverted
mordent, it can be detected already in mm. 36-38 (v5: G C B C) and,
even closer to the original, in mm. 39-41 (v4: B A G F E F G) and
mm. 45-46 (v1: D C B A B). When it finally discloses itself, the third
subject (S3) takes up the more outgoing variants among the traits of M3,
the leaping-fourth upbeat and the inverted mordent in eighth-note motion,
and adds, as a particular characteristic, the repeated-note splitting of the
syncopation. The third subjects character can be described as extroverted
and active. S3 plays a major role in this fugue, from its initiation in
mm. 49-51 (v3) to the end of
the composition. The dynamic
outline in S3 consists in a rise
toward the downbeat (i.e., to
the last of the repeated notes),
followed by a relaxation.
96
WTC I/4
Episodes are generally defined as subject-free passages. In a fugue

with three subjects, only those measures that do not feature any of the
subjects would therefore truly qualify. These are very few. In the first part
of the ternary design, before the entries of S2 and S3, there are only two
short stretches; these are matched later in the fugue by two more:
E1 mm. 102-122
E3 m. 88
E2 mm. 172-192
E4 mm. 1092 -1121
The most prominent feature in E1 is M1; the soothing quality of this
motif defines this first episode as a softening break that allows breathing
space before the plunge into the next span of tension build-up. E2 is
determined by M2. The role of this episode in the overall development of
tension is therefore that of a link continuing the tension increase. In E3, the
little M1 figure that seemed long forgotten is briefly recalled, thus giving
this episode again a calming quality.
Finally, E4 is conceived as a cadential close. Sounding as it does at the
end of a longer dominant pedal that had already provided the backdrop for
a tension increase, this cadence promises relief that it then withholds: its
final chord is an oddity and deserves more detailed discussion. Following
the dominant-seventh chord in m. 111 there are theoretically three different
resolutions listeners might expect: the tonic (C minor), the Picardy-third
modification of the tonic (C major), or the interrupted cadence with a
chord on step vi (A major). What Bach gives us is not one of these three
but a combinationand the oddest possible combination, for that matter.
The left hand in m. 112 plays the third A-C, representing the chord on vi,
while the right hand simultaneously intones the E-G-C of the Picardythird tonic. This results not only in a polytonal effect but also in an implied
clash between the implied fifth E of the A major chord and the sounding E
of the C major harmony. What Bach expresses by using this artful though
daring device seems to be this: the cadentially confirmed ending in the
home key at this moment still leaves so much unresolved tension that it
demands yet another pair of S1/S3 statements. Yet these should not sound
as a coda. Perhaps Bach created the incomplete harmonic resolution that
binds the final cadence as a necessary completion to the one in mm. 111112 in order to achieve this effect of belonging.
The complexity of the rhythmic pattern and the structural design in
conjunction with the predominance of stepwise motion in all but the S3
upbeat all favor the interpretation of this fugue as one of rather calm
character. Upon closer inspection of the emotional implications of this
character it becomes obvious that there is a distinct hierarchy within the
components of the thematic material: The main subject is defined by a
C minor
97
strong introverted urge, the second subject is generally lighter in its mood,
while the third subject shows a jolly, extroverted vivacity.
The three subjects appear in all possible combinations. Determining
what kind of mood or spirit results from each of these juxtapositions lays
a good foundation for an understanding of the entire composition.
The coupling of S1 + S2 brings about a quality that is more serene,
less urgent than that which determines the main subject when heard
alone (S1 introverted/intense + S2 introverted/relaxed).
In the S2 + S3 combination, all the darker drive has gone. What is
left is a momentarily intense quality, soon giving way to one of
relaxation (S2 introverted/relaxed + S3 extroverted/active).
Conversely, when S1 is paired with S3, two strong forces join,
expanding both inwardly and outwardly (S1 introverted/intense
+ S3 extroverted/active)
The range of tempo in this composition is confined on the one hand by
the basic character, which is rather calm, on the other hand by the time signature, which requires an alla breve pulse. Ideal is an eighth-note motion
that is unhurried but can nonetheless be clearly felt in half-measure groups.
The corresponding articulation requires legato for almost all melodic notes.
An exception can be made within the third subject where the first interval,
the ascending fourth, and the ensuing note repetitions may sound in a
bouncing non legato in order to give credit to the special outgoing mood
of this component of the material. The Urtext of this fugue does not
indicate any ornaments.
The relative tempo of the prelude to the fugue that does justice to their
characters sets larger metric units in proportion. Owing to the strong
half-measure feeling in both pieces, a good effect is achieved like this:
half a bar
corresponds with
one bar
in the prelude
in the fugue
(Approximate metronome settings: prelude beat = 100, fugue beat = 66.)
The fugues design is complex. Within the first large part, the full
ensemble of five voices is attained only transitorily in mm. 19-20; the fifth
entry and most of the redundant subject statement sound in still reduced
ensemble. After the ensuing cadence in mm. 21-22 the number of voices
drops to only three, a fact suggesting the beginning of a new round. This
second section ends in m. 35 in full ensemble and with a pronounced
cadence, followed again by a reduction of the ensemble to three voices.
The third section introduces the second subject, which sounds both
against the main subject and independently, both in its original shape and
in inversion (see mm. 41-44). In m. 49, the third subject enters for the first
98
WTC I/4
time. Its immediate combination with both S1 and S2 marks the beginning
of something new: the fourth section. From here to the end of the large
middle part, there are two structurally analogous progressions:
mm. 48-72
correspond with
mm. 73-93
5 three-subject juxtapositions : 4 three-subject juxtapositions
grouped 3 + 1 + 1
:
grouped 2 + 1 + 1
(mm. 48-57 / 59-62 / 66-69
:
mm. 73-79 / 81-84 / 89-92)
interspersed with two passages featuring only S2/S3
(mm. 57-58, mm. 62-64 : mm. 79-80, mm. 84-88)
rounded off by passages with distinct cadential-bass steps
(mm. 69-71
:
mm. 92-94)
After the end of these two sections, the fugues fourth and fifth, the second
subject is dropped. Instead of three-subject juxtapositions, the sixth section
begins with S1 and S3 in simultaneous stretto. All this together provides a
picture of six sections, with the following statements of the main subject:
S1:
I
mm. 1-22
v5, v4, v3, v2, v1, v4;
II
mm. 22-35
v4, v3, v5, v3;
III
mm. 35-48
v4, v3, v2;
IV
mm. 48-72
v1, v4, v2, v1-v1;
V
mm. 73-93
v5, v1, v4, v1;
VI
mm. 94-115
v1v2, v1v5, v4, v1, v2.
As both the second and the third subject enter considerably later, it is
to be expected that their own rounds do not entirely coincide with those
of the main subject. For the fifteen statements of the second subject, four
groups can be made out:
S2:
I
mm. 35-48
v1, v4, v5, v1;
II
mm. 49-65
v2, v3, v1, v2;
III
mm. 66-71
v5, v1;
IV
mm. 72-94
v3, v5, v2, v1, v4.
The third subject enters last but seems the strongest among its rivals. It
surpasses both the other subjects in the number of its statements. It catches
up with the second subject by already finishing its fourth round in m. 92;
and it equals the main subject insofar as it also builds a total of six rounds.
The 37 entries of the third subject are as follows:
S3:
I
mm. 49-64
v3, v1, v5, v2, v4, v1;
II
mm. 64-73
v3v5, v2, v3, v2;
III
mm. 74-84
v4, v2, v5, v3;
IV
mm. 84-92
v4v2, v5, v5;
V
mm. 92-101
v1v2, v4v3-v4v3, v2v4,v1, v5;
VI
mm. 102-115 v4v3, v2v4, v3+v2v4, v3.
C minor
99
The harmonic outline will here be described with reference to the main
subject. Its first six statements, all on the tonic or dominant of C minor,
are concluded with a cadential close in G minor in m. 21-22. The next
four entries effect a modulation to the relative major key, through F minor,
C minor, and B major to E major, with a cadential close in m. 34-35. The
third section finds S1 returned to the C minor field. It closes, interlocked
with the beginning of the next round, in C major at m. 491. The fourth
section begins firmly anchored in the subdominant, with two entries in F
minor and one in the relative key of A major. However, the last two S1
statements revert to C minor, with the repeated entry in the first voice
ending in a floating position on the dominant-seventh chord. The fifth
section confirms the return to the tonic by placing all its four S1 statements
on C itself and concluding this part with a cadential close in C minor
(mm. 93-94). The final section basically remains on the tonic. Stretto
entries occur on various pitch levels but all relating to an implied key of C
minor. From m. 105 onward the G pedal, giving way only momentarily to
its leading-note F, prepares the end of the composition, which is reached
with a final subject statement over a tonic pedal.
100
WTC I/4
Within the first section, a gradual rise in tension is brought about by

the increasing number of voices, enhanced by their rising entering order.
As has been mentioned above, there are forces that serve to hold the
tension back and prevent a premature full-fledged climax: the relaxing
tendency of M1 and the fact that this section, belying the obvious expectation for a fugal exposition, does not close in full ensemble.
The second section brings the modulation to the major mode and with
it a six-bar-long stretch in five voices. In addition, the calming influence of
M1 is replaced by the forward-pushing gesture of M2. Thus the second
tension increase is considerably stronger than the first. It creates the
impression that it means to unite the first two sections under one common
target. After this, the third section maintains a low tension profile as
there is neither an increase in the number of voices nor any harmonic or
structural device that might engender a rise.
The two analogous sections that follow describe analogous dynamic
curves. The climaxes fall at the respective beginnings, where three-subject
juxtapositions create a state of very high intensity that abates only very
gradually, through the interspersed S2/S3 paired statements, the following
entry group, and the final cadential-bass steps. Toward the end of the
fourth section, this tension decay is melodically supported by spans of
chromatic descent. An equivalent in the fifth section can be observed in the
unexpected re-emergence of the soothing M1.
At the beginning of the sixth section, the previously achieved intensity
of three simultaneous subject statements is surpassed by the combination
of the two stronger ones that, for several bars, now both sound in stretto.
(Whether such unheard-of density can actually be grasped by the human
ear and mindmost probably it cannotis beside the point, last but not
least because Bachs music was certainly to a large extent written For the
Greater Glory of God. But the effect of this immense density is undoubtedly stunning.)
The middle of this section then brings forth a sudden strong decline in
tension, enhanced by the chromatic descent in the uppermost voice (see
mm. 101-105). As this descent touches ground, it is met by the entry of the
dominant pedal. This gives new impetus to the tension, which rises again
much faster than it had previously fallen. A parallel statement of S3 tries to
surpass the earlier strettos in intensity and is supported by the harmonic
sidestep to the diminished seventh chord. Even the following cadence does
not yet bring about the deserved relief but instead a climax of yet different
means: the momentarily bitonal cadential close mentioned above.
WTC I/5 in D major Prelude

Underneath a virtuoso surface structure the D-major Prelude is entirely
determined by harmonic processes. Its 16th-note pattern does not develop
motifs. Any recurrences are transpositions that occur exclusively in connection with analogies in the harmonic development.
The first cadence ends at m. 31, after a harmonic progression from the
tonic (m. 1) via ii7 and V7 (m. 2) back to the tonic. This cadential close
comes with a subtle structural break. The next harmonic progression
modulates to the dominant. During this modulation, the right hands ornamental line resembles that of the first cadence (compare mm. 3 and 4 with
m. 2); its end in particular is an exact transposition of the previous end
(mm. 53-61 . 23-31). As before, this cadential close signifies the completion
of a structural unit. There are altogether eight closed harmonic progressions.
Note that each phrase begins with the second 16th-note in the right-hand
pattern. The upper/lower-case letters indicate major/minor keys:
mm. 1-31
DD
mm. 14-201 DG
mm. 3-61
DA
mm. 20-221 GG
mm. 6-121 Ae
mm. 22-251 GD
mm. 12-141 eD
mm. 25-35
DD
This prelude, like several others from the first volume of the WTC,
derives from a much shorter model in Bachs Note Book for Wilhelm
Friedemann. The original version comprises 22 measures. Of these, mm.
1-17 are retained unchanged, while mm. 19-20 can be found in mm. 27-28
of the WTC version. An interesting question is therefore whether the basic
binary form of the Wilhelm Friedemann version has only been extended,
or whether the addition of mm. 18-26 and 29-35 changes the structure.
No portion of the D-major prelude is repeated literally. There is, however, a striking five-measure analogy: the preludes beginning is restated in
the middle of the prelude in the subdominant key (mm. 1-61 . 20-251).
The modulation from tonic to dominant that takes place at the beginning of
the prelude thus serves, when launched from the subdominant, to return to
the home key in the manner of a typical Baroque recapitulation.
Furthermore, there are several smaller analogies: The end of the modulation in the second section recurs as a bracket in the subsequent section:
compare mm. 5-61 with mm. 6-71 and 10-111, and the beginning of this
101
102
WTC I/5
modulation then reappears with some interval adjustments at the outset of

the ensuing section: mm. 3-43 . 14-153. A further correspondence includes
2 measures built on a pedal note, followed by another 1 measures that
conclude the modulation (compare mm. 8-121 with mm. 16-201). The bass
line of these two passages differs slightly while the right-hand part is an
almost exact transposition, with only a tiny adjustment in the last half
measure.
The choice for the preludes tempo should be made after considering
the following two aspects: The bass line is composed as a hidden two-part
structure. This might indicate that an alla-breve pulse should be made
perceptible behind the given 4/4 time signature. The treble shows a pitch
pattern that is easily unmasked as an ornamentation of broken chords. A
melodious treatment of this line would mislead listeners into expecting a
true melody that then does not materialize. The appropriate tempo is therefore fairly fluent: fast enough to give the impression both of a half-measure
pulse and of a non-melodious, arabesque-like right-hand pattern. The
corresponding articulation is legato or, better still, quasi legato for the right
hand. The left hand plays non legato; because of the obvious rests in the
bass part, this indication affects the touch and coloring of the notes rather
than their actual duration.
Exceptions from the patterns established in m. 1 only begin with m. 27
and become more prominent from m. 32 onward. Here, a peak-note line in
the right hand is marked as a suddenly emerging melodious upper voice.
Owing to the features of this line, a syncopated rhythm and a dosido
formula, legato playing is indispensable here. (This can either be done by
fingering 45 on each of the three syncopations, or by playing these notes
with 454 right away). At the same time, the newly created tenor
should sound non legato. The same holds true for the second half of m. 34
where the soprano takes up the typical closing formula while the simultaneous quarter-notes in the left hand continue the non-legato touch.
The only ornaments appearing in this prelude are the two arpeggios in
mm. 33 and 34. They invite the same questions as the arpeggios in the
C-minor prelude: Is the top of the arpeggiated chord primarily a melodic
note, or is it primarily part of the chord, with no more melodic value than
any note in the chords middle register? The answer, as can easily be seen
from the score, will be different for the two cases. In m. 33, the treble D is
both an immediate continuation of the preceding melodic motion and the
point of departure for the continuing line. In m. 34, by contrast, the logical
target of the run in the preceding bar is the lowest note C, while the
upper-register B is only a chordal note. Admittedly, this B expresses high
D major
103
tension, representing the diminished seventh of this chord and moving up

chromatically to the B in the subsequent chord. Yet it is very important to
realize that neither peak note carries the melodic flow, which can rather be
traced like this:
mm. 32 . . . . . . . . . . . . . 33 . . . . . . . . . . . . 34 . . . . . . . . . . . . . . . . 35
F 2 E 2 D 2 C 2 D 2 . . . . . . . . . (D1) C 1 D 1 D 2 C 2 D 2
Recognizing this melodic progression is particularly vital for performers
as it will influence the execution of the two arpeggios: In the first arpeggio,
the melodic note D should sound on the beat, together with the bass note
A, while the remaining notes of the arpeggiated chord follow swiftly.
Conversely in the second arpeggio, the melodic target is the C at the base
of the chord. In keeping with the rules of Baroque ornamentation, it will at
any rate fall on the beat; the treble B sounds here at the end of the broken
chord, slightly after the actual downbeat.
The preludes initial two measures establish the D-major key with a
simple cadence. The harmonic curve contains no unexpected features; the
subdominant harmony, here represented by a ii7 chord, is reached in the
first half of m. 2, followed by the dominant and the resolution onto the
tonic. The ensuing modulation builds up slightly more tension toward a
climax, in the first half of m. 4, on the inverted seventh chord that brings
the shift from one key area to the next. After this harmony, the tension
subsides gradually until the resolution on the downbeat of m. 6.
mm. 1
2
3
4
5
6
p
mp p
mp + mp p
The next measures represent a transposition of the preceding ones, with
a similar climax in the first half of m. 7. The three final notes in m. 7 then
announce a change, which materializes in the F pedal extended for two
and a half measures. During these measures (mm. 8-9), the tension that had
just begun to decrease after the climax, rises again. This rise is more
gradual than the former increases because it is not triggered by an active
harmonic step but rather by the persistence of the pedal note. At the end of
the pedal, m. 10 is the transposition of m. 5. In keeping with this structural
analogy, the tension-decrease should end on the downbeat of m. 11 where
the target harmony of this modulation, the E-minor chord, is reached. Yet
Bach decides not to settle on this harmony. As if the target key had been
reached too suddenly after the build-up over the pedal, he adds another
measure confirming the E-minor tonality at m. 121. Then, with a swift turn,
he reverts to the key of D major, which is consolidated by an extra measure
on the tonic (m. 12 = modulation back to D, m. 13 = confirmation of D).
104
WTC I/5
These two measures, which in the harmonic development represent a

modulation of their own, thus appear more like an appendix to the previous
progression. The entire segment, expressed in terms of dynamics, will
appear somewhat like this:
mm. 6
7
8
9
10
11
12
13
14
+
p
mf
mp mf
mf
mp mp p
p
A process roughly similar to that described above occurs in the subsequent section. The first measure is yet another version of the active part of
the modulation. After a renewed climax in the first half of m. 15, the
tension begins to fall but is suspended in its descent by a pedal of 2
measures on B (mm. 16-18). Here, the hidden two-part structure that so far
characterizes the bass pattern is replaced briefly by a broken chord pattern,
thus giving these two measures the even larger frame of whole-measure
pulses. At the conclusion of the pedal, m. 18 seems to lead to a new key
(A minor). Yet again this key is not confirmed but passed through on the
way to G major. This time, however, there are no repetitions of the two
target harmonies. As in the corresponding measures discussed above, this
final small-scale modulation sounds like an appendix to what precedes it.
mm. 14
15
16
17
18
19
20
p
mp + mp mp + mf mp p
The recapitulation of the first five measures should, of course, retrace
the dynamic outline from the beginning of the prelude; in other words:
mm. 20
21
22
23
24
25
p
mp p
mp + mp p
The preludes last section begins with two measures that prepare the
final pedal on the dominant. Unlike similar final pedals in other preludes,
this dominant bass note is not preceded by its leading-tone. Here, instead,
this leading-tone is made up for in the middle of the pedal (m. 302-31)
where it combines with virtuoso peak notes in the right hand and thus
creates a strong emotional climax. After this interruption, the dominant
pedal continuespartly as a sustained note, partly implied under the
cadenza-like setting of the right hand in mm. 33-34and only resolves
onto the tonic in the very last measure.
For the dynamic design this means a steady increase, with only slight
inflections on the surface in those measures where the harmony seems to
come to a halt, i.e., mm. 30-31 and m. 33. The overall dynamic development in this section may be described as follows:
mm. 25 26 27 28 29 30 31 32 33 34 35
p
mp mp mp + mf mf mf + pf pf
pf + f
D major
105
The following diagram attempts to capture the dynamic layout:
WTC I/5 in D major Fugue

This subject is roughly one measure long. It begins after a quarter-note
rest at m. 12, ends at m. 21, and consists of an indivisible phrase. Its pitch
pattern features a predominance of steps interrupted by one major-sixth
leap. This interval is not one of the so-called high-tension intervals. The
rhythmic pattern includes three kinds of considerably different note values:
the longest (dotted eighth-notes) are six times as long as the shortest (32ndnotes). Later in the fugue, quarter-notes are added. The rhythmic organization alludes to a specific musical genre: The almost constant presence of
the dotted-note pair, interspersed with fast ornamental groups, is typical of
the French overture. The fact that the fugue is modeled on such a distinct
character piece accounts for most of the particularities that would, without
this background, seem disconcerting in a polyphonic composition.
106
WTC I/5
The subjects harmonic background is that of a simple progression,

with the sequence T S D7 T. All musical aspects of the phrase unite in
favoring one note for the climax: the dotted B in the center. Harmonically
this B represents the active step (the subdominant), melodically it is
reached in the most expansive leap, and
rhythmically it represents a surprisingly
long note after a bustling beginning. While
the tension-rise before this climax sets out
from the keynote in complete relaxation
and escalates in almost no time, the subsequent resolution is much more
gradual. As the subject ends on the third (F), the tension does not entirely
fall back to the level of the beginning.
There are altogether 12 subject statements in this fugue:
1. mm. 1-21 B
5. mm. 7-81 B
9. mm. 13-141 S
2. mm. 2-31 T
6. mm. 8-91 S
10. mm. 14-151 T
3. mm. 4-51 A
7. mm. 11-121 S
11. mm. 15-161 B
4. mm. 5-61 S
8. mm. 12-131 A
12. mm. 24-251 B+S
This subject undergoes none of the modifications otherwise encountered in Bachs fugues: it is never inverted, augmented, or diminished, and
its pitch and rhythm patterns remain untouched. The only exception occurs
in the final statement (B: m. 24-25), which some scholars do not consider
equal to the other statements as it is extended in both directions. The
bustling 32nd-note group from m. 244 is preceded by three similar groups
that gradually ascend in fourths toward the tonic level (see the groups from
B, E, and A). Symmetrical to these anticipating figures, the subjects
second half is extended with sequences in m. 26: the original notes of this
half, G F E D, beginning here with an octave displacement, recur in a first
transposition a fifth down (C B A G), while a second sequence is modified
to accommodate the cadential-bass steps (F E D A). Moreover, while
there are no real strettos in this fugue, the final entry is partially set in
parallelsand as the parallel begins one beat (i.e., one whole figure) later,
the deceptive impression of a stretto does arise for a moment. To complete
this unusual entry, in its second half and throughout its sequences the three
upper voices move in rhythmic alignment but in contrary motion (see the
right-hand chords in mm. 25-26). The strong homophonic element in this
final, extended subject statement enhances the outsider position it occupies
in the design of the fugue.
D major
107
The counter-subjects are unusual. They confirm that the French overture,
an essentially homophonic genre, is indeed one of this fugues god-parents.
The contour appearing in continuation of the first subject statement and
against the second entry consists of the notes G-F-E-D-E-A. It meets the
three basic requirements of a counter-subject: to be easily recognizable,
independent, and taken up again later in the fugue. Nonetheless the term
counter-subject seems a little too big for this simple unit, not least because
it sounds very much like a cadential-bass pattern and does not wander from
voice to voice as a polyphonic unit should, but remains essentially at the
base of the texture. A close look at the two other counter-subjects comes to
similar conclusions: Although they are formally independent of the subject
and taken up repeatedly in the course of the fugue, the fact that they
resemble closing formulas prevents one from taking them too seriously as
counter-subjects.
Dynamic shaping in the contrapuntal components allows for but little
independence. The cadential-bass pattern of CS1 should logically approach
the note that represents the subdominant as its climaxa climax thus
coinciding with that of the subject. Similarly, CS3 with its quarter-note/
half-note/quarter-note rhythm will accent the half-note, which falls again on
the same beat as the other climaxes.
Only the syncopated CS2 could sound
in diminuendoif it were left alone;
but more often than not its first note
falls together with that of the subject
beginning (see mm. 4, 5, 8, and 13).
The dynamics of these four parts thus
give a distinctly homophonic result.
Five times are the subject entries
interspersed with episodes: E1 = mm. 3-4, E2 = mm. 6-7, E3 = mm. 9-11,
and E4 = mm. 16-23. Among the three motifs that make up the episodes,
one, appearing in four of the five episodes, uses the 32nd- note figure from
the subjects beginning. Moreover, E4b (mm. 17-23) is dominated by the
dotted rhythm of the subjects second half. Hence, most of the episodes in
this fugue are closely related to its subject.
The close relationship with the primary thematic material explains
some of the dynamic developments in the episodes. M1 is an ornamented
version of the unresolved second counter-subject (mm. 32-41 . A: 132-141),
a descending step in decreasing intensity. In analogy to the subjects
second half, M3 also proceeds in decrescendo. Conversely, M2 with its
relationship to the subject head seems at first to build up tension. Yet there
108
WTC I/5
are three reasons that speak against this rendition: on occasion of its first
entrance in mm. 6-7 the motif ends with the neutral interval of a fourth. In
all other episodes it is launched on a strong beat and ends on a weak one.
And above all: the fugues structure would become incomprehensible were
all M2 entries to sound at the energy level as the subject head.
The role played by the episodes in the course of the fugue is also
puzzling. E1 with its underlying lines of parallel syncopations links two
pairs of subject entries within the same section. Z2 has the same effect,
although it marks the opening of a new section. By creating this relaxing
effect in front of what should be a new impulse, Bach seems to thread the
first two sections closely together, so that the feeling of a structural ending
only arises when a new tonal sphere is reached, i.e., with the downbeat of
m. 9 that establishes B minor. E3 is the first episode to be self-contained.
It seems intended as a buffer between two larger portions of the piece.
The pattern of the subjects statements in mm. 9-16 (S, A, S, T, B) with
its twofold soprano lead suggests a subdivision into two rounds. The
surprising deceptive cadence at the end of the alto entry and the M2 variant
accompanying the concluding tone underline a caesura that is otherwise
hardly tangible. By contrast, the extensive final episode is subdivided in
itself: the brief first segment (E4a) is a cadential close and thus brings a
feeling of relaxation (mm. 16-171), while the longer second segment (E4b)
develops the material of E3 (mm. 17-19 + 21 . 9-11). After descending
sequences in mm. 17-20 it uses a dense web of imitations to rebuild tension
within a single measure. The subsequent cadential close with its soprano
ascent to the third does not allow a complete resolution. After another
measure with M3 in three-part homophonic texture Bach repeats the
cadential close, now with conventional formulas in the soprano and bass.
The character of this fugue is determined more by the mood of a French
overture than by the conventions it shares with other polyphonic compositions. It is stately, lively but not light. Articulation in French overtures is
an intricate affair: The 32nd-note groups are legatovirtuoso figures that
should never sound melodious, which gives an indication for a minimum
tempo. The dotted-note groups reflect the stately character of this piece
best if played in a rather heavy non legato. The longer note values in the
subjects companions require three slightly different types of articulations:
neutral non legato in the cadential figure of CS1, a more melodious non
legato in CS3, and true legato in the dosido formula of CS2.
Of the 16th-notes in the upper part of M3, the first threea written-out
inverted mordentshould be played legato, while the fourth 16th-note
may sound as non legato as the accompanying voices.
D major
109
As the prevalent features in this fugue and its prelude are so different,
a risk of monotony does not arise. The tempo proportion may therefore be
simple: a half measure (two quarter-notes) in the prelude corresponds with
one quarter-note in the fugue. (Approximate metronome settings: prelude
beat = 120, fugue beat = 60.)
The fugue features three ornaments that, based on their print size in the
Urtext edition, stem from Bachs own hand: the cadential ornaments in
mm. 20 and 22 and the grace-note in m. 10. The first, a mordent, should
consist of four notes (D-C-D-C), with a triplet of 64th-notes against the
first two 32nd-notes in the left hand. The symbol for the compound trill in
m. 22 indicates a turn + mordent figure: D-C-B-C-D-C-D-C must be
fitted into less than a dotted eighth-note as the anticipation should come
after a short stop. A brilliant 64th-note execution, with the final C on or
before the dot, is the solution. The grace-note in m. 10 represents an appoggiatura, held for only a 16th-note to avoid the D-C parallel with the bass.
If played, it must be transferred to S: mm. 11 and B: mm. 18, 19, 20
Other ornaments stem from copies. They may or may not go back to
the masters suggestions but are worth considering. The subject could be
played with ornaments on two of its dotted notes. The first is an inverted
mordent (to be played with a whole-tone neighbor in all cases apart from
mm. 8 and 15). The second is a mordent with three or, better still in overture style, five notes. Both ornaments must be either omitted or included in
each subject statement. The mordent in M1 (T: m. 3) seems inconsequential
but sounds pleasant. The inverted mordent in m. 15 may serve to enhance
the fact that this subject statement, unlike all others, ends unresolved. The
two cadential ornaments in mm. 16 and 26 are very typical. Both begin
with the main note, ascend twice to the upper neighbor note, and end in a
suffix. Their rhythmic realization depends on a decision that, unique to this
particular fugue, has an impact on the rendition in general: Performers who
wish to underline the French-overture character of the fugue will render
these ornaments in the corresponding style, i.e., as fast as possible, concluding them in the time span of a dotted eighth-note or less and leaving
ample room for a point darrt. Performers preferring to stress the fugue
over the genre piece will interpret the compound symbol as an indication
for a note-filling trill; they will therefore sustain the initial main note for a
16th-note, followed by three regular 32nd-note-pairs.
This brings us to a rhythmic feature to be mentioned in this context.
French overtures are known to feature over-dotted rhythms: some of the
notes written with single dots were traditionally played as if they were
double-dotted. Again: whether one wishes to play this composition in the
110
WTC I/5
style of a true French overture or regard it as after all, a fugue depends

on personal taste. In the first case, however, it is vital to know that the
over-dotting does not apply equally to all dotted notes but only to those
that are complemented by a single note, fall on one of the strong beats in a
bar, and are not accompanied by lines that indicate normal dotting. To give
an example: over-dotted eighth-notes are possible in mm. 13, 21, 23, 43, 51,
and 53. Conversely, in the episodes in mm. 3-4 and 6-7 the strong-beat notes
are complemented by three faster notes that must sound in their written
value. In mm. 9-10, the spelled-out ornament of the upper part indicates
that the parallel voices, too, are to be played as written. As a result, the
following is possible: Subject and CS1 can be double-dotted in their
middle; the subjects closing note can be double-dotted when followed by
another subject statement (mm. 2, 5, and 12) or a sequence (mm. 25-26),
but must be simply dotted when followed by a group of notes (mm. 8 and
14). All episodes in this fugue seem not to allow for double-dotting, except
perhaps for the notes in the cadential formulas (mm. 203 and 223).
There are a few indicators of structural design in the surface features.
The first subject statement to sound in reduced texture after the initial
ensemble buildup is that in m. 11. The two-part setting here constitutes a
dramatic reduction from the full ensemble of mm. 5 and 7-8. Furthermore,
as mentioned above, the fugue features only one episode that manifestly
concludes a section: that in mm. 15-171. We can thus deduce that there are
two partitions in this fugue: one in m. 11 and another in m. 17. However,
the assumed first section with its six entries contains one subject statement
more than it should. This section must, therefore, be further subdivided
below the surface. The same holds true for the second section: what
points to sub-grouping here is not so much the number of entries but the
fact that the soprano sounds both at the beginning (mm. 11-121) and in the
middle (mm. 13-141). The third section is unusual in a fugue for other
reasons. It consists mainly of an extensive episode that is twice interrupted
by a cadential formula reconfirming the return to the home key. Its last
portion finally leads into a subject statement that sounds quite different
from what one might expect: luxuriantly extended at both ends, with the
expected contrapuntal texture abandoned for parallel motion in the first
and homophonic chords in the second half.
These results are supported by the harmonic development. After five
entries in D major, the sixth is in the subdominant. This launches a modulation that passes through the relative key of B minor (m. 9) and returns to
the subdominant G major (m. 11). The point at which this new key is
established coincides with the subject statement in reduced ensemble and
D major
111
thus indicates the beginning of a second section. Moreover, although the

fifth entry (mm. 7-81) sounds in full four-part texture, it is accompanied by
only one of the contrapuntal figuresthe other two are substituted by
episode-style material. This statement is also the first in the fugue to end
not in a resolved root position but instead in an inverted seventh chord. It
is thus weakened in two respects, and the concluding D7 chord links it to
the target key, G major, of the subsequent modulation. Bach seems to create
two subtly distinct structural levels. The first ten measures consist of two
only slightly separated sections, one that presents the buildup of the
ensemble in ascending order and introduces the three contrapuntal figures
but is harmonically static, and one that begins with the subject bolstered
only by CS1 and two neutral lines in homophonic pattern, is harmonically
weakened by its resolution onto a seventh chord tending toward the key of
G, but followed by an entry that provides both the expected new key and
the four-part ensemble with all three contrapuntal figures.
The second part shows a similar sub-surface split into two sections.
Indicators appear once again in the harmonic development: The statement
in m. 11, in the new key of G major and in two-part texture, is followed by
an entry in three-part ensemble modulating to the relative minor key. This
relates it directly to the sixth entry, which brought about the same modulation. The next entry returns to D major; in addition, its resolution chord
contains the seventh (C) that relates this statement to the fifth entry in the
fugue (the second sections first statement). Furthermore, the three-part
texture is retained and not yet complemented with the missing fourth
112
WTC I/5
voice. The next entry sounds in full ensemble and in the expected G major;
but again, its resolution comes with a seventh (this time the major seventh).
The third entry in this group reaches E minor, the subdominant relative and
the key of the explicit cadence at the end of this part. The ensuing sevenbar episode is longer than any of the preceding sections; it is even longer
than the entire six-measure second part.
Within the first part, the four subject statements of the first section
bring forth an increase in texture. The increase in loudness, however, is
comparably small because already the first statement should have fully
declared the stately character of the piece. Throughout sections I and II, the
tension remains almost unchanged. The two similar episodesone within
the first section, the other between the two sectionscause slight but inconsequential relaxations. It is only the third episode, at the end of the first
part, which succeeds in conveying a more noticeable tension-decrease.
Within the second part, the buildup is again caused by the growing
number of voices. While among the two statements in three-part texture,
the alto statement in mm. 12-13 with its exposed pitch position sounds
more outgoing than the soprano statement in mm. 13-14, the ensuing tenor
statement in four-part texture and the last entry in the bass, accompanied
by a parallel of CS3 in the two middle voices, share in the climax. As the
end of this statement does not bring about the expected harmonic resolution but keeps the middle voices in suspension, the tension is allowed to
subside only very gradually in the subsequent cadential measure.
The long episode opening the third part begins with the material from
the tension-releasing third episode and thus creates a sense of continued
relaxation until m. 201. Immediately following this, the fourfold imitative
interplay of M2 propels the tension so high that even the cadential pattern
with its typical closing-formula mordent cannot dissipate it fast enough.
Another measure with M3 and the varied repetition of the cadential close,
this time without the stormy 32nd-note stretto, is required to achieve relaxation. The following measure with its descending tendency continues this
direction. As a result the final, unusually extended subject statement with
its powerful parallel motion begins from a relatively soft level, only to
unfold its tremendous buildup all the more forcefully.
The relationship between the three parts is obvious from all that has
been said above: the first two parts are similarly built, with rises both times
from about mf to a good f. The third part seems to delay the outbreak, and
when it finally allows its only subject statement to make its declaration,
this by far outranks that of all the previous entries.
WTC I/6 in D minor Prelude

This prelude begins in purely harmonic design. Against this backdrop,
melodic ideas gradually emerge, repeatedly emancipating themselves from
simple bass patterns. The prelude should therefore be analyzed under the
criteria of both harmony and motivic development.
The first harmonic progression ends with the lower D that serves as a
tonic pedal during the first five beats of the piece. (In the right hand, the
conclusion falls on the fifth triplet-16th-note of m. 2). Since this progression unfolds above a pedal-note bass, it conveys the impression that
the piece has not yet quite begun. This cadential close is thus not to be
regarded as structural, but as a first segment of a larger unit. The full first
section modulates to the relative key, F major. It closes in m. 6 (right hand:
on the eighth triplet-16th-note). The structural caesura is enhanced by a
distinct cadential-bass pattern in mm. 5-6. Regarded under harmonic
aspects alone, the sections of this prelude are as follows:
I mm. 1-6 tonic to tonic relative (D minor to F major)
II mm. 6-8 on to subdominant
(F major to G minor)
III mm. 8-10 on to dominant
(G minor to A minor)
IV mm. 10-15 back to tonic
(A minor to D minor)
V mm. 15-26 tonic confirmed
(D minor)
A close look at the melodic structure reveals that the three sections in
the center are closely interrelated: section II sets up a motif in the lower
voice, which is then sequenced one tone higher (L: mm. 6-8 . 8-10) and
taken up again after a short interruption (L: mm. 12-14). This time, the
expected cadence is delayed and appears only after an additional measure
in the lower octave. Moreover, the interruption just mentioned features
a short motif (see L: m. 10, A A G F) that is sequenced three times.
Extending the analyses to motivic development, the previous layout can
thus be condensed into three major sections:
I mm. 1-6
(D to F, no motif yet present )
II mm. 6-15 (F to D, two motifs and their development)
III mm. 15-26 (D to D, motifs 1 and 2 no longer present)
The sequential structure in the three harmonically closed progressions
in the central section apart, the prelude does not contain any transpositions,
variations, or other analogies. Only one very short correspondence should
113
114
WTC I/6
be mentioned as it will add to the understanding of the overall design:

m. 15 recurs in m. 23, transposed to the dominant and enriched by an
additional middle voice.
The continuous flow of broken-chord patterns in the right hand might
indicate an ideal tempo for the piece. As in preludes determined solely by
harmonic processes, these broken chords should sound fast enough to allow
listeners to perceive them as harmonic entities rather than melodic zigzag
lines. Neither should they appear as sparkling virtuoso figures, so as not to
divert listeners from grasping the inherent harmonic design. The articulation reflects the rather lively basic character with legato for the triplet
16th-notes and non legato for the eighth-notes. Within this non legato, color
shading distinguishing between neutral bass notes and more emotional
motivic notes is, of course, very much encouraged.
The first section in mm. 1-6 displays all the characteristics of a piece
determined solely by harmonic processes. Each eighth-note represents a
complete chord, and unlike the more complex progressions in the preludes
in C major, C minor, D major etc., most of these chords (up to the third
eighth-note in m. 4) appear in root position. Moreover, the secondary
processes that often occur in harmonically-determined pieces can both be
found here: a pedal and sequences. These processes are connected with the
following dynamic lines: Above the tonic pedal, a D-minor cadence
unfolds with the active step to the subdominant taking place at m. 13. The
tension-enhancing step is enhanced by an upward shift in the pitch level. A
complementing descent accompanies the tension-decrease toward the
resolution in m. 2: mm.1-21 = p mp p. After an octave leap that brings
about a sudden tension-increase, a model of four eighth-notes is introduced
and sequenced one step lower. In the model, the pitches and with them the
tension are again falling: mm. 22-31 = mf p + - mp + - p . In the next
measure, the density of sequential pattern and harmonic change doubles.
Each eighth-note now establishes a new chord, and the model to be
sequenced is only two eighth-notes long. (The sequences contain one slight
adaptation in the bass, which allows the left hand to move to the preferred
lower octave.) The model and its three sequences move upward, thus expressing a significant tension-increase: mm. 32-41 = mp mp+ mf mf.
The next measure provides a yet different pattern, with harmonic changes
now occurring on the quarter-note beats while the bass inserts non-harmonic
passing notes on each unaccented eighth-note. The two-eighth-note long
model is followed by three descending sequences. Here, the tension-release
is prolonged for an additional measure with a short cadential close:
mm. 42-62 = mf + p.
D minor
115
The second section is determined by its two motifs. The longer first
motif (M1), introduced as an independent melodic unit in the lower-voice
of mm. 6-8, is actually not completely new. Its stepwise descent recalls the
bass line in m. 4, and its ending is reminiscent of the cadential-bass pattern
in mm. 5-6. Some details, however, are new, and they are decisive: M1
begins on a pitch that is the seventh of its chord. This creates high melodic
tension right from the start of the motif. In its second half the line returns
once more to the same note, which is now redefined as the root of the
deceptive chord in the subsequent cadence. The tension created here is of
a harmonic nature. In the right hand, the uppermost chordal notes present
several short parallels to the left-hand line. Particularly the ascent in the
middle of m. 7 serves to enhance the impact of the deceptive chord. The
much shorter second motif (M2), introduced in m. 10, depicts a melodic
descent combined with a harmonic process of relaxation and thus counterbalances the gesture of M1. The dynamic processes in this section may be
described as follows:
M1
mm. 62-82
mf mp mf + p
sequence
mm. 82-101
mf + mp + pf p +
M2 + sequences mm. 101-122 mf mp + mp p +
M1 + extension mm. 122-151 pf mf pf mp mf + p
The third and concluding section combines various secondary melodic
processes of a different nature. The lower part sets out from the newlyfound keynote D. After a broken chord that propels it almost two octaves
upward, it splits into a hidden two-part structure with D as a pedal in the
lower part and a melodic descent in the upper notes, accompanied by a
peak-note line in the right-hand part. The anticipated release of all tension
is delayed when the peak-note line takes on a virtuoso form in mm. 18-19.
But, as this virtuosity seems to bear little effect against the decreasing
power of the bass line, the previous peak-note pattern is reinstated.
Soon after the bass descent has been concluded, the concealed two-part
structure of the left-hand part turns into a real one: an independent middle
voice evolves. As middle and lower voices proceed, two distinct tritone
intervals mark points of high tension at mm. 213 and 223. In the right-hand
part, the peak note line is taken over alternately by the top and bottom
notes of the broken chords. The overall impression is one of descent, so
that the downbeat of m. 23 is reached in a softness almost comparable to
that of its corresponding measure, m. 15.
From here, both hands are once again propelled up almost two octaves,
with the now firmly established middle voice adding extra drive. In an
116
WTC I/6
unaccompanied cadenza-like treble line the ascent continues up to the high

B, from where diminished chords descend chromatically until the keynote
D is regainedand with it the beginning of a short, chordal-style final
cadence.
WTC I/6 in D minor Fugue

Beginning after an eighth-note rest and ending at m. 31, this subject
spans exactly two measures. Despite the sudden interruption of the sound
flow that is indicated by the wedge on the B and creates a tensionsustaining hiatus, the subject consists of one indivisible phrase. Its pitch
pattern features a predominance of steps and minor thirds. The single
larger leap in the middle of the phrase, a minor sixth, represents one of the
established high-tension intervals. The rhythmic pattern comprises four
different note values: 16th-notes, eighth-notes, quarter-notes and the tied
quarter-note (which, as will be seen later, does not belong entirely to the
subject but is nonetheless heard as a rhythmic unit).
As one compares the rhythm in the remainder of the piece with the
subject, if becomes obvious that there is only one additional note value in
the thematically active parts (i.e., not counting the longer values in the
cadential closes): a syncopated dotted eighth-note. As will become apparent,
this is the value with which the counter-subject implicitly opens.
The subjects harmonic background is that of a simple progression; the
tonic gives way to the subdominant on the second beat of m. 2, followed
on beat 3 by the dominant seventh, and
resolving onto the tonic at the beginning
of m. 3. The fact that the active harmonic
step takes place in the middle of m. 2
and not on one of the downbeats endows
the B with the quality of a syncopation. This impression is enhanced by
the rhythmic standstill (this is the feeling given by the quarter-note that
appears suddenly, after the initial eighth-note motion had developed into
16th-notes) and the interrupting wedge. This B is therefore the obvious
choice for a climax. Melodically reached in a high-tension leap, harmonically representing chord iv, further enhanced both by its suddenly larger
rhythmic value and by the unexpected articulation mark, and metrically
appearing as a quasi-syncopation, this note combines all possible features
that could characterize a climax.
D minor
117
The preparation and the resolution of this climax are slightly irregular.
The process of tension-growth leading up to the climax is called into
question by the three-note slur in m. 2, and after the climax there are only
two notes entrusted with resolving the rather powerful tension. Yet instead
of ending on one of the two melodically relaxed notes of the tonic chord,
F or D, the subject comes to a melodically incomplete-sounding halt on A,
the fifththe one scale degree that regularly fails to convey a satisfactory
feeling of resolution. The slur in the subject becomes more comprehensible
if we remember that articulation symbols in Baroque polyphonic music
derive from markings for string and wind players. Violinists approaching
an unmarked line would play each note with a separate bow movement.
They would thus be able to increase the tension through each note. But,
seeing the slur in the subject, they would combine the group of notes
comprised under it in one single bow movement. The dynamic effect is
that the initial note under the slur is active while the notes following it
sound passive and do not continue the tension. The slur thus creates a
two-leveled structure: the dynamic gesture leads through each of the notes
in the first measure to the first note
under the slur and from there implicitly on to the climactic B, after
which it breaks off and then
resolves through the trill into the
subjects ending note.
There are seventeen complete and seven incomplete but structurally
relevant statements (the latter are here marked with an asterisk):
1. mm. 1- 3 O
9. mm. 17-19 U
17. mm. 28-30 M
2. mm. 3- 5 M
10. mm. 18-20 M
18. mm. 29-31 U
3. mm. 6- 8 U
11. mm. 21-23 U
19. mm. 33-34 M*
4. mm. 8-10 O
12. mm. 22-24 O
20. mm. 34-36 U
5. mm. 12-13 M* 13. mm. 23-25 U
21. mm. 35-36 O*
6. mm. 13-15 O
14. mm. 25-26 M* 22. mm. 35-36 M*
7. mm. 14-16 M
15. mm. 26-27 U* 23. mm. 39-41 U
8. mm. 14-15 U*
16. mm. 27-29 O
24. mm. 40-42 M
Two kinds of changes can be observed in the complete statements of

the subject: Its ending may be varied (melodically to a rising fifth in mm.
22-23, or rhythmically with eighth-notes instead of the two quarter-notes
118
WTC I/6
in mm. 35-36), or it may appear inverted (see entries nos. 5, 7, 12, 13, 14,
15, 16, 18, and 22). Moreover, there are two conspicuous strettos, both
involving complete entries appearing in the original shape (mm. 17-20 and
39-42). These are surpassed by a stretto sounding in a chain of three entries
two of which are inverted (mm. 21-25). Finally, there are several strettos
made up of a mixture of complete (2/2) and incomplete (1/2) statements, using both the original shape (O) and the inversion (I):
mm. 12-16: 1/2 I (M) + 2/2 O (U) + 1/2 O (L) + 2/2 I (M)
mm. 25-31: 1/2 I (M) + 1/2 I (L) + 2/2 I (U) + 2/2 O (M) + 2/2 I (L)
mm. 33-36: 1/2 O (M) + 2/2 O (L var) + 1/2 O (U) + 1/2 I (M)
Parallel statements do not occur. There are, however, two instances
where the (inverted) second entry of a stretto coincides with an incomplete
entry in the original shape: see mm. 14-15 M + L and mm. 35-36 U + M.
Bach invents only one counter-subject for this fugue. CS grows out of
the final note of the preceding subject statement. The fact that the subject
seems to ends in a tied note thus brings up the question of the note value
with which the counter-subject begins. To what proportion the A on the
downbeat of m. 3 is divided between the subject and its counter-subject
can be understood when comparing later counter-subject statements: the
few times when CS does not follow immediately after a subject statement,
it opens with a first note of dotted eighth-note duration (see, e.g., U: m. 6).
The counter-subject thus begins, like the subject, on the second eighth-note
of a measure and ends after two measures on the downbeat. It consists of
two subphrases. These are interrelated through their rhythmic structure.
Both set out with a syncopated dotted eighth-note (U: mm. 3 . 4) and
progress in regular 16th-note motion until the subsequent downbeat. Both
are structurally analogous: in the first subphrase, four descending 16thnotes are sequenced; in the second subphrase, the same happens to a turn
figure. The two subphrases are most clearly distinguished in mm. 6-8
where the first is allocated to the upper voice while the second follows in
the middle voice. Considering these features, each subphrase will find its
climax on the initial syncopation, after which the falling 16th-notes effect
a relaxation. These dynamic developments are very gentle. With regard to
the relationship between the
two subphrases, the second
subphrase expresses a higher
tension level than the first
both because of the shift upward and, more importantly,
D minor
119
because of the greater emphasis the real syncopation creates in comparison to the merely implied syncopation at the beginning of CS.
Determining the number of episodes in this fugue depends on how one
evaluates the incomplete subject statements. If one assumes that, as in
many other fugues, considerably shortened subject statements function as
episode material, then the number of episodes would amount to eight.
However, since the incomplete statements in this prelude often engage in
primary-level strettos with complete entries, the label episode seems illfitted for measures characterized by these shortened entries. The following
table lists all the episodes in this fugue, but places measures with
incomplete subject statements in brackets:
E1 mm. 5-61
[E5 mm. 25-271]
E2 mm. 10-121 [-131]
E6 mm. 31-331 [-341]
E3 mm. 16-171
E7 mm. 36-391
E4 mm. 20-211
E8 mm. 42-44
Almost all episode material in this fugue seems closely related to the
subject. This is particularly true for the first three episodes and their
analogs later in the piece: In E1, the upper voice imitates the subjects
second half (see m. 5) while the middle voice recalls the beginning of
CSincluding the way in which CS emerged out of the subjects ending
note (M: m. 5 . U: m. 3). In E2, both the upper and lower voices extend
the preceding entry in sequences (mm. 9-101 are sequenced in mm. 10-111
and 11-121). Only the middle voice gains some independence by creating
a motivic figure of one-bar length. In mm. 12-131, the first incomplete
entry in the middle voice is combined with the subjects second half in the
lower voice (as it appeared in E1) and a long note in the upper voice. E3 is
a varied sequence of the measure preceding it (mm. 16-171 . 15-161). By
contrast, E4 is a typical cadential close: the upper voice provides an
elaborately ornamented dosido formula, the middle voice contributes the
syncopated trill with resolution on the next downbeat that often appears in
closing formulas, and the lower voice plays a cadential-bass pattern. In E5,
two incomplete subject entries in the middle and lower voices combine
with a figure derived from the end of the counter-subjectextended and in
inversion. The three final episodes are related to three episodes in the first
half of the piece: For E6 see E2, although the voices are inverted and considerably varied (mm. 31-331 or 341 . 10-121 or 131). For E7 see E3, the
latter being an extended version of the former, with three bars similarly
continuing as a sequence of the preceding bar. For E8 see E4, as the first
measure of E8 is an exact transposition of E4 (mm. 42-431 . 20-211). This
120
WTC I/6
is followed by a 1-measure coda in which the three voices of the fugue

are split to six. The four inner voices of this new texture present a fourfold
parallel of the subjects first halfthe right-hand double thirds moving in
inversion, those of the left-hand in the original direction.
Not a single episode displays independent motifs, and none reaches the
level of a self-contained unit. One can, however, distinguish four different
ways in which these episodes behave within the dynamic design: The two
cadential closes (E4 and the first half of E8) create the most noticeable
relaxation. In the case of E8, this relaxation is soon counteracted by a
particularly powerful tension increase, but this falls already into the coda.
The episodes that appear as an extension of the previous development
create a gradual relaxation but never quite lose the color of the primary
material (see E1, E2, and E6). The episodes that extend the preceding
measure with sequences involving all voices (E3 and E7) serve to prolong
the tension of the subject entry; in both cases this impression is further
enhanced by incomplete statements balancing the resolving tendency in the
subjects second half (see L: mm. 15-161 and particularly U: mm. 35-361).
Finally, the episode appearing in brackets in the table above because of its
stretto of incomplete subject statements (E5) has the strongest drive of all.
It sets out from a point of complete relaxation and creates, in both its
incomplete stretto and the inverted (i.e., ascending) counter-subject figure,
a powerful increase preparing the subsequent complete-entry stretto.
Both the pitch pattern and the rhythm advocate a rather calm basic
character. Yet there is one detail that seems to support a character in which
the single notes carry less weight: it is the fact that, upon close inspection,
all 16th-note groups are turn-figures or scale segments ornamenting longer
notes (m. 3 = A-G ornamented). Taking into consideration both the main
constituents of the character and the ornamental structure of the 16th-notes,
the ideal tempo is one that creates a sense of calm within a rather fluent
quarter-note pulse. The corresponding articulation requires legato for all
melodic notes, with slightly less weight in the 16th-notes. Non legato is
reserved for cadential-bass notes (mm. 20 and 42) and obvious brokenchord patterns (as in L: mm. 31-32). The wedged note in the subject should
sound actively interrupted after about half its note value. (For performers
who like to work on shades: such an active interruption stands in contrast
to the passive ending of unmarked notes before a rest or phrase cut.)
The proportion of the preludes tempo to that of the fugue is simple: a
quarter-note in the prelude corresponds with a quarter-note in the fugue.
(Approximate metronome settings: 60 for all beats.)
D minor
121
For the ornament in the subject there are two possible solutions; they
depend on how one interprets the wedge on the climax. For performers
who feel that the wedge creates a sudden, rather dramatic halt in the
melodic flow and that the subsequent G comes in after something like a
phrasing, the trill should begin according to the rules for ornaments at
phrase beginnings, i.e., on the main note. For performers who feel that the
wedge creates articulation rather than phrasing and that the tension is
suspended throughout this interruption and picked up at an almost undiminished level in the G, the trill should begin according to the rules for
ornaments within a phrase, i.e., on the upper auxiliary note. In each case
the trill shakes in 32nd-notes and ends in the whole-tone suffix F-G.
There are several indicators that help to determine the design of this
fugue: A subject entry in reduced ensemble can be found at the beginning
of the four-measure passage with multiple strettos (mm. 212-251). This
reduction of voices occurs after the episode identified as a definite cadence
with strongly conclusive powers. The fugue is thus laid out in extended
structural correspondence. Working backward from the analogous cadences
one finds: mm. 20-211 . 42-431, mm. 17-191 . 39-411, mm. 12-(161) .
33-(381), mm. 9-111 . 30-321. Moreover, the subject statement following
the first distinctly cadencing episode, launching the four-bar phrase with
dual stretto (mm. 212-251), is set in reduced ensemble. If we accept that the
first four single entries in the fugues initial half (mm. 1-10) correspond to
the first two strettos in the latter half, the rest of the analogy falls into
place. The two passages form the first and third sections, concluding in
mm. 131 and 341 respectively. The two passages just ascertained as
analogous, mm. 9-21 and 30-43, constitute the second and fourth sections.
122
WTC I/6
Harmonically, the three initial statements sound in D minor, the fourth

hovers between keys, and the subsequent episode modulates to the
dominant. The second section is anchored in A major/minor. In the fugues
second half the first stretto modulates back to the tonic, followed by statements that once again fluctuate between D minor and major. Interestingly,
the harmonic endings of the first and third sections overlap with the
incomplete first entries of the strettos with which the second and fourth
sections begin. Such an overlapping creates a particularly tight-knit linkage
between two consecutive sections, so that the main impression of the
design is one of two parts, each with two coupled sections.
Within the first section, the tension rises gradually owing to the growing number of voices. The descending sequences in the episode that ends
this section serve to dissipate this tension, so that the second section sets
out from a rather soft level. Its many strettos, however, create a powerful
buildup until immediately before the concluding cadence. Both sections
feature an episode in their middle that, as has been shown above, is so
closely related to the primary material that the loss of tension is negligible.
The third section begins in reduced ensemble on a soft level into which it
falls back in the return modulation (m. 25) after only very little increase
in-between. Yet from here on, all forces join in a stretto that reaches an
intensity considerably above that heard in the corresponding measures of
the first part. The fourth section begins again slightly softer than the third
section had endedbut more intensely than its equivalent in m. 13. After
the initial stretto, the tension-sustaining episode with its ascending
sequences is extended (compare E7 with E3) and thus leads to a final
statement pair that, once again, outranks the corresponding one in the first
half of the piece.
It can thus rightly be said that each of the two halves of the fugue is
composed as one long buildup of tension, interrupted only slightly at the
section ending in its middle. The second half combines several additional
intensifying features and thus appears, in terms of tension, as a heightened
variation of the first.
WTC I/7 in E major Prelude

The prelude in E major is long and complex. Its three segments are
visibly distinct in the use of material and the degree of virtuosity. The first
segment (mm. 1-10) is built along the lines of a prelude determined by
motivic development, ending with a virtuoso passage and a cadential
close. The second segment (mm. 10-25) displays a polyphonic texture in
which voices enter one after the other with a short motif. At first, this motif
is imitated in stretto; later the imitative pattern loses some of its density.
No indications for any of the well-known structural models (such as fugue
or invention) materialize: there is no cadential close in a second key
followed by a new beginning corresponding in any way to that in m. 10.
As a passage in free imitative style based on a single motif, this segment
may thus be referred to as a fugato. The third segment (mm. 25-70), by
far its longest, is also polyphonic, beginning with two contrapuntal voices
in complementary keys. As there are several definite cadences and new
presentations of the material after these cadences, this segment within the
E-major prelude qualifies as a fugue. To sum up, Bachs prelude in E
major is conceived as a through-composed three-movement composition
consisting of prelude, fugato, and fugue.
The first cadence concludes at m. 43. However, as the bass has not yet
begun to participate in the harmonic progression but remains locked in a
tonic pedal, this cadential close should not be regarded as structural. A
second reason is that the flow of the lines continues uninterrupted, without
even the slightest sign of phrasing. The next harmonic progression ends
again over a pedal noteat m. 101 in B major, the dominant of E. It
marks the end of the prelude in the prelude.
Within the fugato, similar circumstances prevail as in the prelude.
The only structurally relevant cadence, however, appears again at the end
of this segment, i.e., at m. 251 (once again in B major). Two earlier
cadential closes are worth mentioning because of the unusual way in which
the listeners expectations for a resolution are deceived in the first and
turned around in the second. Both cadences set out to close in the home
key of E major but are prevented from doing so. The first seems to draw
to a close in m. 15. Yet before the typical closing formula has reached the
resolution, three of the four voices proceed quite differently from what is
123
124
WTC I/7
anticipated.1 The second cadence is initiated in m. 18. The return to the

tonic seems almost achieved on m. 191, with only a soprano appoggiatura
lagging behind with its resolution upward to E. Yet when the uppermost
voice moves, it does not rise but fall. Soon thereafter, the other voices also
give up hope for a conclusion. After this second attempt at a resolution
onto E major, the section returns to the dominant tonality from which it
had set out.
Neither the prelude nor the fugato thus contain any structurally
relevant internal cadences. The fugue, however, is variously subdivided.
Its six sections bring the total of structural blocks in the E-major prelude
to eight, each of them individual and without any analogy. In the table
below, the tonality in brackets denotes with which cadential target each
section closes.
1. mm. 1-10 =
prelude
(B major)
2. mm. 10-25 =
fugato
(B major)
3. mm. 25-35 =
fugue, section I
(G minor)
4. mm. 35-41 =
fugue, section II
(C minor)
5. mm. 41-49 =
fugue, section III (B major)
6. mm. 49-58 =
fugue, section IV (A major)
7. mm. 58-68 =
fugue, section V
(E major)
8. mm. 68-70 =
fugue, coda
(E major)
The ideal tempo for this prelude is one that accommodates the character of each of the three movements without any need for an adjustment
of pace. To be specific, the tempo should allow the 32nd-notes of mm. 8-9
to sound fluent enough to give the impression of an ornamental rather than
a melodic line. At the same time, the quarter-notes of mm. 10-24 should
sound calm enough to express the basic character of this fugato.
The appropriate articulation for the entire piece is legato. While the
lively character in the prelude would require any eighth-notes and longer
values to be played non legato, there are no notes to which this applies:
virtually all longer values come with tie-prolongations, and the only
separate long notes (tenor mm. 7-10) form a do-si-do group that is legato
in any case. In the calm character of both the fugato and the fugue,
only cadential-bass notes and consecutive leaps are detached. This applies
to mm. 29-30: B-B-E , mm. 34-35: C-D-G, mm. 39-40: G-C-F, mm.
40-41: E A F G, and mm. 58-59: E A F B .
1
Try playing m. 15 as the penultimate measure of an E-major cadence. The soprano would
resolve onto E, the alto descend to G, the tenor to E, and the bass leap down to E. As m.
161 shows, this is not what happens: all voices except for the alto take unexpected turns.
E major
125
The only ornament in this prelude occurs in m. 9. It is a trill ending in

an anticipated resolution, i.e., in a point darrt. The motion should be
faster thanpreferably twice as fast asthe fastest regular note values in
the piece.2 The trill itself thus sets out in sixty-fourth-notes from the upper
auxiliary F, comes to a sudden halt (arrt)
on its main note and ends with the anticipated resolution. The following figure gives
a written-out version for the entire compound ornament.
The first segment of the composition, the prelude within the prelude,
is built entirely on an eighth-note motif. M1 is introduced in the upper
voice where it begins after a downbeat rest, describes a curve in 16th-notes,
and ends with a large leap upwards. This leap is written as a split in the
voice, thus emphasizing the interval it spans. Moreover, the impact of the
target note is enhanced by its duration, which exceeds that of the entire
preceding 16th-note group. M1 is immediately imitated but not copied in
the tenor of this four-part texture: the leap is much smaller and there is no
voice splitting. Instead, a pedal note E materializes in the bass. The
subsequent two measures feature ascending sequences of the pattern
established in m. 1 together with a corresponding tension-increase. After a
short resolution onto the home chord the right hand propels M1 up to the
peak C. At the same time, the left-hand part starts developing the motif: the
leap appears substituted by a note that is integrated into the 16th-note
figure both in pitch and in note value, and the prolongation is given up.
The result is a continuous run downward covering two octaves. In m. 6, a
descending scale in the right hand joins the left, and both voices reach a C
minor chordstep vi in E major, thus a chord with comparably high
tensionat m. 71. Two further statements of M1 are followed in the treble
by a continuous development similar to that previously heard in the lower
part (compare m. 8 right hand from C onward with mm. 4-6 left hand). The
prelude section ends with a freely virtuoso run and the above-mentioned
cadence in mm. 9-10.
The second segment, the fugato within the prelude, is also based on
a single motif. M2 originally spans five notes (mm. 10-11: B E D C B )
but soon drops its final resolution. In its four-note version it is strikingly
related to the main motif of the Prelude in B minor, also from Book I of
2
It need not, however, exceed the tempo of the spelled-out turn figure immediately preceding the trill, as closer inspection reveals that this figure is in fact part of a longer compound
ornament beginning on the tied eighth-note E.
126
WTC I/7
Bachs Well-Tempered Clavier. As M2 is the one and only driving force of

the fugato, it seems important to identify all its statements:
1. m. 10
T
6. mm. 13-14 S**
2. mm. 10-11 B
7. mm. 14-15 T**
3. mm. 11-12 A*
8. mm. 16-17 B***
4. mm. 12-13 S***
9. mm. 17-18 S*
5. mm. 12-13 B**
10. mm. 19-21 S**3
While the four-note version is still very close to the original, later
modifications of M2 appear more and more transformed.4
The final segment of the fugato features only two rudimentary statements of M2 (see S and B: mm. 19-21). More influential here are the
descending lines in all voices: soprano mm. 20-25 = E down to D, alto
mm. 20-25 = D down to B, tenor mm. 22-24 = B down to E, and bass
mm. 20-23 = G down to A. Each of these descending lines comes in
diminuendo, so that the last five measures of this section describe a long
relaxation. Not only is the ending of the fugato section thus well prepared,
but, more importantly, the impending beginning of the third major
segment, the fugue, is strongly suggested.
This fugue turns out to be based on two leading ideas. As each of them
is later used quite independently, it seems appropriate to speak of two
subjects rather than of subject and counter-subject. In a proper double
fugue, these would be presented separately at considerable distance, each
determining a major portion of the fugue before being juxtaposed. Yet in
this fugue within a prelude, the two melodic ideas enter simultaneously.
S1 is introduced in the alto. It spans two measures, setting in on the second
16th-note in m. 25 and ending on the downbeat of m. 27. This subject is
closely related to M1 from which it borrows the first seven notes. S2 first
appears in the bass: it stretches from m. 251 to m. 271. This subject is
closely related to M2. The beginning recalls the first three notes of M2;
these are followed, after the third note has doubled as the beginning of a
sequence, by the entire four-note version of M2.
3
A missing accidental in the Urtext, found also in several other editions, should be carefully
considered: in m. 20, the tenor features B-A-B despite the A in both soprano (m. 19) and
alto (m. 20). As the soprano statement of M2 suggests B major for these measures, adding
a natural to the leading note seems essential.
4
The variations indicated by the asterisks above are: * The third note may be prolonged to
twice its value, delaying the fourth note to a weak beat where it follows usually as yet
another syncopation. ** The motif may sound with its first note lengthened to a syncopation. This may then be followed either by only the third note or by the two remaining notes.
*** The fourth note may bend back instead of leading downward.
E major
127
The dynamic design of the two subjects is, at least at the beginning,
determined by the motifs from which they are derived: S1 begins with a
tension-increase toward what was the final note of M1; the second, longer
subphrase begins anew and carries an even stronger crescendo through the
motifs varied sequence. The climax falls preferably on the quarter-note E
and is followed by the resolution through the subsequent do-si-do figure.
The dynamic shape of S2 reflects its evolution from the fugato motif: a
climax on the syncopation E
followed, after slight relaxation serving as a new start, by
a stronger second climax on
m. 21. The descent to the keynote of the target chord brings
the relaxation.
S1 appears twelve times in the course of the fugue and S2 twentytwo times. The following chart lists the statements of both subjects:
mm. 25-27
S1
A + S2
B
mm. 27-29
S1
S + S2
T
mm. 30-32
S1
B + S2
A
mm. 35-37
S2
A + S2 S
mm. 38-40
S2
S + S2 T
mm. 41-44
S2
A + S2 T + S1 S
mm. 46-49
S2
B + S2 T + S1 A
mm. 49-51
S1
A + S2
B
mm. 53-56
S1
S/B + S2
T + S2 A
mm. 56-58
S1
B + S2
A + S1 S
mm. 58-60
S1
A + S2
S
mm. 60-63
S1
S v S2
A + S2 B
mm. 64-67
S2
A + S2 S + S2 A
mm. 68-70
S1
A + S2
T
Both subjects appear with several irregularities and variations: S1 may
come without its final note (as in mm. 27-29) or even without the last two
notes (as in mm. 60 and 63). It may be deprived of its entire first subphrase
128
WTC I/7
(as in mm. 30-32), shortened in the middle so that the first subphrase leads
directly into the note repetition of the ending (as in B: mm. 56-57), or it
may, while shortened in the middle as described above, find its concluding
note repetition replaced by a single tied note (as in S: mm. 56-57). S2 may
come without its final note (as in mm. 27-29), it may have lost both its
final notes (as in mm. 30-32, 35-37, 38-40, etc.), or its beginning may
sound ornamented (as in m. 60).
S2 builds several strettos (see particularly mm. 35-49 and 64-67), the
last of which sounds briefly like a parallelalthough it is not. In addition,
there are a few interesting voice crossings: in m. 28 the tenor does not
descend directly after the two rising fourths of the S2 beginning. Instead it
M1
M1
M1
M1
M1
M1
M1
M2
M1
M1
M1
M1
M1
M2
M2 expanded
M2 expanded
M2
M2
M2
M2
M2
M1
M1
M1
M1
M1
M1
M1
M1
M1 S1
S1
S2
S2
M1
S2
M1
M1
S2
M1
S2
M1
M1
M1
M1
M1
S1
M1
S2
M1
S2
S2
S1
S2
S2
S1
M1
S2
S1
M1
S2
M1
S2
M1
M1
M1
S1
S2
S2
M1
M1
S2
M1
M1
M1
M1
M1
S1
M1
S2
M1
M1
S1
M1
S1
S2
M1
M1
S1
S2
S1
M1
S2
M1
M1
M2
S2
S1
S2
M1
M1
M1
E major
129
continues, using scattered M1 quotations, to climb as high as A (m. 29).

The alto with its statement of S2 thus begins underneath the tenor and only
regains its rightful position in m. 31. In mm. 41-43 the S2 entries in alto
and tenor rise high while the S1 entry in the soprano crosses below both of
them before it corrects its position with an octave adjustment. Finally, in
mm. 53-54 the S1 statement seems confused: it begins in the soprano
with a varied first subphrase, descends correctly to E (m. 541), after which
it switches to the bass! This, however, does not pick up the right pitch:
instead of the expected A-A-A-G-A we hear F-F-F-E-Ab.
As in any other fugue there are a number of subject-free passages:
E1 m. 29
E4 mm. 40-41
E7 m. 63
E2 mm. 32-34
E5 mm. 44-45
E8 m. 67
E3 mm. 37-38
E6 mm. 51-52
The material employed in these episodes derives mostly from M1; the
remainder is neutral. Features to be pointed out include typical cadential
patterns (see B: at the end of E2 and E8, and S: at the end of E3 and E4),
as well as particular pitch patterns that engineer a tension-rise in all voices,
thus preparing for the subsequent entries (as in E5 and E6). In contrast to
these six episodes with definite closures, E1 and E7 use fragments of the
first subject, thus serving merely as links.
Many indicators help in presenting an easy overview of the design of
this fugue: The four-part ensemble is built up gradually, from the initial
two voices (alto + bass) in m. 25, which are joined by the soprano in m. 26
and the bass in m. 30, to a first four-part cadence in mm. 34-35. A tonic
pedal in mm. 68-70 separates the coda from the fugues trunk. The
grouping of subject statements includes simple S1/S2 juxtapositions in
mm. 25-35, S2 strettos in mm. 35-41, S2 strettos with additional S1 statements in mm. 41-49, S1/S2 juxtapositions with irregular S1 entries in mm.
49-58, and merging entries / juxtaposition / S2 stretto in mm. 58-68.
WTC I/7 in E major Fugue

The subject of this fugue spans 1 measures. It begins on the initial
downbeat and ends at m. 23 on B. As indicated by the natural sign preceding the A at the end of m. 1, the subject features a modulation to the
key with only two flats: B major, the dominant of E major. A is not only
the leading-note to the new key but also part of a broken chord F-A-C-E
in mm. 1-2, the V7 chord of B major. The subject note B in the middle of
m. 2 represents the resolution onto the new tonic.
130
WTC I/7
There are two subphrases within this subject. This can be deduced not
only from the rest in its middle but much more from the varied partial
sequence with which the second subphrase begins: the two prominent
eighth-notes in the subjects second half sound like an elevated sequence
of the two eighth-notes before the rest, while the preceding groups of
16th-notes is shortened but made up for by the final strong-beat note.
The pitch pattern in the subject features mainly broken chords:
m. 11-2
= E-major chord (with auxiliary note)
m. 12-3
= A-major chord (with auxiliary note)
mm.14-22 = F7 chord (with passing note5)
A brief look at the remainder of the fugue shows that these and other
broken-chord patterns are prevalent throughout the piece. The rhythm in
the subject consists mainly of 16th-notes and eighth-notes, with the
exception of the quarter-note carrying the trill. The same predominance of
these note values can be detected in the entire piece.
The subjects harmonic background is determined by the modulation
within its confines. The active step (tonic to subdominant) occurs melodically immediately after the second beat, but is metrically confirmed only
on beat 3. This subdominant harmony is sustained even when the second
eighth-note of beat 3 returns to B (a note that, in its unaccompanied
melodic version, is necessarily heard
as a return to the tonic). The pivot
chord (V7/V) beginning melodically
with the A can be felt metrically
either in the rest at m. 14 or at m.21.
The E at m. 21 seems the obvious choice for a climax: it falls on a
downbeat, it represents the metrical place of the pivot chord that triggers
the modulation, and it is reached in the high-tension interval leap of a
minor seventh. Moreover, as mentioned above, it is conceived in varied
sequence to the process in the first subphrase that finds its climax on the C
at m. 13, the note that represents the subdominant harmony and sounds
rhythmically as the first halt after the opening 16th-notes.
The tension in the subject thus moves in two curves. The first begins
with a moderate crescendo and is complemented with a one-note relaxation
(C-B). A second crescendo then surpasses the first and is complemented
with a longer diminuendo for a more perfect resolution of the tension.
The subject appears altogether nine times in this fugue:
5
The use of the second eighth-note in m. 2 as a passing note is confirmed in Bachs harmonization of later subject statements; see, e.g., mm. 4, 7, and 18.
E major
1. mm. 1- 2
2. mm. 3- 4
3. mm. 6- 7
131
U
M
L
4. mm. 10-12 U
5. mm. 17-19 M
6. mm. 20-22 L
7. mm. 25-27 L
8. mm. 28-30 U
9. mm. 33-35 M
The subject always sounds in its complete length. Its last note appears
most often as a 16th-note but may be extended to an eighth-note (as in mm.
27 and 30) or even to a quarter-note (as in m. 35). Moreover, several
statements begin with a syncopated anticipation of the first note (see mm.
10-11, 25-26, 28-29, and 33-34). In addition to these small changes in the
appearance of the subject, interval adjustmentsbetween the first two
notes and across the restoccur in all tonal answers (see m. 3 etc.). No
stretto or parallel are used.
Bach has invented only one counter-subject for this fugue. CS is introduced in mm. 3-4 against the second subject statement, where it begins
slightly later than the subject with the eighth-note A and ends, together
with the subject, on the G at m. 43. This counter-subject appears as a
faithful companion to the subject in all but the initial and final entries,
featuring only one slight variation of its beginning in m. 20. It fulfills its
task of counter-balancing the subject in several ways: Against the brokenchord patterns dominating the first two-thirds of the subject, CS sets
stepwise motion, and against the stepwise motion at the end of the subject
it sets a broken dominant-seventh chord (see m. 42). While the subject is
made up of two subphrases, the counter-subject is conceived as one
indivisible line. And while the subject has two climaxesa softer one on
the third beat and a stronger one on the fifth beat after its beginningthe
counter-subject features a single
unbroken tension-curve. Its
climax falls either on the first
eighth-note, followed in this case
by a single long diminuendo, or

(more likely) on the syncopation
that coincides with the rest in the
subject.
In this fugue, the number of episodes equals that of the subject
statements:
E1 = mm. 2-3
E4 = mm. 12-17
E7 = mm. 27-28
E2 = mm. 4-5
E5 = mm. 19-20
E8 = mm. 30-33
E3 = mm. 7-10
E6 = mm. 22-25
E9 = mm. 35-37
132
WTC I/7
E6 is the only episode to show a relationship to the subject: its first two
measures quote a variation of the first subphrase in their upper voice. All
other episodes are entirely independent from the primary material. They
present a number of characteristic motifs that are used with great consistency: E1 introduces the half-measure motif M1 featuring two broken
chords, each with a subsequent step downward to the next beat (mm. 2-3:
F-D-B-A, F-D-A-G). This motif plays a leading role in all episodes of
this fugue, occasionally in an extended version with a final unaccented
jump upward as in E3 and E4. E2 combines a prolonged M1 in the middle
voice with M2, a figure in the upper voice that also consists of broken
chords, this time in ascending direction. Significant features are the long
syncopations. M2 also recurs, sometimes unchanged, at other times represented only by the accented upward leap in eighth-note rhythm. E3, the
first episode in three-part texture, introduces M3 (M: mm. 7-8: F-D-B-E ),
a third motif that is subtly related to the other two: like M1 it begins with
a descending broken chord, and like M2 it ends with an accented upward
leap. Yet as its rhythm is built exclusively from eighth-notes, it is at the
same time quite distinct. These three motifs appear in a variety of combinations (see particularly in E3, E4, E5, and E8). The second half of E6
then introduces a new pattern that recurs shortly afterward in E7. In it,
sequences of ascending eighth-notes in the bass and a 16th-note dialogue
with varied segments of M1 in the two higher voices create a pattern that
we shall call M4.
Several relationships exist among the episodes of this fugue: the shape
of E5 and E2 is similar although the hands are inverted (mm. 4-5 . 19-20),
E3, E4a, and E8 are analogous except for the inversion of voices (mm.
7-10 . 12-15 . 30-33), and the second half of E6 is taken up, as was
mentioned above, in E7. The only episode segments to remain without any
correspondence are the cadential close of E4b with its preparation in mm.
15-173), the first half of E6 (E6a = mm. 22-241), and the final episode E9.
As all episodes apart from E6a are conceived as independent in
material and character from the subject and counter-subject, they should
sound like self-contained units, in a color and intensity distinctly different
from those characterizing the subject and its counter-subject. The dynamic
gestures within this contrasting and much lighter color are as follows: In
E1 the descending sequences create a natural decline of tension. Similarly,
the final episode E9 sounds in manifold falling lines. E2 with its analog in
E5 and E3 with its analogs in E4 and E8 all show relaxing tendency in
descending sequences. The first half of E6 also features a pitch direction
that points downward; here again the dynamic tendency is decreasing. The
E major
133
second half of E6 and, correspondingly, E7 are the only episode segments

to engender dynamic build-ups. These are created both by the ascending
eighth-note lines in the bass and by the ascending sequences.
Both the steady rhythmic pattern with its predominance of 16th- and
eighth-note values and the pitch pattern with its high content of broken
chords characterize this fugue as rather lively. The tempo should be fast; a
conductor beating an energetic four-four meter may give a good orientation for the minimum tempo, with the tricky trills in mm. 21, 27, and 35
marking the upper limit. The articulation corresponding with the character
of this fugue requires a bouncing non legato for the eighth-notes and a
quasi legato bordering on leggiero for the 16th-notes. The only longer note
values that must be played legato appear in a dosido formula (U: m. 35
E-D-E ). Furthermore, legato is harmonically indicated in the two-part
version of M1 in U: mm. 27-28 (D-C and E-D) and in the chromatic
descent in the split-off upper part of the middle voice in the final measure
(D-C-C-B). The relative tempo of the E major prelude to the subsequent
fugue should best be kept simple: a quarter-note in the prelude becomes a
quarter-note in the fugue. (Approximate metronome settings: 84 for all
beats.)
The only ornament in this fugue is the trill in the subject. As it is
approached stepwise, it begins on the main note with a 16th-note, proceeds
in two pairs of 32nd-notes, and ends in a suffix. This trill must be retained
in every subject statement, even where it is not at all or only ambiguously
indicated in the score, as in m. U: 12, L: m. 27, and M: m. 35.6
The design of the fugue reveals itself from several features, all of
which are underlined by the harmonic development. The build-up of the
three-part ensemble is followed by a fourth statement that is to be considered as a redundant entry belonging to the first section since it remains in
the home key. The subsequent episode modulates to the relative key and
confirms it with a cadential close. The fifth statement sounds in reduced
ensemble and is linked to the sixth by way of the shared minor mode. The
next closure, bringing about a modulation back to the home key, appears in
E6a. The second half of E6 with its rising tendency not only prepares the
next statement but also recurs in variation immediately afterward, when it
6
Note that the trills in mm. 7, 12, 21, and 27 require a double note to be played in the
middle of the trill motion. This becomes much easier in m. 30 if the twofold middle-voice
E is taken by the right handwhich is more comfortable than one might think. In m. 35,
the trills main portion can be played with the right hand while the suffix sounds smoother if
taken by the left.
134
WTC I/7
acts as a bridge to the subsequent subject entry. The remaining group of

subject entry + episode + subject entry is built in obvious symmetry to the
similar group in the fugues first section, encompassing the third entry +
the third episode + the redundant entry. This symmetry hints at a structural
analogy between the second to fourth entries of section I and the same
number of entries in section III. The final entry with its chromaticism and
deceptive cadence is harmonically the most daring.
Within the first and second sections, the tension-increase engendered

by the gradually growing ensemble appears constantly interrupted by the
regularly interspersed episodes of contrasting color. Thus each subsequent
entry sounds only slightly louder than the previous one. In the third section,
the preparation of entry no.7 creates more tension at the very beginning
(thus making up for the missing first statement in the recapitulation?).
This process is repeated before the next statement, so that the tensionincrease within this section is more pronounced than that in the exposition.
As the initial subject entry of the fugue expresses considerable strength
owing due to the subjects lively and bouncing character, the dynamic
equivalent to this tension-increase might be approximately mfpoco f.
Within the second section, the change of mode results in a considerably
less exuberant mood and, consequently, in a much softer touch. The third
section thus not only balances the first one but even exceeds it slightly.
WTC I/8 in E minor Prelude

The E-minor prelude is primarily determined by melodic processes.
These unfold in the context of an essentially homophonic accompaniment.
The two predominant melodic ideas are introduced in the treble line of
mm. 1-4 and 20-21 respectively. The former begins in the second half of
m. 1 and ends on the E in m. 4; the latter stretches from the high C in
m. 20 to the downbeat of m. 21. While subsequent developments of these
melodic ideas show a large degree of variation in the treatment of pitch
progressions, their rhythmic patterns remain consistent.
The first harmonic progression, a simple cadence with the subdominant
in m. 2 and vii7 (replacing V) in m. 3, concludes in m. 4. This cadential
close coincides with the end of the melodic entity and is thus of structural
relevance. However, as the subsequent development is composed as a
continuation with its own structure, phrasing at this point only marks a
smaller breathing within a larger unit. The next harmonic development
embarks on a progression of secondary dominants which each resolve into
their related tonics. Listeners thus repeatedly experience resolutions
without truly arriving at any new key.
mm. 5
6
7
8
9
10
11
12
V7/vi vi
V7/iv iv
V9 I
V9/v v
The following twofold cadence in B minor in mm. 13-142-161 secures
the modulation to the minor dominant. Yet again, this new key is not
established with a truly convincing resolution. In m. 14 the new tonic is
reached in metrically unaccented position on the second half-note, and in
m. 16 the melodic ascent after the downbeat also creates a sense of
openness. The subsequent measures then modulate to the subdominant A
minor, which is established with a iv-V-i progression in mm. 19-20. The
point of completion of this cadence marks the end of the first larger section
of this prelude.
There are altogether four structural sections in the E-minor prelude:
I
mm. 1-4
(E minor) tonic confirmed
II
mm. 4-16 (B minor) modulation to the dominant
III
mm. 16-20 (A minor) modulation to the subdominant
IV
mm. 20-25 (E minor) modulation back to the tonic
V
mm. 25-40 (E minor) tonic confirmed
135
136
WTC I/8
The fifth section shows three subdivisionsnot only on the level of

melodic development but also on that of harmonic progressionsinsofar
as there are two deceptive cadences prior to the final perfect ending: In
mm. 25-29, the Neapolitan-sixth chord in m. 26 and the vii7 in m. 27 seem
to prepare a definite ending. This expectation is deceived on m. 291 by a
chord that combines features of the traditional deceptive-cadence VI
(which would read C E G) with the A of step iv, reached by way of an
appoggiatura B. The i6 of E major that follows in m. 30 is reached
plagally from iv6 and thus does not conclude this cadential progression
either. Similarly in mm. 31-37, the subdominant chord at m. 311 is redefined on beat 2 as yet another Neapolitan-sixth chord, the vii7 is extended
through four measures (mm. 32-35), and the preparation of the ending is
almost exactly the same as before (mm. 36 second half . 28 second half).
Once again an expectation is deceived, this time because the resolving
chord is topped by its minor seventh. As a V7/iv progression, this initiates
another iv-vii7-i progression that, above a four-measure tonic pedal, finally
leads to the resolution into the Picardy-third tonic.
The appropriate tempo in this prelude is determined by rhythm and
meter. On the one hand, the 16th-notes have to be calm enough to retain
their melodic expressiveness and at the same time to allow for 32nd-notes
in the trills. On the other hand, the time signature demands a tempo fast
enough to convey a pulse in half-notes.1
The articulation requires legato for all notes pertaining to the melodic
process; this can be achieved with the fingers alone and thus poses no
problems. The accompanying chords also sound best if smoothly linked.
Yet connection between chordsparticularly if they involve so many
repetitionsis not possible on the piano without the help of the sustaining
pedal.2
1
This is very important, particularly since the somewhat complex rhythmic structure in the
piece may tempt performers into a transitional stage of counting in quarter-notes. While
this may be essential during the process of preparation, it is equally vital that this stage be
ultimately overcome in favor of the larger swinging pulse with only three beats to a
measure.
2
The use of the pedal is a very delicate matter and may appear somewhat complicated
where 16th-note motion in one of the lines requires a clean and unblurred rendition. The
only solution is thus to pedal on or slightly after the last rhythmic value in a measure. As
every performer will aim at evenness in tone color throughout each measure and each
phrase, this means that the pedal cannot be depressed before the last 16th-note fraction of
each half-note. The only stretches that might do without any pedal are the beginning of the
second half (mm. 20-21) and the cadenza-like measure near the end (m. 35).
E minor/D minor
137
There are three kinds of ornaments: trills, arpeggios, and a grace-note.

The grace-note in m. 36 poses the easiest task. In keeping with the Baroque
rule of equal appoggiatura-resolution distribution in binary note values,
the A, played together with the notes in the other voices, is held for one
quarter-note, leaving the second quarter-note for the resolution G.
Although the trills are indicated by different symbols, and tr,
their execution is determined much more decisively by their context.3 The
desirable rendition can be found by addressing the nature of the resolution.
If the ornamented note resolves directly into a strong beat, the trill fills the
entire value, ending with a suffix immediately before the resolution. If,
however, the ornaments resolution occurs on an unaccented fraction of a
beat, or if it is preceded by written-out notes, the shakes obviously have to
stop before the written-out bridge. Also, if the ornamented note belongs to
the harmony of the subsequent pitch, which therefore does not provide an
actual resolution, the trill, short or even relatively long, must end on the
main note without a suffix. Since we know that Bach often wrote simple
mordent symbols for any length of trill including suffix-ended long ones,
these considerations take care of all ornaments. Regarding the onset, as
always in this era, a beginning from the main note is required only if the
ornament is approached in stepwise motion; in all other cases, it is
launched from the upper auxiliary.
The many arpeggios all begin with the lowest note on the beat. They
sound best if the rhythmic distance between the lowest and the highest note
remains constant, even though this may seem more difficult than retaining
the rhythmic distance between any two consecutive notes. As a result, the
long arpeggio at m. 61 should be considerably faster than the short threenote arpeggios at m. 62 and 63, etc. Another question is whether all notes of
a broken chord belong to the level of accompaniment, or whether the treble
or the bass form part of a melodic progression. Whenever the latter is the
case, performers must consider both the coloring (melodious vs. neutral)
and, in the event of a treble line, the metric placement of the melodic note.
The horizontal context would sound awkwardly distorted if such a note
appeared offbeatas it would if it were played as the end of a rolled
chord. One creative way of finding out what is appropriate in any given
context is to imagine the entire texture of this prelude rewritten for an
3
On the one hand, if the extended mordent in m. 4 is interpreted as a note-filling trill,

consistency requires that we play an ornament of similar duration in mm. 8, 10, and 12
where only the simple mordent sign appears. On the other hand, the tr in m. 15 decorates
a dotted-note figure and must therefore be executed with a point darrt ending.
138
WTC I/8
ensemble of instrumentse.g., the oboe for melodic lines in the higher

register, the bassoon for melodic lines in the lower register, and the harp
for the accompanying chords. This aural image will greatly improve the
color pattern of keyboard performers. Questions like the following should
then be asked: Who plays
the bass-clef B at m. 51 as opposed to that at beat 2?
the treble-clef E at m. 61 as opposed to that at beat 2?
the treble-clef C-B-A at m. 81 and the notes at m. 82?
the treble-clef G-F-E at m. 41 and the high G at m. 42?
the treble-clef F at m. 122 and the G at m. 131?
the treble-clef G at m. 251 and the F at m. 261?
the treble-clef E-D in m. 28? (This measure might allow for two
equally meaningful solutions.)
A second question worth considering is whether the accompanying
chords following exactly the same pattern as the others but appearing without an arpeggio symbol should be treated in the same way. This applies to
the left-hand chords in mm. 29-30, the two-hand chord in m. 311 (with the
exception of the C continuing the oboe line), the right-hand chords in
mm. 32-34, possibly the chords in m. 362 + 3 (with the exception of the
melodic treble notes to be played on the beat, together with the root of the
arpeggio), and possibly the left-hand double notes in mm. 38-39.
As was already mentioned, the prelude contains two relevant motifs,
both of which undergo considerable changes in the course of the piece.
Furthermore, both motifs are determined much more by their rhythmic and
metric shapes than by their pitch patterns or the number of notes they
encompass. M1 rests on a figure that could be described as a question +
answer phrase with
two similar halves, each
consisting of an upbeat
of varying length, a
dotted-note group with
or without an ornament,
and a downbeat with or
without a female tail.
M1 is introduced as an extended four-measure phrase consisting of
three subphrases. (The upward stem of the B at m. 11 is confusing, yet
later statements leave no doubt about the motifs upbeat beginning.)
Dynamically, the first and second subphrases describe small crescendos
toward their successive climaxes and preferably no heavy accent in the
third subphrase. The motifs character is graceful and mild. This can be
E minor/D minor
139
heard particularly well after the completion of the first perfect cadence: the
first subphrase sounds in the lower melodic voice, leading harmonically to
a dominant-seventh chord, while its answer, given in the upper voice,
supplies the respective resolution. In mm. 12-14 the motif is freely developed in the upper voice, followed by a variant in what appears as a middle
voice (mm. 14-16). A last statement, presented in the lower voice in m. 16,
is extended and merges into the cadential-bass pattern that ends this
section. Yet another variant emerges in mm. 22-25. Finally, toward the end
of the prelude, M1 recurs stripped of its question + answer design in the
form of a chain and in different harmonic guise.
M2 is introduced in mm. 20-21, in the very center of the 40-measure
prelude. Spanning only one measure, it shares its rhythmic and metric
features with M1. After only one beat, the lower voice sets in, creating a
stretto. The whole imitative pattern is then sequenced. The tension curve is
quite different from that in M1, mainly because the entire figure expresses
a single harmony. The climax falls on the beginning, the character is
stately. M2 is developed once, in mm. 26-28, where it appears in a singlevoiced version without imitation. Owing to the harmonic change at the end
of the first of these statements and the inverted pitch pattern in the second,
the dynamic curve is also inverted, with a crescendo to the final note of the
motif.
The interpretation of this prelude is determined by its broadly swaying
basic character and its texture: a melodic dialogue before the backdrop of
softly arpeggiated chords. Large-scale dramatic buildups are not the aim in
this meditative composition.
WTC I/8 in D minor Fugue

With a little less than three measures, the phrase length of this subject
is not metrically oriented. The keynote D that falls on the middle beat of
m. 3 does not only act as a harmonic resolution of the dominant-seventh
chord represented by the G and E on the first two accented beats of m. 3,
it also marks the melodic return to the pitch from which the phrase set out.
While intuition tells us that the subject is made up of two subphrases, it
seems that there could be two equally valid solutions for phrasing. One
option is to view the complete curve from the initial D to the D in the
middle of m. 2 as an entity and thus as the main thought, with the scalar
descent from G onward as an afterthought. Another option is to regard
the leap D-G in m. 2 as a varied sequence of the opening D-A which
140
WTC I/8
makes sense rhythmically and metrically as each D is a quarter-note on a

strong beat leaping up to a syncopation of 3/8 length. Most performers
have fairly strong feelings as to which version they prefer. However, as it
turns out in the further course of the fugue, the choice is not really ours for
there are two statements that reveal what phrasing Bach himself had in
mind. The middle-voice entry in mm. 19-22 and the upper-voice entry in
mm. 20-22 both feature a rest within the phrase. This rest occurs after the
end of the curve; it cuts through the assumed sequential leap and thus
provides undeniable evidence for the phrasing into main thought and
afterthought, i.e., the first option described above.
The subjects pitch pattern contains two fifth intervals, at both ends of
the main thought; all else is stepwise motion. (As we know now that we
have understood the phrase structure, the step from D to G does not count
as an interval since the two notes do not belong to the same subphrase.)
Rhythmic features include eighth-notes, quarter-notes, and syncopated
dotted (or tied) quarter-notes. The harmonic background is that of a simple
cadence with an extended first chord. The complete curve of the main
through is rooted in the tonic, but interspersed with dominant chords that
quickly resolve back into D minor.
The syncopation at the beginning of
the afterthought represents the
subdominant, followed by a gradual
return to the home chord.
The shaping of tension within the subject leaves no room for doubt in
the second subphrase where a simple decrease after the syncopation is
certainly the only logical answer. In the first subphrase, however, it allows
for two slightly different interpretations, depending on the chosen tempo
and on individual preference for rhythmic versus melodic processes. Performers who feel that the rhythmic feature should take precedence over the
melodic one will place the main climax on the second note, the dotted A,
while performers who favor the pitch curve may wish to play a tension-rise
past this A to B, the (high-tension) minor-sixth degree, thus giving this
subject a very special emotional quality.
The D-minor fugue contains 35 subject statements: in this fugue, the
subject seems all-important. There are no rivaling counter-subjects or even
any transitorily prominent motifs. Instead, subject entries encompass an
unusually large portion of the composition, and the subject appears in so
many different shapes, distributed with such obvious determination and
purpose, that nothing else seems to matter.
E minor/D minor
1. mm. 1-3
2. mm. 3-6
3. mm. 8-10
4. mm. 12-14
5. mm. 19-22
6. mm. 20-22
7. mm. 24-26
8. mm. 24-26
9. mm. 26-29
10. mm. 27-30
11. mm. 30-32
12. mm. 36-38
M
U
L
L
M
U
U
M rh
U
M
U inv
M inv
141
13. mm. 39-41
14. mm. 44-47
15. mm. 45-47
16. mm. 47-50
17. mm. 47-50
18. mm. 52-53
19. mm. 52-53
20. mm. 52-53
21. mm. 54-55
22. mm. 54-55
23. mm. 54-55
24. mm. 57-60
L inv
L inv
U inv
M inv
U inv, rh,
L
M
U
L inv
M inv
U inv
U
25. mm. 61-64 M

26. mm. 62-67 L augm
27. mm. 64-67 U inv
28. mm. 67-69 L
29. mm. 67-72 M augm
30. mm. 69-72 U
31. mm. 72-75 M
32. mm. 77-79 L
33. mm. 77-80 M rh
34. mm. 77-83 U augm
35. mm. 80-83 M
Apart from the usual adjustments in the tonal answer, the most frequent
guise of the subject is the inversion (inv) in which all intervals appear
upside down. Augmentations (augm), i.e., statements in which each note
value is doubled, and rhythmic variations (rh) each occur three times. In
these variations, the dotted-quarter-note value of the subjects second note
also affects the fourth and sixth notes, thus stretching the first subphrase
in such a way that the second subphrase has to be shortened.4 Furthermore,
two-part and three-part strettos occur frequently.
Since this fugue does not feature any counter-subject, the intriguing
juxtapositions are those three-part strettos in which one of the voices
appears in augmentation. The example shows mm. 77-83:
Moreover, the first note is abridged in mm. 12 and 61, lengthened by anticipation in m. 26,
and ornamented in m. 39. The final note is omitted in m. 50, delayed in mm. 14 and 26, and
reached after an escape note in mm. 29 and 79 or after a chromatic passing note in mm. 6364. Finally, in the very dense three-part strettos of mm. 52-53 and 54-55, only the subjects
main thought can be tracked, but even this fragment contains irregularities at its end.
142
WTC I/8
There are ten subject-free passages. As can be seen from a cursory

glance at the music, these occur more often in sections where the subject
appears in separate statements, but episodes are scarceand shortwithin
the context of the strettos.
E1 mm. 6-7
E6 mm. 50-51
E2 mm. 103-111
E7 mm. 56-573
E3 mm. 14 -193
E8 mm. 60-613
E4 mm. 33 -35
E9 mm. 75-771
E5 mm. 413-443
E10 mm. 83-87
No episode material is related to the subject. The only subject-related
component between entries occurs in mm. 22-23 where the upper and
middle voices extend their respective subject statements by sequencing the
afterthought. These sequences, however, are inseparably linked to the
preceding entries. As they do not give the impression of a contrast between
statements, these measures are counted as a subject extension rather than as
an episode. There are no independent episode motifs, but a few note
groups that reappear occasionally.5 Comparing the measures featuring
these recurring note groups with the list of episodes drawn up above, one
finds that the only episodes to develop any melodically lasting figures are
E3, E5, and E10. These are not only the longest episodes in the fugue, but,
as will emerge later, indicators of important structural caesuras. Beyond
this correspondence on a higher level, there is no further remarkable
relationship between the episodes.
5
They should be mentioned not because of their thematic importance but because a
performer would wish to pay attention to shaping them consistently.
In mm. 14-16, the middle voice line C-F-E-D-C is imitated an octave higher (see
U: mm. 16-17 from C).
The same line may also be recognized, with a variation in the first interval and a cut
before the last, in mm. 18-19 (see M: C-B-A-G).
Also in mm. 17-18, the upper-voice line C-D-E-F-E-D-C is followed by a varied
sequence (see U: mm. 18-19 from D).
The same curve recurs in inversion twice in mm. 33-34 (see U: from E; M: from G varied), and six more times, in several slightly different variations, in mm. 41-44 (see U:
m. 41 from G, m. 42 from F, m. 43 from C; M: m. 42 from C, m. 43 from F, m. 44
from F).
In mm. 82-85, overlapping with the end of the preceding subject entry, the same curve
recurs three times in inversion (see M: m. 82 from A, m. 83 from A, m. 84 from F).
Also in mm. 82-85, the syncopated final note of the subject entry, together with the
subsequent falling broken chord, sets the model for a little curve that recurs twice (see U:
m. 83 from A, m. 84 from F.
Finally, all three voices are involved in stating a four-note scale segment as an upbeat
figure in mm. 85-87.
E minor/D minor
143
Almost all episodes serve as periods of relaxation between the measures

of tension built up by the subject statements. The only exception is E6,
where the tension seems to rise, both because of the unrelentingly driving
eighth-notes in all voices and because of the harmonic intensification
toward the A-major chord in m. 52.
Both the variety of note values (half-notes, quarter-notes, eighth-notes,
16th-notes, and various syncopations) and the predominantly stepwise
motion in the fugue indicate a rather calm basic character. The range of an
appropriate tempo is suggested by two inherent characteristics of the work:
The calm character requires a tempo slow enough to allow for each eighthnote to be felt with full melodic impact, while the augmented subject
statements require a tempo fluent enough to allow for the main thought
to be sung in one breath. The tempo proportion between prelude and fugue
may be simple: a half-note in the prelude equals half a measure in the fugue.
(Approximate metronome settings: prelude beat = 33 [quarter-note = 66],
fugue beat = 66.) The articulation that best expresses this character is an
overall legato, only interrupted by phrasing. These interruptions, which
occur frequently, should be gentle but distinct.
The only ornament, a complex trill in the lower voice of m. 74, seems
somewhat arbitrary. Proving neither structurally important (like the trills in
cadential closes) nor thematically consistent (like trills embellishing a note
in the subject or a counter-subject), it appears redundant if not outright
confusing. If played, it should consist of five notes: B-C-B-A-B.
The design of this fugue is obvious in the grouping of the material:
I The first section contains four subject statements: M U L / L. The
regular lower-voice entry (mm. 8-10) remains in the pitch range of
a tenor while the fourth (redundant) entry sounds like a bassthus
giving unsuspecting listeners the illusion of a four-part fugue.
II M. 19 introduces the first of three strettos, all of them using the
subjects original shape: MU UM UM. In the first stretto, the
imitation is at the octave at two quarter-notes distance, with both
voices equally faithful to the original. (The middle voice is in the
lead merely because it enters first.) In the second stretto, the subsequent voice enters after only a quarter-note but sounds in rhythmic
variation and with a shortened afterthought. In the third stretto, the
upper voice enters first but shows modifications at beginning and
end, while the middle voice gives a faithful rendition of the subject
and also relates more immediately to the target key of F major.
This is why the middle voice should be regarded as the leading
voice in this stretto.
144
WTC I/8
III
IV
VI
Section III begins in m. 30 with three statements of the inverted

subject: U inv M inv L inv. As has been shown above, the closing
episode of this section bears resemblance to the episode that
closes the first section (mm. 41-44 . 17-19), thus confirming
the structural analogy of these two sections.
M. 44 introduces two strettos comprised exclusively of subject
inversions: LU MU. In the first of these strettos (just as in the first
of the earlier strettos), the imitation is at the octave at two
quarter-notes distance, with both voices equally faithful to the
original. (The lower voice is leading since it enters first.) In the
second stretto (just as in the second of the earlier strettos) the
subsequent voice joins in after only a quarter-note but features
rhythmic variation and a shortened afterthought.
A new grouping is launched with three-part strettos and a return
to the subjects original shape after a long stretch of inversions.
The two strettos sum up and further expand the earlier developments. They sum them up in that they bring, in rapid succession,
the original subject and its inversion. The expansion is caused by
the increase in the number of entries in the strettos from two to
three. Both strettos show the lower voice in the lead, followed
by the middle voice and complemented by the upper voice. The
imitations are at the octave and separated by one quarter-note
each. All statements appear in a shortened version. After this
quite dramatic buildup, a single entry in the original shape (mm.
57-60) presents no new pattern but serves to release tension.
The final section is distinguished by an entirely new feature:
strettos with an augmented subject entry. From m. 61 onward
there are three such strettos: MLU inv, LMU, and L,MUM. In the first
two, the augmented entry imitates a normal-sized one at a
distance of two quarter-notes; another statement in original
values follows once the first is completed, i.e., halfway through
the augmented statement. All voices remain close to the original
shape. The second group is slightly extended in that the voice
carrying the augmented entry adds a repeated statementa
separate entry reminiscent of the similar one after the second
stretto in the fifth section. As before, this entry serves to release
tension before the impending greater climax. The last stretto is
the densest of all. The augmented leader is preceded by two
normal-sized entries, the second of them rhythmically varied,
and followed half-way through by a fourth entry.
E minor/D minor
145
There are few explicit cadential closes in this fugue. The work features
a twofold harmonic development, with a return to the tonic and a new
progression occurring roughly in the middle of the fugue. The first four
separate statements are in D minor, followed by a modulation to the
dominant (A minor; cadence in m. 19). The first stretto then modulates to
F major, which is reached in closing formulas at m. 301. After this, the
three separate inverted statements return to the home key of D minor. The
second stretto group sets out from the tonic and leads to the dominant on
m. 521, thus repeating the harmonic process observed in the first section.
Subsequently, the fifth section begins and ends in A, with a short detour
to F in the second three-part stretto. Similarly, the augmented-entry strettos
begin and end in A while the middle group touches several other transitory keys. It is left to the final episode to modulate back to the home key.
Owing to this harmonic device, the impressive fugue ends without a coda.
146
WTC I/8
Within the first section, there is a very gradual tension-rise supported

by the growing ensemble. At the same time, this growth in tension is
counteracted on various levels: by the strong tension-release in the subject
itself, by the lack of contrapuntal tension, and by the neutral character of
the episodes. In the second, third, and fourth sections, the dynamic increase between one entry or stretto and the next is negligible; emphasis
thus rests entirely on the contrast between subject-carrying passages and
episodes. The fifth and sixth sections allow us to distinguish three dynamic
levels: a level of extremely high tension in the strettos, a level of
drastically reduced, medium tension in the sudden separate statements
(mm. 57 and 72), and a level with a strong tendency toward relaxation in
the subject-free passages.
The overall dynamic development in this fugue shows the outlines of
continuous tension-growth, with two analogous departures and two small
internal relaxations. The steep drop or rise in tension between adjacent
sections is so pronounced that in comparison each section seems built on a
solid plateau. The only remarkable increase within a section occurs in the
sixth section. Here the final augmentation-led stretto is evidently conceived
as a climax outranking anything that preceded it.
The following simplified diagram shows the structural processes in
relation to the dynamic levels:
I
II
E
separate
entries
(original)
p-mp
III
E
two-part
strettos
(original)
dim
mf
separate
entries
(inverted)
dim
IV
E
V
E
two-part
strettos
(inverted)
p-mp dim
mf
VI
E
three-part
strettos
(orig/inv)
three-part
strettos
(w/augm.)
dim poco f dim f-ff
WTC I/9 in E major Prelude

This prelude is composed in polyphonic style. Apart from a few
missing rests and the not so unusual voice-splitting in the final measures,
its three voices are presented in consistent part-writing. Prominent material
within this prelude is based upon one motivic idea only. The composition
can thus be regarded as built along the lines of an invention. Sections are
determined by their harmonic patterns as well as by their development of
material. In addition, there is a striking structural analogy, revealing a
ternary form with coda.
The first harmonic progression ends at m. 31. Its steps are: m. 1 = I IV,
m. 2 = I V7, m. 3 = I. The cadential close coincides with the end of the
motif (G at U: m. 31). For this reason, and also because the bass has not
yet taken part in the polyphonic texture, these first measures do not
constitute a self-sufficient unit. The next harmonic progression concludes
in the middle of m. 8, after a modulation to the dominant. This cadence
represents a strong close and serves as a structural caesura. The dominant
key is first reached in m. 4. In m. 5 it gives way to its dominant (F major)
which, reinforced by sustained notes in the bass and several F-major
broken chords in the treble, governs the three measures preceding the
perfect cadence in B major.
Harmonic progressions indicate four structural sections: mm. 1-87
(E major to B major), mm. 8-131 (B major to F minor), mm. 13-14 (stepwise return from F via B to E, which serves here as dominant to A), and
mm. 15-24 (A major to E major, with an interrupted cadence in m. 227
followed by a full cadence in m. 24). The first section recurs in the fourth
(mm. 1-87 .15-227). There are only two small deviations from a literal
transposition: melodically, the ornamented note at m. 41 is substituted by
an eighth-note group at the beginning of m. 18, and harmonically, the
perfect cadence at the end of the first section is replaced by an interrupted
cadence (see the C-minor chord in the middle of m. 22).
The basic character in this composition is ambivalent, but the differences in articulation resulting from the two choices are small. On the one
hand, the interval structure in the principal motif features an entire broken
chord. If, with regard to the rhythmic pattern, one focuses on the polyphonically involved voices and their frequent three-eighth-note groups
147
148
WTC I/9
against dotted quarter-notes, one would have to interpret the rhythm as

simple. This underlying simple rhythm, in conjunction with the melodic
surface features, could then be read as indicating a rather lively basic
character. In this case, the dotted quarter-notes and quarter-notes would be
played non legato, the eighth-notes (quasi) legato. On the other hand, one
might read the melodic line in the first phrase as an elaborate hidden twopart structure based, after an initial rising sixth (E-C), on the stepwise
descent C-B-A-G set in double thirds, and consider the syncopations in
U: m. 3 and M: mm. 4 + 5 along with the 16th-notes in the cadential figures
and the bridging m. 14 as essential constituents of the rhythmic pattern.
In this case, the piece would be interpreted as representing a rather calm
basic character in which all melodic notes are to be played legato. In the
motif, the sequential structure implies phrasing after each E. In either character, cadential-bass notes (mm. 12 + 22-23) are non legato while dosido
formulas (L: mm. 2-3, M: mm. 12-13, and U: mm. 23-24) as well as notes
with tie-prolongation into an eighth-note (M: mm. 1-2, U: mm. 3, 4 etc.)
are legato. This leaves only very few notes that will be played legato in a
rather calm but non legato in a rather lively character.1
The preludes tempo, too, does not necessarily differ too much in the
two approaches. As always in Baroque pieces, the compound time signature indicates that the dotted quarter-notes and not the eighth-notes are to
be perceived as the pulse of the piece. At the same time, the ornaments and
the written-out 16th-note runs must be accommodated. A combination of
these two criteria limits the choices to a moderate pulse with gently flowing eighth-notes.
The prelude features several ornaments. The principal motif, in its first
presentation (mm. 1-31) and its recapitulation (mm. 15-171), carries two
mordents; these are dropped in all shorter developments of the motif. Both
mordents begin on the upper neighbor note and thus contain four notes, the
first of which falls on the beat. The mordent adorning the preparation of
the cadence in m. 7 is not derived from Bachs manuscript but from a
copy, as the brackets indicate. Because it is very appropriately placed and
embellishes the piece in an unpretentious way, playing it is a good idea.
(Consistency would then require integrating an equivalent mordent on the
middle beat of m. 21.) In both cases, the ornament begins on the main note
because it is approached stepwise. Finally, the compound ornament in m. 4
is launched from the lower neighboring note, as the little convex curve
1
These are: U: none; M: mm. 1: G, 5: E, 6: A, 10: all, 13: F, 15: C, 19: A, 20: D;
L: mm. 6: F, 7: G, 9: E, 10: all, 14: B, 20: B, 21: C.
E major
149
preceding the symbol designates. For performers with good finger dexterity,
a rendition with eight ornamental 32nd-notes sounds most convincing. A
longer ornament might blur the lower-voice entry of the motif.
m. 1
m. 7
m. 4
The initial phrase, which ends in a perfect cadence at m. 31, introduces

the motif of this invention. This motif has several outstanding features,
some of which are transitorily lost in the development and only recur in the
transposed recapitulation. As mentioned above, the line consists of a rising
broken chord whose octave is prolonged with a written-out inverted mordent
E-D-E and a melodic descent through the strong-beat notes C-B-A-G.
The first two of these melodic notes recall the inverted-mordent figure,
enhancing this lyrical embellishment by additional mordents and returning
in-between with an escape note to the octave E. This E thus serves as a
background layer in a hidden two-part structure. The last two melodic
notes are then linked by a free ornamental line. Horizontally, one could
regard the motif as consisting of head and body, the head comprising
the broken chord, the original inverted-mordent figure ornamenting the
octave, and the melodic step downward (which entails a harmonically
active movement), leading into a body in which the climax relaxes with a
gradual melodic descent and an equally gradual harmonic return to the
tonic. The motifs original accompaniment shows in the lower voice a protracted dosido formula and in the middle voice a (differently embellished) melodic parallel to the descent in the motifs body: A-G-F-E.
The development focuses on the motifs head, which appears twice in
the lower voice (m. 3 = tonic and m. 4 = dominant) before returning with
two statements on F (V/V and V7/V respectively) to the upper voice.2 The
remaining 1 measures borrow the idea of gentle curves from the end of
the original motif and close in m. 8 with a B-major cadence.
2
Note that in these measures, the motifs head carries on into another broken chord. For a
meaningful performance it is vital to distinguish between the original, active broken-chord
rise, found here on the first beat of each measure, and the passive one that serves as a
rhythmic extension to the melodic target note.
150
WTC I/9
From a perspective of harmony, the first and second sections perform

the pattern of head and body on a larger scale. This large-scale head
the first phraseis firmly rooted in the tonic before its partial imitation
undertakes an active harmonic step (from the E-major chord to the inverted
F9 at m. 33). The gradual melodic descent of the body, begun in m. 3 with
G, continued in mm. 5-6 with F-E, E-D, and completed after a lyrical
ornament in m. 8 with C-B, ultimately resolves into a new tonic as the
F-major chord, implied from m. 33 to m. 82, finally gives way to B major.
In the second section, the imitation of mm. 8-9 with the active harmonic
step leading to an inversion of the C-major ninth chord represents the
structural head, followed here again by a sweeping descent (U: mm.
9-12: B-A-G, D-C-B-A) accompanied by the gradual resolution into the
tonic of F minor at m. 131.
The subsequent brief passage encompasses two statements of the motif
in the lower voice that are even shorter than the previous ones. The sudden
16th-note runs, launched by the right hand and later taken up by the left,
add to set these measures apart from the remainder of the piece. Structurally, these two measures serve as a retransition to the recapitulation in m.
15; harmonically they lead to the function that, in Baroque compositions,
commonly opens a recapitulation: the subdominant (m. 15: A major). The
fourth section repeats the design of the first but ends in an interrupted
cadence instead. The coda presents three shortened versions of the motif in
the split middle voice before it concludes the work with a final cadence in
E major. This cadence is of particular interest as it is a hybrid between a
plagal cadence (represented by the bass step A-E) and an authentic cadence
(represented by D and C , the third and the ninth of the dominant chord).
WTC I/9 in E major Fugue

The subject of this fugue is short. It begins between beats (on the
upbeat to m. 13) and ends already at m. 21, encompassing six notes and one
rest. The dominant, implied on beat 4 where the unaccompanied line
features a rest, resolves onto the tonic at the return to E.3 The harmonic
movement on the rest, i.e., in the step from the implied prolongation of the
3
It is possible to assume a female ending, i.e., an unaccented extension without any

modification of the harmonic pattern. This would include the four 16th-notes up to the G
at m. 22. However, an analysis of the whole fugue shows that it makes more sense to relate
these notes to the counter-subject where they form an essential part of the first subphrase.
E major
151
highest pitch and its subdominant function to the

dominant, underlines the indivisibility of the phrase.
The harmonic base is a simple cadence in which the
subdominant is replaced by a seventh chord on its
relative. The subjects pitch pattern comprises seconds and one fifth
interval. The rhythm includes three different note values: 16th-notes, one
eighth-note, and one quarter-note. Later in the fugue, this rhythmic pattern
is occasionally interspersed with tie-prolongations.
Identifying the climax is easy in this subject. The second note F is the
melodic peak, the rhythmic focus as a sudden halt, and harmonically enhanced by the subdominant function and the implied change of harmony
on the rest. The rest after the climax sustains the tension, which is then
gradually released through the ensuing four 16th-notes.
The subject appears altogether twelve times in the course of this fugue.
1. mm. 1-2 M
5. mm. 7-8 M
9. mm. 20-21 U
2. mm. 2-3 U
6. mm. 9-10 L
10. mm. 21-22 M
3. mm. 3-4 L
7. mm. 16-17 M
11. mm. 25-26 U
4. mm. 6-7 U
8. mm. 19-20 L
12. m. 28
L
Only two of the subject statements are varied. The upper-voice entry in
mm. 20-21 includes an interesting octave displacement: a 16th-note-group
sets out one beat early from middle B. The extra time created by this
metric shift is filled with a figure leading to the climax C in the higher
octave. Similarly, the final entry also begins on the wrong beat. This time,
however, no correction is made. Instead, this statement concludes in its
metrically anticipated position and thus makes room, as it were, for an
accented cadential-bass leap to the tonic at the end of the piece.
The fugue features only one counter-subject. Introduced in the middle
voice where it is pitted against the second subject statement (from the
second 16th-note in m. 2 up to m. 31), it is almost omnipresent. This
companion is thus considerably longer than the subject. As it begins immediately after the downbeat, it bridges the three-eighth-notes rest that
precedes the first subject note. The counter-subjects pitch pattern and
perfectly even rhythm confirm the ornamental nature and, by extension,
the basic character of this fugue. A spiraling motion around a central note
(mm. 2-3: around E) and an obvious lack of harmonic or melodic features
that might enhance tension discourage any explicit dynamic shaping other
than a very soft rise at the beginning and a relaxation during the second
152
WTC I/9
half. For the same reasons and because of the very even rhythmic pattern,
subphrasing is also unlikely. Nevertheless, one can clearly distinguish two
segments. They are of equal length (each spans eight 16th-notes and ends
on a strong beat), which allows them to appear in various combinations.
Separate or joined, they play an important role in the fugue. The first
segment, CS-a, consists of a scale section from D to B followed by a
broken chord B-G-E. The second segment, CS-b, features the inverted turn
D-E-F-E and its varied repetition D-E-F-D. These segments appear
both as companions to subject entries and as episode motifs. They are
therefore worth being traced in detail.
As a companion to the subject, the complete version of CS-a + CS-b
appears only against the subject entries in mm. 2-3, 7-8, 20-21 and, with a
slight variation, in mm. 21-22. Other subject entries are accompanied by
1. CS-b + CS-b in sequential pattern (see mm. 3-4, 25-26)
2. CS-b + CS-b in a two-part imitative setting (see mm. 9-10)
3. a free figuration followed by CS-b (see mm. 6-7)
4. CS-b dissolving into a free figuration (see mm. 28-29)
5. no counter-subject material at all (see mm. 16-17, 19-20).
Within the episodes, the segments may appear in the counter-subject
order of CS-a + CS-b (mm. 4-5 L: D-D; mm. 8-9 M: A-A). In this order,
they may form descending sequences (mm. 11-12 and 12-13 L: G-G and
E-E; mm. 13-14 and 14-15 U: B-B and A-A; mm. 17-18 and 18-19 M:
B-B and G-G; mm. 22-23, 23-24 and 24-25 M: D-D, C-C and B-B).
They may even incorporate small variations that change the harmonic
outline (mm. 26-27 and 27-28 U: D-G and F-B). Or the episodes may
recall only one of the counter-subjects segments, either separately (m. 3
first half U: A-B) or in ascending sequences of CS-a (mm. 5-6 L: C-C,
mm. 15-16 U: G-G). In other words, there is not a single measure in the
E-major fugue without some quotation of the counter-subject segments.
Because of its unusual nature without any explicit dynamic shaping,
the counter-subject does not challenge the subjects supremacy. Yet one
must distinguish different levels of intensity in its 16th-notes and shape the
beginning carefully: while the
four ascending 16th-notes that
S
end the subject constitute a distinct diminuendo, the ensuing
group of ascending 16th-notes
represents a tension increase
(though a mild one owing to the
ornamental structure).
E major
153
There are seven subject-free passages in this fugue.

E1 = m. 3
E4 = mm.103-163
E6 = mm. 22-253
E2 = mm. 43-6
E5 = mm. 17-193
E7 = mm. 26-281
E3 = mm. 83-9
No episode is related to the subject, but all, as has been shown above,
feature at least one of the counter-subject segments. This consistency of
material throughout all structural portions creates a strong impression of
unity. In addition, the episodes contain three motifs. While inconsequential
for the design of the composition, their dynamic impact for the episodes is
important. M1 appears in the upper voice of mm. 5-6 (B D F B); its
broken chord pattern, immediately imitated in the middle voice but never
recurring thereafter, describes a graceful increase in tension. M2 is of
greater importance. Consisting of a syncopated appoggiatura and its
resolution, it is introduced in the middle voice of mm. 13-14 (A-G) and
followed by two descending sequences, each representing an emotional
accent and relaxation. Its accompaniment, a non-melodic eighth-notes
figure in the lower voice that is also sequenced twice, finds its climax on
the middle beat of each measure, preceding the high-tension interval (see
the minor seventh leaps A-G, G-F, F-E). M2 is taken up in E6 where
its appoggiatura-resolution pattern appears in the upper voice of mm. 22-25.
The accompaniment, again in the lower voice, is a non-melodic eighthnote figure in similar dynamic design, albeit in a different pitch pattern.
M3 is presented in the upper voice of E5 (see m. 17, G to tied note A) and
sequenced once in the next measure; it never recurs thereafter (although U:
m. 9 could perhaps be regarded as remotely related). It describes a tension
curve with the climax on the first 16th-note.
The episodes comprise two unequivocal closing formulas that are
structurally significant as they provide essential information for the understanding of the nature of the seven episodes. This is crucial in the case of
the first closing formula, which occurs in mm. 43-53, where the upper voice
features the typical dosido figure. E2 thus consists of two halves of very
different structural importance: the first measure provides a cadential close
in the dominant key of B major, while the continuation introduces M1 and
modulates back to the tonic, re-establishing E major in m. 64. The second
closing formula is found in m. 12, with the perfect cadence in C minor,
the tonic relative, completed at m. 131. Once again the upper voice presents
one of the typical figures, and once again this closing formula appears in
the middle of a subject-free passage. E4 is thus built similarly to E2; it also
consists of two segments, the first of which provides the cadential close
while the second introduces a new motif: M2. Two other closing formulas
154
WTC I/9
appear in subject statements. The first, characterized with the typical

cadential steps in the lower voice, accompanies the middle-voice entry in
mm. 16-17; the second, with a similar lower-voice pattern and a do-si-do
figure in the upper voice, supports the final entry in mm. 28-29. Concerning structural analogies, E6 is a variation of E4b (mm. 13-163 .
22-253).
The role the seven episodes play in the development of this fugue
reveals a strikingly regular pattern: E1 provides a bridge between the
second and third entries; it sustains the tension that has already been built
up without sounding too important lest it diminish the premonition of the
next subject statement. E2 engenders a release of tension in its first half;
after the cadence, M1 introduces a contrasting color with its own relaxing
tendency. E3, like E1 though longer, serves as a bridge between two
entries. E4, like E2, injects a tension release into the (also considerably
extended) first segment; after the cadence, the appearance of M2 provides
an even more distinct color contrast.4 E5 like E1 and E3 serves as a bridge.
Like its predecessors E1 and E3, it also appears in two-part texture, with
the lower voice resting. Yet because of its unique motif, this bridge sounds
more individual than the two others. E6 repeats the color contrast and the
overall relaxation of its model, the second half of E4. E7 once again
sustains the tension between the two final subject statements.
The basic character of the E-major fugue is rather lively. This is
determined above all by the rhythmic structure that, from the middle of the
first subject statement onward, presents an unbroken continuity of 16thnotes, and by the ornamental nature of the 16th-notes. Several brokenchord figures (M1, M3, M2 accompaniment, lower voice E7 etc.) confirm
this choice of character. The appropriate tempo of this fugue is very fluent,
almost virtuoso, with a swiftness that avoids any heaviness. The articulation encourages many nuances: an energetically bouncing non legato in the
subjects upbeat, a lighter non legato in the broken-chord patterns of M1
and M3 as well as in all other upper- and middle-voice eighth-notes, a
neutral, medium-length non legato in the M2 accompaniment and all other
cadential-bass patterns, a crisp quasi legato in the 16th-notes figures, an
ordinary legato in the melodic closing formulas of the treble (mm. 4-5,
12-13, and 28-29), and an intense legato in the appoggiatura-resolution
pairs of M2.
4
M2 is more prominent than M1 both in length and in further impact. It is set more clearly
apart from the ornamental character of this fugue owing to the melodic legato found in its
appoggiatura-resolution figure, and its descending sequences effect an overall relaxation.
E major
155
The tempo proportion of prelude and fugue also allows for two
choices, depending on just how fluent or how gently flowing the prelude is
taken. If the prelude is played with a pulse of metronome 80 or above, the
proportion can be simple: a beat (dotted quarter-note) in the prelude
becomes a beat (quarter-note) in the fugue. If the prelude is played with a
considerably slower pulse, a translation of eighth-notes into eighth-notes
works best. (Approximate metronome settings: [a] 96 for the dotted quarternotes in the prelude and the quarter-notes in the fugue; [b] 72 for the dotted
quarter-notes in the prelude; 108 for the quarter-notes in the fugue.)
The paramount force determining the design is this fugue lies in the
two explicit cadential formulas as well as in the structural analogies. Some
of these correspondences were already mentioned above, some should be
added here. Summing up, these are the factors determining the layout of
the fugue: E2 and E4 both consist of two segments, the first ending with a
cadential close emphasized by a closing formula, the second creating a
color contrast by introducing a motif that is unrelated to the primary
thematic material. Interpreting these two analogous cadences as the ends of
the first and second sections respectively, and looking backward from
these ends, one can observe that both are preceded by the same pattern of
two subject statements (on tonic and dominant), followed by an episode
that serves as a bridge, and complemented by another statement in the third
voice (mm. 1-43 . 64-113). The second section differs from the first owing
to the extensions of roughly one measures length each at both sides, i.e.,
additional episode measures both before the above-mentioned pattern.
Another analogy occurs between the second segment of E4 and E6.
Both present almost exactly the same material, albeit in different keys and
with the upper and middle voices inverted. If the second segment of E4 is
regarded as the beginning of the third section in this fugue, which follows
from what was said above, it would be logical to accept E6 as the beginning of the fourth section. Looking back again from the respective section
endings one can detect the following analogy: The three subject statements
preceding the end of the third section recall those of the first section. Like
them they are presented on the tonic, the dominant, and the tonic (mm. 1-43
. 193-22). The episode that, in both the first and the second sections,
serves as a bridge inside this pattern, now connects the entire pattern with
the additional subject statement at the beginning of the third section. While
we found that the second section was extended in comparison to the first,
we can now confirm that the third section continues this process of
extension at least at the beginning: the episode-segment opening the
section is longer (mm. 13-163 . 53-6), the linking episode inside the
156
WTC I/9
section is also longer (mm. 17-193 . 83-9 . m. 3), and there is an additional subject statement.
While the third section reveals itself as related to the first by the key of
its three subject statements, the fourth section similarly recalls the second
one. The two statements in the fourth section take up not only the keys of
the first and last statement in the second section, they even appear in the
same voices (mm. 25-26 . 6-7: statement in U on the tonic, and mm.
28-29 . 9-10: statement in L on the dominant).
Owing to the ubiquitous presence of counter-subject segments and the

absence of intensifying modifications of the subject, large-scale dynamic
developments are not called for in this fugue. The overall impression
should emphasize the ornamental nature of the piece and the subjects
cheerfulness.
WTC I/10 in E minor Prelude

This is another prelude developed from a forerunner in the Notebook
for Wilhelm Friedemann. The original, only 23 measures long, displays a
16th-note bass line below block chords in recitativo style (1/8-note chords
followed by 3/8-note rests, sounded on the first and third beats of each
measure). In this earlier setting, the prelude is harmonically determined.
The version Bach developed for the Well-Tempered Clavier comprises 41
measures. The first 22 measures rest on an almost unchanged 16th-note
bass line.1 The recitativo-style block chords appear now in the middle layer
and are often reduced to only two notes. The two strands are topped by an
elaborate cantilena reminiscent of a violino-obbligato part in a baroque
oratorio aria. This line sounds so unique that it may make listeners overlook that it is only an ornamented variation of the treble in the original
block chords. In other words: in its initial half the E-minor prelude is a
harmonically determined piece luxuriantly adorned.
The preludes second half is marked Presto. Although the word hints
primarily at a changed pace of the prevalent pulse, it leads performers and
listeners alike into expecting entirely new material. It may then come as a
surprise that these 19 measures rest on a 16th-note figure that, on close
inspection, proves strikingly related to that of the first half. What is new is
that here, the right-hand part also appears in 16th-notes. For the most part,
these move in parallel motion to the bass, but they nevertheless create the
impression of ornamented broken chords. Although the final nine-measure
portion features several details that will deserve special mention later, we
can safely maintain that this part of the prelude is also, above all, harmonically determined.
1
The Wilhelm Friedemann version differs little from the Well-Tempered Clavier version:
The G in m. 91 avoids the seventh jump downward by appearing one octave higher; the line
is thus smoother, but the bass-note descent from m. 6 to m. 9 is less satisfactory without its
expected target in the lower octave. In m. 13, the re-sharpening of the F already occurs in
the second half of the measure and not, as in the later version, only in m. 14. In the second
half of m. 15, the original version features G-B-C-D-C-D-C-B, the harmonically logical
figure that, in the adaptation, is modified to G-A-B-C-B-C-B-A in favor of an assimilation
with the ensuing measures. Finally, mm. 21-22 were originally erected on a repeated tonic
pedal. As the reworked prelude for the Well-Tempered Clavier is destined to continue with
a weighty second half, these measures are changed to support the intended modulation.
157
158
WTC I/10
The first cadence ends in m. 4 (m. 1 = i, m. 2 = ii7, m. 3 = V9, m. 4 = I;

the chords on ii and V sound in inversion over a do-si-do bass line). The
close should be marked with a melodic caesura after m. 43. On the last
three eighth-notes of m. 4, a melodic upbeat prepares the beginning of the
ensuing phrase. The next harmonic progression modulates to the relative
key G major, which is reached at m. 93. This progression coincides again
with a melodic phrase.
Throughout the entire prelude one can distinguish 8 structural sections.
They are grouped in a proportion of 5:3 into two encompassing sections:
I = mm. 1-43 (E minor cadence), mm. 43-93, 93-151, and 153-211 (modulations to G major, on to A minor, and back to E minor), and mm. 213-231
(new modulation to A minor); II = mm. 23-263 (A minor cadence), mm.
263-281 (modulation back to E minor), and mm. 28-41 (E minor affirmed).
The initial four-measure cadence recurs, transposed to the subdominant, at
the beginning of the Presto with a corresponding bass line. The diatonically
descending bass line of mm. 14-17 is taken up in mm. 27-30, surrounded
by an analogous harmonic pattern. Moreover, the return modulation from
the end of the first half is used to close a phrase in the middle of the Presto
(mm. 19-21 . 31-33). In other words, the final eight measures aside, the
Presto presents a recapitulation in a chain, without links, of three separate
passages from the first half of the prelude: mm. 23-26 . 1-4, mm. 27-30
.14-17, and mm. 31-33 . 19-21.
The tempo should accommodate the change in the middle of the piece
in such a way that the second half develops naturally from the first. The
most balanced way of achieving this is to play the Prestos 16th-notes twice
as fast as the initial 16th-notes. With regard to the articulation one may
assume that the preludes first half, with its cantilena in highly complex
rhythm, represents a rather calm basic character. The 16th-notes should
therefore begin in legato and only change to a crisper touch in the Presto.
The melodic line in the first half is also legato (Bachs paired slurring in m.
3 and the uncommon interruptions of the trills just before the suffixes in
mm. 10 and 12 add variety). The original block chords in the middle strand
are separated by rests and best played in a neutral tone color. In the Presto,
as in any lively piece, longer note values would have to be detached. Yet
apart from the leap in U: m. 24 and the half-notes in M: mm. 36-37 and
U: 38-39, there are no notes to which this would apply.
The first half of the prelude features five ornaments. As all of them are
long trills, the paramount concern will be the tempo of the shake. The
immediate choice of performers familiar with the rules of ornamentation in
Baroque polyphonic music would be to shake in notes twice as fast as the
E minor
159
shortest appearing values (i.e., with four notes to a 16th-note). For pianists
who find this technically too demanding or musically too congested, there
are, in this particular piece, indicators why a trill speed in slower values
might be not only acceptable but, some believe, even advisable: The fastest
occurring note values, the 32nd-notes, all appear in a strictly ornamental
context. These groups could thus be interpreted as spelled-out unaccented
ornaments and embellishments (see e.g., m. 1: an upbeat turn, m. 3: an
embellishment of the step A-G, etc.). It would therefore make sense to play
the ornaments appearing as symbols in the same motion. Moreover, four of
the five trills end in suffixes that are spelled out. This is either because an
interrupted trill is normally not expected to conclude with a suffix (as in
mm. 10 and 12), or because Bach desires a suffix differing from the
common shape (as in m. 14). Note, however, that the very regular suffix in
m. 1 is also written out. These suffixes appear in 32nd-note notes, a fact
that supports the rendition of the entire trills in 32nd-notes.
The four trills that are approached in note repetition (mm. 1, 10, 12,
and 20) start on the upper neighbor note. The fifth (m. 14), preceded by
stepwise motion, begins on the main (16th-) note. Regarding their ending,
only the trills in mm. 1 and 14 feature regular motion up to the written-out
suffix. The others are either interrupted by a sudden rest (cutting out one
32nd-note pair but leaving the remainder intact; see mm. 10 and 12), or
they end in an anticipated resolution, requiring a point darrt stop before
the anticipation (see m. 20). The example shows the complete rendition of
these trills in the slower of the two possible speeds.
The question of dynamic development in the preludes first half needs

to be addressed separately and in detail for the harmonic background and
the cantilena because of occasional contradictions of harmonic progression
and melodic features. Inside the completed cadences, the harmonic climax
falls on the beat that first manifests the subdominant harmony (in mm. 1-4
on m. 21, in mm. 23-26 on m. 241). In mm. 23-26, both hands support this
design. Conversely, in mm. 1-4 this is presented by the left-hand line and
the chords in the middle strand while the cantilena is headed independently
toward a melodic climax on the middle beat of m. 3.
160
WTC I/10
In passages where the original prelude featured sequences, separate

rules apply, which override those of the harmonic relationships. Such a
sequential pattern can be found in mm. 5-8 (B: E-D, D-C, C-B, B-A and
the paired pattern of the double thirds in the middle strand). In the model,
the chord in the first half of the measure relaxes toward the chord in the
second half (the inverted F7 resolves into the inverted B-minor chord).
Throughout the four-measure descent, this pattern of tension/resolution
should be repeated, each time on a slightly softer level. The final harmony
of this section, the G-major chord in m. 9, thus emerges as a soft ending.
The cantilena enhances this by preparing each of the seventh chords with
an elaborate upbeat while sustaining a long note into the ensuing moment
of resolution. Another succession of sequences appears in mm. 93-133.
Here the model is two measures long; the bass line G-G-F-E is sequenced
as E-E-D-C, and the harmonic progression of a modulation from G major
to E minor is sequenced as a move from E minor to C major. The climax
within the model falls on the step that actively leaves the momentary tonic,
i.e., on m. 101, after which the tension subsides gradually up to m. 113.
This dynamic pattern is then taken up in the sequence, where it sounds
slightly softer because of the generally descending direction. The cantilena
supports this pattern again by approaching these climax points in large
leaps (see mm. 9-10 = octave, mm. 11-12 = minor sixth).
Similarly, in passages where a bass-note pattern creates a line, this line
should outrank any small-scale increases and decreases in the harmonic
tension. In mm. 14-17 and mm. 27-30, the bass descends in diatonic halfmeasure steps. The superimposed harmonies suggest an increase, except
for one definite resolution at the respective ends of the short modulation
(onto A minor at m. 151, onto E minor at m. 281). The bass line Bach
composes reflects this break in the dynamic development by a break in the
straight pitch line (in m. 15 the A appears an octave higher than expected,
and in m. 28 the same holds true for the E.) The descent is prolonged with
chromatic steps that create an even stronger increase in tension as they
surpass the boundaries of the home key and bring about harmonies of high
tension. The resolution onto E minor then provides a relaxation.
The corresponding phrase in the Presto features an extension over a
diatonic bass line that reaches its climax on the chord corresponding to the
one mentioned above (i.e., that at m. 321). The ensuing relaxation is not
quite as complete, both because the dominant appears in minor mode (thus
lacking the leading-note) and because the A-minor chord in m. 33 is
inverted. In all these developments, however, the right-hand part supports
the general harmonic design.
E minor
161
This leaves only the final measures of each half of the prelude to be
discussed. In the first half, mm. 21-22 modulate actively toward A minor,
thereby suggesting an increase in tension. In mm. 33-41, a rather abrupt
initial increase is followed by the considerably softer beginning of the
pedal note B. For the duration of this pedal (mm. 343-39), the dynamic
growth is smooth and gradual. However, it manifests itself here not only in
the obstinate repetition of the bass note but also in an intensification of the
texture from three to four voices (compare m. 34 with m. 38. Note that this
increase of texture does not change the essential figures in the least). The
logical target for this crescendo is the interrupted cadence in m. 40, after
which the tonic of the home-key is approached in an overall diminuendo.
The simplified outline of the prelude in E minor given in the following
music example tries to capture these processes.
WTC I/10 in E minor Fugue

This is the only two-part fugue in the Well-Tempered Clavier. As its
subject modulates a fifth up to B minor, the closing cadence sets out with
the dominant of this target key, i.e., the F-A-C in the latter part of m. 2.
Harmonically, the resolution falls on the first beat of m. 3. Structurally
however, the subjects indivisible phrase ends properly only on the second
eighth-note of m. 3. The reason why this second eighth-note must be
162
WTC I/10
regarded as the final note, thus creating an unusual overlapping of the end
of the first subject statement with the beginning of the subsequent entry,
lies in the subjects melodic structure. The pitch pattern is unusual in that
it is composed in hidden two-part structure. In an attempt to determine
which notes belong to the melodic strand and which to the harmonic background, the following layers can be revealed: The melodic strand comprises an initial E (represented by either or both of the key notes in the first
chord), followed by a chromatic descent to the B on the downbeat of m. 2.
The end of the line is then provided by the A on the second beat in m. 2
and the two upbeat groups
G-F-G and F-E-D. The harmonic background is represented first by the third and
fifth of the tonic and then by
a keynote pedal in an offbeat
repeated-note pattern. This
segment of the backdrop is concluded by the ornamented keynote pedal
(E-D-E) after the downbeat of m. 2. The modulation to B minor manifests
in this strand with the offbeat notes C and A in m. 2. And these find their
resolution onto the target key on the equally offbeat B in m. 3. The example
above spells out this hidden
two-part structure. Consecutive notes hardly ever belong
to the same strand, thereby
transcending the concept of
intervals.
The rhythm in the subject comprises only 16th-notes and eighth-notes.
These two basic values remain predominant throughout the composition,
and even the third rhythmic unit, the tied-over quarter-note in the countersubject, does not change the fundamentally simple rhythmic pattern. The
subjects harmonic progression, too, is very simple. Aside from the initial
tonic and the two chords representing the modulation it only contains a
rudimentary subdominant in m. 13.
The swift motion permits no emotional involvement in melodic details.
At the same time, neither the simplicity of the harmonic design nor the
rhythmic pattern in uninterrupted 16th-notes offers any climax. Therefore,
the only feature capable of influencing the development of tension in this
subject is the descent in the melodic part of the structure. This leaves the
subject with a very straightforward dynamic gesture: an energetic beginning followed by a gradual decrease in tension.
E minor
163
The eight complete entries of the subject appear as follows:

1. mm. 1-3
U
5. mm. 20-22
L
2. mm. 3-5
L
6. mm. 22-24
U
3. mm. 11-13
U
7. mm. 30-32
L
4. mm. 13-15
L
8. mm. 32-34
U
The only modification the subject undergoes is the omission of the

unaccented second eighth-note at the end; this shortened form (in which
the background strand thus does not resolve) occurs in L: mm. 13-15 and
in U: mm. 32-34, i.e., in the fourth and eighth statements.
There is only one counter-subject. It is introduced against the second
subject statement but is shorter at both ends, beginning after m. 32 and
ending at m. 51. Its predominant note value is again the 16th-note. The
only longer note is a quarter-note on the downbeat tied to an additional
16th-note on beat 2. The counter-subject is not composed in hidden twopart structure but in a contour consisting of various ornamental figures.
Each four-note group comprises three neighboring notes, one of which is
repeated as an axis. All figures differ from one another. The placement of
the climax poses no problem: it falls on the sudden long note value that, in
addition, highlights a strong metric position. The two sketches show the
phrase structure and dynamic design in the E-minor fugue.
The first represents the

two-part result as it is
apparent in the written
version.
The second is a more

accurate depiction of the
effect created in a
listener.
164
WTC I/10
The eight subject statements in this fugue are very neatly grouped in
four pairs, each of which is followed by a subject-free passage.
E1 mm. 5-10
E3
mm. 24-29
E2 mm. 15-19
E4
mm. 34-42
The second half of E4 recalls subject segments (compare U: m. 39 and
L: m. 40 with the first subject measure, as well as U: mm. 41-42 with the
second half of the subject). E1 and E3 quote the counter-subject, with the
expected final note replaced by a large leap downward that is followed by
the rising sixth with which the subject concludes. One might describe this
hybrid figure as counter-subject body with subject tail. Its companion is
the first genuine, i.e., independent episode motif (see U: mm. 5-7). M1
consists of two rising broken chords followed by a four-note ornament, a
descending scale, and a long closing note. Two dynamic designs are
possible in this episode motif. The climax can fall either on the peak note
of the broken-chord sequence (e.g., the G in m. 5) or on the rhythmically
exposed tied note (e.g., the D in m. 6). The first rendition underscores the
motifs independence while the second, owing to the analogous length of
the climax note with that in the counter-subject, makes it sound like a free
imitation of the other voice. E2 and E4 provide two more motifs. The
lower voice in mm. 15-16 introduces M2, which is characterized by a
simple broken chord in eighth-note motion complemented by the repetition
of the chords third and octave. The upper voice contrasts this with M3,
which consists of two descending scale segments in an uninterrupted
16th-note pattern. With these two motifs Bach achieves a very balanced
effect by subtle means: the two ascents in M2 are contrasted in M3 by two
descents, and while in M2 the four-note broken chord was followed by the
shorter two-note leap, M3 begins with the shorter scale segment that is
then followed by the longer one.
The combination of M1 and the hybrid motif described above as a
counter-subject body with subject tail is followed by a descending
sequence and rounded off by two measures of almost completely parallel
structure in the two voices (see mm. 9-11 and 28-30). Similarly, the combination of M2 and M3, covering only a single measure, is followed by a
double imitation and a descending sequence before this episode-type, too,
is rounded off by a parallel motion of the two voices (see mm. 19-20 and
38-39). The relationship between the episodes is thus very straightforward:
E1 recurs in E3 in inverted voices. Similarly, E2 is recapitulated in
inversion in the first half of E4 (mm. 15-19 . 34-38).
The role the episodes play in the development of the composition is
only slightly more complex. E2 and the first half of E4, consisting entirely
E minor
165
of material not related to the subject or its counter-subject, provide a

certain color contrast. This is enhanced by the fact that M2 is the only
component in this fugue to appear in a longer succession of eighth-notes.
The descending direction of the sequences and the ensuing twofold descent
in the parallel-motion measure create a distinct relaxation. E1 and E3 are
more closely related to the primary material of the fugue and thus create
less contrast. The descending sequence suggests a tension release, but the
ensuing parallel-motion measures build a powerful curve that brings about
its own virtuoso climax. Hence this episode type ends on a level similar to
that from which it begins. The fugues three final measures serve as a
relaxing coda, owing not only to the decrease in the subject that is partially
quoted here, but also to the softening shift toward the conclusion in the
major mode and to the graceful ornamental arpeggio that replaces the more
austere sixth leap originally ending the subject.
The simplicity of the rhythm, combined with the ornamental structure
in the 16th-notes and the leaps and broken chords in the eighth-notes,
suggest a lively basic character for this fugue. The tempo requires moderate beats in 3/4 time, so that the eighth-notes will sound swift and the
16th-notes truly ornamental. (Ornaments indicated by symbols do not
occur in this fugue.) The articulation encompasses few non-legato notes.
These include the final leap in the subject and the equivalent in the hybrid
episode figure, the eighth-notes in M2, and the left-hand eighth-notes in
the penultimate measure. All 16th-notes as well as the quarter-notes tied to
a 16th-note are quasi legato, i.e., played with a technique of basic legato
but with crisp fingers. The tempo proportion between the prelude and the
fugue is best determined by the continuity of the 16th-note motion. As the
character of the fugue allows for a fairly swift tempo, the pulse of the
preludes Presto section can be continued directly into the pulse of the
fugue. In other words: an eighth-note in the preludes initial tempo equals
a quarter-note in the preludes Presto and a quarter-note in the fugue.
(Approximate metronome settings: prelude basic beats = 60, Presto beats =
120; fugue beats = 120.)
The design of the E-minor fugue is reflected unequivocally in its
analogies. The analogy of E1 with E3 and of E2 with E4 has already been
expounded in detail. In addition, the first two pairs of subject statements
are faithfully recapitulated, in inverted voices and different keys, in the last
two pairs. This perfectly symmetrical structure with two corresponding
sections (mm. 1-19 . mm. 20-38) is then complemented by the threemeasure coda that, as has been shown above, was the only episode portion
to state larger segments of the subject.
166
WTC I/10
The harmonic outline of this fugue comprises a very active movement

from one tonal area to another. This is caused by two facts: the subject
itself is modulating, and the answer is conceived as a real (not a tonal) one,
thus modulating further away from the original tonic. As the modulation in
each entry pair thus invariably leads two fifths up, one wonders how and
when the music can smoothly find the way back to the home key. Bach
solves this ingeniously by launching the two subject pairs in the second
half of the fugue from the subdominant (one fifth down) and the double
subdominant (two fifths down) respectively, so that the modulations will
logically return to the tonic at the end of the very last entry.
The tonal areas as represented by the subject entries are as follows:
mm. 1-3 (E minor to B minor), mm. 3-5 (B minor to F major),
mm. 11-13 (G major to D major), mm. 13-15 (D major to A major),
mm. 20-22 (A minor to E minor), mm. 22-24 (E minor to B major),
mm. 30-32 (D minor to A minor), mm. 32-34 (A minor to E major).
The dynamic outline follows the large-scale analogy with two gradual
build-ups. The tension continues to increase through the bridging episodes
E1 and E3 and is released in the concluding episodes of each half.
WTC I/11 in F major Prelude

This prelude, written in almost consistent two-part texture, is determined by a figure of consecutive broken chords with occasional auxiliary
and passing notes. With 16th-notes in the compound time of 12/8, this
figure dominates the entire composition without a single interruption.
The first cadence appears at the end of the figures first statement,
shortly before the end of m. 2. It thus defines the thematic component and
not a structural section. The next harmonic progression effects a modulation to the relative minor key, which is reached in the middle of m. 6.
Contrary to the first cadential close, this one indicates the completion of a
structural process.
There are altogether three sections:
1. mm. 1-67
F major to D minor
2. mm. 6-127
D minor to G minor
3. mm.12-18
G minor back to F major
The prelude comprises a striking symmetry in its first two sections:
mm.1-27 . 67-81 (transposed, voices inverted)
mm.3-67 . 9-127 (transposed only)
As the entire prelude consists of broken-chord patternsboth in its
16th-note figures and in their eighth-note accompanimentand does not
feature any truly melodic passages, a swift tempo is indicated. An ideal
rendition is one that allows listeners to feel the four-beats-to-a-measure
meter implicit behind the written compound-time designation. The articulation requires a light non legato for the eighth- and quarter-notes and a
leggiero-style touch for the 16th-notes.
The most noticeable ornaments in this prelude are the compound trills
on the half-notes in mm. 3, 4, 9, 10, 12, and 13. The long trill in mm. 14-15
may be included here as it fulfills basically the same function. The
execution is as follows: the speed of all these ornaments is in 32nd-notes.
Each of them ends with a suffix (the ones in U: mm. 4 and 9 including the
raised lower neighbor notes B and F respectively). The onset is as visually
indicated: concave curves signify a beginning with the four-note group of
upper neighbor/main note/lower neighbor/main note followed by the usual
trill motion, whereas convex curves indicate a launch with the inverted
shape of lower neighbor/main note/upper neighbor/main note followed by
167
168
WTC I/11
the complementing number of two-note shakes.1 The ornament in the final

measure is a point darrt trill, consisting only of four notes followed by a
stop before the final sweep. The same applies to the ornament in U: m. 17,
which should stop well before the end of the dotted quarter-note because
the resolution is delayed by a tie-prolongation. The example below gives
the execution for one representative of each ornament-type.
There is only one relevant figure, the

broken-chord in 16th-notes introduced
in the upper voice in mm. 1-2. It can
easily be reduced to the harmonic progression shown. This figure describes
a sequence in which a harmony when
inverted is expanded with its seventh,
then resolves, only to find the resolution itself turning into another seventh
chord and resolving again.
The dynamic development equivalent to this harmonic process is twofold:
the slight waves of tension-increase and relaxation appear under the
umbrella of an overall diminuendo, either from the beginning of the entire
figure to the return to the tonic (as in mm. 1-2) or, in the shorter versions
later on, only through a single chord/seventh-chord/resolution group. In its
development, the figure is shortened and complemented by a trill. As such,
it occurs in stretto imitation (see mm. 3-4). As the harmonic progression
departs from the tonic, the expected resolution is repeatedly substituted by
an entirely foreign note (see the C-major seventh chord followed by C in
m. 5etc.) without changing the basic pattern.
The dynamic outline of the first section presents an S-curve. After a
fairly assertive beginning and the two-measure decrease mentioned above,
1
In other words, the ornaments begin with the following notes:

m. 3: U on A and L on D m. 4: U on D and L on E,
m. 9: U on A and L on B
m. 10: U on E and L on A m. 12: U on A
m. 13: M on F
m. 14: L on C.
It is interesting to observe that Bachs basic idea seems to be that of a trill from below;
but wherever such beginning from the lower neighbor note would clash with a pitch just
heard before, he substitutes this beginning for the opposite one; see e.g., in U: m. 4, a
beginning on the lower neighboring note B would clash with the B in the preceding figure;
in U: m. 9, a beginning on F would sound wrong after the recent F.
F major
169
the modulation to D minor features an increase, with each figure in the

stretto beginning slightly louder than the preceding one. This leads to a
climax (which should, however, be gentle enough not to rival the section
opening) followed by a gradual decrease up to the end of the section.
The second section begins similarly with a two-measure decrease. The
passage featuring the development of the figure is preceded by an extra
measure (m. 8) that initiates the increase of tension, although it does not
contribute to the subsequent modulation. From the middle of m. 10 onward,
a gradual relaxation leads to the end of the second section.
The third section sets out with two stretches of 1-measure extension
each. Both feature continuous ornamental motion in one hand and a new
development of the figure in the other, and in both cases the harmonic
progression moves in active steps (i.e., in a pattern of ascending fourths;
see mm. 127-13: g7-C-F-B, mm. 14-157: d7-g7-C7-F). These developments
engender a dynamic rise much stronger than any of the previous ones, with
the preludes climax falling on the middle beat of m. 15. In the final descent,
the false imitation beginning in L: m. 167 may be given extra emphasis,
but neither the trills nor the eighth-notes in mm. 17-18 should interrupt the
gradual relaxation that continues through to the last measure. The prelude
ends in a gentle hue.
WTC I/11 in F major Fugue

The subject of the F-major fugue is slightly less than four measures
long. It begins with an eighth-note upbeat and ends at m. 41 where, after
the dominant harmony of m. 3, the A represents the resolution onto the
tonic. The subject includes two features that indicate subphrasing, both
occurring in m. 2 after the downbeat: a change in the rhythmic pattern (the
continuous eighth-notes give way to 16th-note runs with eighth-notes only
at their ends), and a considerable shift in the pitch level (a major sixth
below the earlier slow turn around the C). Thus the 16th-note run sounds
almost like a new beginning.
The rhythmic pattern is simple, consisting only of eighth-notes and
16th-notes. The interval structure contains no surprises either; apart from
the above-mentioned leap, all the subjects intervals are seconds. However,
there is a hidden superimposed line, created by the downbeats of the four
measures in the subject: D-C-B-A. Harmonically, the steps of this line
represent the steps of a simple cadence.
170
WTC I/11
The dynamic outline of the subject is determined by three criteria: the

shape of the two subphrases, their intensity relation, and the influence of
the superimposed four-note descent. Within the first subphrase, the D in
m. 1 is metrically supported by falling on a strong beat and harmonically
by representing the subdominant, and as no rivaling melodic or rhythmic
features claim attention, this D is the obvious choice for a climax. Within
the second subphrase, the eighth-note B represents the dominant-seventh
chord. It is also the only longer note value and, because of the ascending
scale leading to it, the note with the greatest melodic emphasis. The combination of the two subphrases results in two structurally very similar curves.
The first is more exposed both harmonically (the step to the subdominant
being more active than that to the V7) and rhythmically (the eighth-notes
sounding more forceful than the ornamental or scalar 16th-notes). Considering the line created by the four downbeats, the concept will be modified
slightly to accommodate the continuity
behind the two-subphrase structure. This
means that the C in m. 2 is no longer the
almost complete relaxation it would be if
seen only as the end of the first subphrase,
but appears integrated into the superimposed diminuendo D-C-B-A.
There are fourteen subject statements:
1. mm. 0-4 M
6. mm. 25-29 L
11. mm. 46-50 L
2. mm. 4-8 U
7. mm. 27-31 M
12. mm. 48-52 M
3. mm. 9-13 L
8. mm. 36-40 U
13. mm. 50-54 U
4. mm. 17-21 U
9. mm. 38-42 M
14. mm. 64-68 U
5. mm. 21-25 M
10. mm. 40-44 L
Besides the adjustment of the initial interval in the answer, which

occurs only once (U: mm. 4-5), the subject undergoes two other variations.
Both are introduced toward the end of the work. Three statements feature a
fill-up 16th-note immediately after the second climax (mm. 49-53: L, M, U).
As a result, the second subphrase consists exclusively of 16th-notes, reinforcing the contrast with the all-eighth-note first subphrase. The final
statement, substituting the four eighth-notes in the upbeat and the first
measure with an ornamental line, comprises even more 16th-notes.
F major
171
On three occasions, statements follow one another with considerable

overlap (see statements nos. 6-7, 8-10, and 11-13). In the latter two groups,
the third statement is not strictly in stretto with the first. However, the fact
that the two consecutive voices in each compressed group repeat the
leading statement literally on the same scalar step and with the same
variation supports a stretto-group interpretation. Parallel subject entries do
not occur.
There is only one companion to the subject, and it is not particularly
faithful. It is introduced in the middle voice against the second subject
statement (m. 4, second 16th-note G, to m. 81 C). Already when accompanying the third subject statement in mm. 9-13, CS appears shortened at
its beginning. The fourth statement finds a considerably distorted version
of CS shared between two voices (see M: mm. 18-19, complemented by L:
mm. 19-21, from C onward). The fifth statement is once again accompanied by the complete counter-subject. The two subsequent statements,
which build the first stretto, are surrounded by the second half of CS (see
U: mm. 26-29). Hereafter, the counter-subject is dropped for good.
As the sequence found in the first two measures is ornamental in
nature, the counter-subject should be regarded as an indivisible unit. With
regard to dynamic shape,
the swift 3/8 time favors
the downbeats. In the
absence of any other
features likely to enhance
tension, the climax falls
on the downbeat that is
highest in pitch, i.e., on E
in m. 6, etc.
There are altogether six subject-free passages in this fugue:
E1 mm. 8-9
E4
mm. 44-46
E2 mm. 13-17
E5
mm. 54-64
E3 mm. 31-36
E6
mm. 68-72
None of the subject-free passages is related to the subject. By contrast,
the counter-subject is quoted partially or even in its entirety (see E1, where
the first measure from CS is anticipated in the upper and middle voices,
and E2, lower voice). No genuine episode motifs are presented, but short
sequence-models gain local importance: In E2, the model established in m.
13: U + M is sequenced in mm. 14 and 15. In E3, the complementary
model introduced in mm. 31-32: U + M, which picks up the end of S/CS, is
sequenced in mm. 33-34. And in E5, the little one-measure figure from
172
WTC I/11
M: mm. 56-57 is imitated through all voices in mm. 57-61 and 63-64. By
contrast, E4 consists of nothing but a cadential close, as does the first
segment of E5, which repeats this cadence in varied transposition. The
final episode, too, is only a slightly more elaborate cadential formula. This
structural equivalence of the cadential closes aside, the episodes in this
fugue show neither analogies nor other relevant patterns.
The role each episode plays in the development of this composition can
be easily deduced from what has been observed so far:
E1, anticipating the ensuing counter-subject appearance, links two
statements, as does E2, quoting the entire counter-subject against
rising sequences.
E3 with its descending sequences brings a slight release in tension.
E4, acting as a cadential close, represents a definite relaxation.
E5a creates the same effect of relaxation, whereas E5b, determined
by its ascending eighth-note scales, creates the impression of an
increase in tension.
E6 sets out with the same ascending eighth-notes but then relaxes in
the final cadence.
The simple rhythmic pattern in conjunction with the scalar or ornamental structure of the 16th-notes indicates a rather lively basic character.
The tempo should be fast enough to reflect the bouncing mood of the
eighth-notes and the ornamental nature of the 16th-notes. The appropriate
articulation is quasi legato for the 16th-notes and non legato for all eighthnotes except for those in appoggiaturas and closing formulas.2
The relative tempo of the prelude to the fugue poses a problem because
of the inherent triple meter in both movements. The simple proportion, in
which an eighth-note in the prelude corresponds with an eighth-note in the
fugue, is possible. It strings the two pieces so firmly together as to almost
blur the confines of each. Moreover, there is the danger of monotony in
two consecutive pieces both based on mainly two note values in exactly
the same speed (dotted quarter-notes = 63). I propose another proportion,
which may seem more complicated but has the advantage of giving a
livelier, more interesting result. In it, an artifical group of three 16th-notes
in the prelude corresponds with an eighth-note in the fugue. (Approximate
metronome settings: dotted quarter-notes in the prelude = 72, eighth-notes
in the fugue = 144.) The different speed helps to differentiate touch and
color in the fugues heavier eighth-notes from the preludes lighter ones.
2
In mm. 45, 55, and 71, the ornamented notes in the upper voice are not detached from the
notes that follow. Likewise in the middle voice, the do-si-do figure in mm. 55-56 is legato.
F major
173
The fugue features mainly two kinds of ornaments: a long trill on the
penultimate note of the counter-subject and mordents in the typical closing
formulas. The trill in the counter-subject abides by the same rules that
would apply to a similar trill in the subject: it is approached stepwise and
thus begins on the main note, its speed is in 32nd-notes as these are twice
as fast as the shorter written-out note values, and it ends in a suffix before
resolving stepwise. The ornaments in the closing formulas (mm. 45, 55,
and 71) are short since their resolutions are in each case anticipated and
fall before the next strong beat, thus demanding a point darrt. They all
begin on the upper neighbor note and move in 32nd-notes or slightly
faster. In the two cases with dotted-note rhythmic figures, the ornament
stops short before the third eighth-note beat. The ornament in m. 28 is a
variation of the long trill mentioned above since the upper voice in mm.
26-29 features a variation of CS. Although shorter here, its beginning,
speed, and end should match the long trill. Finally, the mordent at m. 481
begins on the upper neighbor note and fills the note value with four swift
sixty-fourth-notes.3
Several features in this fugue provide indications for its design. The
subject statements beginning in mm. 21 and 46 both sound in reduced
ensemble and should be earmarked for possible section beginnings. In a
third instance, the stretto launched in m. 36 is accompanied by a sustained
pedal and thus also creates the effect of not actively involving all three
voices. The episodes between the second and third statements (E1) and
between the third and fourth ones (E2) are both related to the primary
material insofar as they quote the counter-subject. They thus suggest a link
between entries belonging to a single structural section. Furthermore, both
episodes have an ascending pitch outline. E3, by contrast, serves as a kind
of extension to the preceding subject statement and shows a descending
pitch line. The triple stretto + closing formula in mm. 36-46 is structurally
analogous to the triple stretto + closing formula in mm. 46-56. The final
cadential close is built along the same pattern as the two preceding ones
(compare particularly mm. 71-72 with mm. 45-46).
Harmonically, the fugue remains rooted in F major until m. 31. The
subsequent measures undertake a shift to D minor (see m. 32, reinforced
by its dominant, the A-major chord in mm. 34 and 36), which becomes the
basis of the triple stretto from m. 36 onward. A cadence in this relativeminor key closes this segment in m. 46. The following measures pick up
3
As this ornament seems somewhat out of keeping with the remainder of the piece, it may
be worth considering whether one might ignore the indication.
174
WTC I/11
the triple stretto in inverse order of entering voices and in G minor, the
relative minor to the subdominant. This key is confirmed with an explicit
cadence in m. 56. Two measures after this cadence the fugue is back in the
home key of F major, in which the final, varied subject statement appears.
This is again rounded off with a cadential formula similar to the two
previous ones.
The design that becomes apparent from these observations is unusual
insofar as it reveals a fugue consisting of parts rather than sections.
The first part, encompassing two sections, is united by the home key; its
two sections are glued together by the bridging power of E2 (which does
not allow for phrasing before the fourth statement, although the entering
order would suggest this) and by the density of material immediately afterward (which does not allow for phrasing after the fourth statement either,
although the reduced texture would suggest this). This part is rounded
off by E3, the relaxing episode in descending sequences that introduces the
modulation. Although the second part does not literally encompass
several sections, the explicit closing formulas in E4 and E5a give the
impression of three segments.
The first half of the fugue, welded together of two sections, describes
a gradual increase of tension. In the second half, each of the three segments
begins with a climax in mighty intensification followed by an episode
concluding in complete relaxation.
WTC I/12 in F minor Prelude

This prelude is truly intriguing. There are, or so performing pianists try
to convince us, three quite different approaches to its interpretation.
a) The melodic intensity in
the 16th-note lines may
compel concentration on
them as primary features
of the composition.
b) The almost uninterrupted
quarter-note pulse may
elicit an interpretation of
the piece as a metrically
determined composition
conveying a meditative
character.
c) The frequent recurrence of the figure
introduced in m.1
in the soprano and
the near-consistent
4-part writing hint
at an invention.
The examples above give an idea of what exactly is at stake. In view of
these choices, performers obviously face a host of interesting decisions:
melodic intensity, texture, coloring, and tempo, to name just a few, will be
different in each of the three cases.
The first harmonic progression concludes at m. 23. In it, the subdominant as an inverted seventh chord falls on m. 21 and the dominant, as an
inverted ninth chord, on m. 22. In terms of an invention, this cadence
coincides with the end of the motif, which has been presented over a bass
pedal, and thus does not define a structural section. The larger harmonic
progression ends at m. 91 in A major, the relative major key. An explicit
closing formula in mm. 8-9 suggests a significant caesura.
175
176
WTC I/12

I
mm. 1-9
i - III
F minor to A major
II mm. 9-12
III - V A major to C major
III mm. 13-163
V-I
C major to F minor, ending with an
interrupted cadence on a D bass
IV mm. 16-22
I
confirmation of F minor
Three of the sections constitute closed harmonic progressions while the
fourth ends in an interrupted cadence that can only be ascertained as structurally relevant because of its obvious correspondence: the beginning of
the prelude recurs as a seeming recapitulation that then turns elsewhere
(mm. 1-23 . 163-17). This is the only structural analogy in the prelude.
The choice of tempo, dynamics, and articulation depends largely on the
performers interpretation of the prelude type.
a) Performers who feel that the message of the composition is conveyed primarily by the 16th-notes will want to play in a tempo slow
enough to allow these note values to develop full melodic power. In
this case, the main concern is with the arabesque-like lines winding
through all voices. They will require careful and expressive shaping
endowed with all shades of emotional quality. Care must be taken
to ensure the constant flow of one voice into the next. As for the
articulation, the calm character requires that notes contributing on
any level to the melodic fabric be played legato, and only cadentialbass patterns and consecutive leaps be non legato. However, since
the first note in several of the leaps is at the same time a quarternote in another voice and therefore cannot be shortened (see, e.g.,
the notes in B: mm. 41-2 and 51-2), the few eighth-notes that actually
sound detached appear above all in the context of cadential
formulas (B: mm. 6, 8-9, and 11-12) or in similar bass lines (B:
mm. 9-10, 13).
b) Performers who wish to express a meditative character by focusing
on the continuous quarter-note pulse will want to choose a faster
tempo so as to gear the listeners attention to the metric feature
rather than the surface details. In this case, all emphasis will be on
the four beats in each measure, which will then be perceived as a
regular rocking motion with a rather low-key melodic intensity.
The 16th-notes (and to a lesser degree the eighth-notes) will be
treated as purely ornamental and played with an extremely light
touch, like delicate lace-work behind a more solid lattice work. As
for the articulation, all quarter-notes and eighth-notes will have to
be gently detached.
F minor
177
c) Performers who decide to draw on the inherent structural features

and interpret the prelude as an invention can choose any tempo as
long as this allows for the thematic figures to be perceived as units.
In this case, the interpretation will emphasize a dynamic shading
that distinguishes clearly between the different components of the
material, giving most intensity to the main motif and its possible
companion, less to any other recurring figures, and least to all other
passages. In terms of the articulation, either of the two approaches
explained above is possible, depending on whether the 16th-notes
are regarded, and played, (a) as essentially melodic or (b) as
ornamental figurations.
There are four kinds of ornaments, all appearing within the first ten
measures. The symbol embellishing the counter-motif (i.e., the main motifs
regular contrapuntal companion as introduced in S: mm. 2-3) denotes a
long trill that begins on the upper neighboring note, moves in 32nd-notes,
and ends in the suffix as indicated by Bach. This ornament recurs three
times in the course of the prelude, on the fourth beats of mm. 3, 4, and 10.
The symbol in m. 5 also designates a note-filling ornament. It begins on
the main note and, including a suffix its five notes, appears like a turn. The
trill in m. 8 lacks a resolutionboth rhythmically (since it does not fill the
time up to the next strong beat) and with regard to pitch (since the resolution note A does not materialize until after three intermittent notes). This
is thus a short ornament, a mordent that, launched from the upper neighbor, comprises one double shake. Similarly, the trill on the middle beat of
m. 10 is a point darrt trill. Approached stepwise and thus starting on
the main note, it comprises five notes, stopping with noticeable separation
from the following 16th-note.
As any description of a composition sets out from a premise as to what
the distinguishing features are, the F-minor prelude allows for three fairly
different descriptions, depending on the three choices outlined above.
a) If the prelude is regarded as a melodically oriented piece, the
features that determine the performance will all be sought in the
16th-note lines. Within the first nine-measure section there is an
initial curve in which the focus is entirely on the upper pitches and
the climax falls on the prolonged D on the downbeat of m. 2, with
a subsequent diminuendo up to the middle beat of the same measure.
This curve is followed by the first of many embellished parallels
(see mm. 23-3, S: A-B-C and T + B: F-G-A). These parallel
ascents, different in detail but joined in their general outline, are
sequenced twice (see mm. 33-4 and 43-5) and create a gradual
178
WTC I/12
crescendo. In mm. 5-63, the tension decreases slightly as all melodic
figures point downward (see in the soprano the central notes C-B,
in the alto the descending sequence, and in the bass the target notes
B, preceded by its leading-note A, and A, preceded by cadential
steps). The extended parallel ascent in mm. 63-73, in which all four
voices participate, then provides an even more intense increase of
tension, which peaks on the syncopated high F, relaxing afterward
toward the cadential close on m. 91.
The second section is distinguished from the first by a much higher
portion of highly emotional intervals. The recurrence of the hidden
parallels (mm. 10-11) prepares the ascent to the climax in m. 113,
after which the successive pairs of leading-note + resolution in
m. 12 bring about a prompt relaxation (see alto: B-C, F-G; tenor:
B-C; alto: F-E; tenor: F-E). The ensuing 3 measures of the third
section contain two tension-curves: a gentle one in mm. 13-14 and
a slightly steeper one beginning with the last 16th-notes in m. 14,
climaxing in m. 152, and ending in the interrupted cadence. The
fourth section begins with a dynamic curve identical to that at the
beginning of the piece. Its descent, however, is extended all the
way to the end of m. 18. A final curve (short increase and longer
decrease) concludes the prelude on a soft note.
In the overall design, the moments of highest tension are at the end
of the first section (where the drive is greatest in the extended
parallel) and at the beginning of the second section (where the
interval intensity is highest). As this interpretation renders the 16thnote motion as an uninterrupted line to which all voices contribute
in complementing manner, no melodic phrasing occurs.
b) If the prelude is regarded as a metrically oriented piece, the features
determining the performance will be sought in the quarter-notes.
The focus in the texture now tends toward the lower voices, and all
dynamic inflections are of a delicate nature. Within the initial
section, a first gentle curve with a climax in the middle of m. 1 is
followed by sequencing one-measure groups. As the bass in these
groups falls ever further, a gradual diminuendo results. This leads
to a state of floating weightlessness in the measure where the
quarter-note motion is temporarily suspended (see mm. 53-63). The
subsequent extended ascent leads to a slightly more pronounced
climax on the downbeat of m. 8, followed by a relaxation through
the cadential close.
F minor
179
The two sections that build the middle of the composition are less
regular in their pulse, featuring several passages in which quarternotes are interspersed with half-note beats (see the first halves of
mm. 9, 10, and 12 as well as the second half of m. 14). This impairs
the effect of meditative calm that prevails in the outer sections.
The soothing continuity of the quarter-note beats is restored with
the interrupted cadence and not broken again before the prelude
ends. Dynamic increases and decreases are even more subdued in
this section than in the preceding ones, concluding the composition
in a state of complete calm.
c) If the prelude is regarded as a structurally oriented piece with the
outline of an invention, the focus will be on the components of the
thematic material. These components are introduced within the
initial two measures. The example shows the components, in the
order of their importance in the composition, with their inherent
dynamic shape:
the main motif
the relaxing link
its regular companion
the syncopated figure
The main motif appears eleven times, more often than not with
a variation of its end. These are its statements: (ext = extended, var
= varied; inv = inverted)
m.
1 S
mm. 6-7 ext T
mm. 13-14 var T
mm. 2-3 T
mm. 9-10 var T
mm. 15-16 inv B
mm. 3-4 T
mm. 10-11
T
mm. 16-17
S
mm. 4-5 T
m. 13 var
S
In four of these eleven statements (mm. 2-3, 3-4, 4-5, and 10-11),
the motif is accompanied by its regular companion. On two other
occasions (mm. 6-7 and, in inversion, mm. 15-16), it comes with a
varied version of the stepwise ascent characterizing the companion.
The syncopated figure (mm. 2, 4, 5, 6, 11, and 17) and the relaxing
link (mm. 2, 3, 5, 5-6, 12, 15, 17, 18 [2x]), 19 [2x]), and 21 [2x])
account for much of the remainder of the piece.
180
WTC I/12
The first section is thematically dense, with main-motif entrances in
almost every measure and, after a short relaxation in mm. 5-6, an
extended quotation of the motif in the left-hand part. The two
sections that form the preludes center feature thematic measures
interspersed with cadential or imperfect closes. The final section
provides a liquidation of the thematic material. In this structurally
oriented interpretation, this liquidation together with the extended
pedal bass determines this last section as an overall relaxation in
which the pitch curve in mm. 19-20 plays only a subordinate role.
WTC I/12 in F minor Fugue

The subject of this fugue spans three measures. It begins on the second
beat of a 4/4 measure with three quarter-notes suggesting a long upbeat. Its
ending is defined by the return to the keynote after the half-note G that
represents the dominant. With regard to phrase structure, two views are
theoretically possible: The subject can be regarded as a single phrase or as
consisting of a main body (comprising the first six quarter-notes) followed
by an afterthought (the descent from B to F). As this mostly chromatic
afterthought may not appear quite substantial enough to represent a
subphrase of its own, the most convincing concept is a combination of the
two ideas. In it, the paramount buildup and decline of tension take place
within the first six notes, followed by a mere extension of the dynamic
decrease in the remaining descent.
The pitch outline is most intriguing. Beginning from the elevated
position of the fifth scale degree, the stepwise motion touches first the
natural leading-note above C (i.e., the minor sixth D), then the artificial
leading-note below C (i.e., the raised fourth B). This triggers an unusual
step: not only is the expected resolution to the leading-note omitted, but
what is more, the interval that follows, while looking like nothing more
special than a perfect fourth, is in fact located entirely outside the F-minor
scale. To top it all, its target is yet another leading-note (the raised seventh
leading up to the keynote) followed by its resolution F. The ensuing
descent seems to pick up from the recently heard B and continues through
three chromatic steps before falling onto the lower F.
Within the subject itself, the rhythm consists only of quarter-notes and
one half-note. However, a cursory glance beyond the boundaries of the
first subject statement reveals that throughout the fugue, eighth-notes and
16th-notes abound, as do tie prolongations.
F minor
181
The subjects harmonic background is unusual in that it sets off from

the dominant, reaches the tonic for the first time in the middle of m. 2 and
only then continues in an ordinary subdominant/dominant/tonic progression. Another extraordinary feature in the harmonic outline is the fact that
the first steps (the ones melodically approaching the leading-notes to C)
are both composed as interrupted cadences: the dominant (i.e., the implicit
harmony below the initial C) proceeds not to the tonic F minor but to D
major; the ensuing F7 chord (i.e., the implicit harmony below the next C),
which one expects to move into B minor, leads instead into a G7 chord.
In his very thorough two-volume work Analyse des Wohltemperierten
Klaviers, Ludwig Czaczkes takes into account all transitory harmonies
created by the counter-subjects. Here is the full harmonization of the
subject as Czaczkes conceived it, followed by a Roman-numeral analysis:
The dynamic outline should mirror the melodic and harmonic particularities. Beginning on the exposed fifth degree requires a slightly elevated
tension in the first note. The ensuing steps toward the natural leading-note
and on to the artificial leading-note create a powerful increase. Next, the
extra-scalar fourth interval that reaches the third leading-note represents
the apex of this dynamic curve. (For any wind or string instrument, the
climax would be between the two notes B and Ea solution that is not,
unfortunately, open to keyboard players.) The resolution onto the keynote
F brings such a relief after all these daring harmonic twists that it involves
an almost complete decrease of tension. What little is left will abate
gradually through the chromatic descent.
The 58-measure fugue comprises only ten subject statements:
1. mm. 1-4
T
6. mm. 27-30
B
2. mm. 4-7
A
7. mm. 34-37
A
3. mm. 7-10
B
8. mm. 40-43
T
4. mm. 13-16
S
9. mm. 47-50
S
5. mm. 19-22
T
10. mm. 53-56
B
182
WTC I/12
The subject undergoes only minimal variations. The adjustment of the

initial interval in the tonal answer occurs only once, and the final note is
only once delayed by a tie-prolongation. No parallels or strettos are used.
However, Bach invents three counter-subjects of very distinct character.
CS1 is introduced against the second subject entry in mm. 4-7. It
consists of five subphrases, thus creating the greatest possible structural
contrast to the subject. The predominant note value is the 16th-note, another
major contrast to the subject and its prevailing quarter-notes. The first
subphrase of CS1 (see T m. 4: F-G-AA-B-C) presents a slight increase
in tension. This is separated from the remainder of the counter-subject by
a leap of a minor ninth, thus launching a completely new rise in tension.
The second to fifth subphrases follow as sequences ascending toward a peak
(and dynamic climax) on the syncopated C.
CS2 is first heard against the third subject statement in mm. 7-10. The
first of its three subphrases comprises four eighth-notes, with a gentle
climax on C. The second begins almost like a partial sequence (F-E-D in
mm. 7-8 can be heard as a sequence of C-B-A in m. 7), with the stronger
climax on its first note F and an extended descent to the lower F (the rest is
tension-sustaining and does not indicate phrasing, as in CS1). A third small
subphrase ascends to a mild climax on the dotted B before relaxing
throughout the last three notes.
CS3 only manifests twice. First heard in its expected place against the
fourth subject entry, it begins later than any of the other counter-subjects
(see the upbeat to m. 14), ascends in a gradual sweep interspersed with
tension-sustaining eighth-note rests, climaxes on D, and resolves shortly
thereafter. The music example shows the contrapuntal play of the primary
thematic material as it is found in mm. 13-16:
S
CS3
CS2
CS1
F minor
183
The further development of the counter-subjects includes several irregularities that may make a convincing rendition of their phrase structure and
dynamic design difficult. The two most confusing details are the crossover
of voices1 and the swapping of sub-phrases.2
The F-minor fugue contains eight subject-free passages.
E1 mm. 10-13
E5 mm. 37-40
E2 mm. 16-19
E6 mm. 43-47
E3 mm. 22-27
E7 mm. 50-53
E4 mm. 30-34
E8 mm. 56-58
No episode is related to the subject. Instead, all seem to derive, in one
way or another, from the first counter-subject. It is helpful to distinguish
two main types of episode in this fugue: E1 is dominated by a motif that
uses the first subphrase of CS1 and continues with a segment from CS2 (A
mm. 103-11 . T mm. 83-91: a quarter-note followed after a 16th-note tie or
rest by a descent in 16th-notes). The CS1-segment is imitated in the bass
with a new ending, partly doubled by the tenor. The ensuing measures
sound like descending sequences.3 This episode recurs in exchanged voices
and with some variation in E4. Here, the bass is the leading voice, imitated
by the tenor, while the soprano adds an extended parallel to the descent.
1
After the middle beat in m. 7, the tenor interval A-F crosses over the descending alto leap
F-E; they regain their normal position in the middle of m. 8. In m. 13 immediately after the
middle beat, the same tenor interval moves the voice into what seems like alto position;
when the alto comes in with the upbeat to m. 14, it does so considerably lower than the
tenor; they recoup their ordinary places with the fourth beat of m. 14. In mm. 292-303,
soprano and alto cross. In mm. 47-48, there is such a gap between soprano and alto that one
easily mistakes the alto for a lower voice. When the tenor comes in an octave above the
alto, the confusion is complete. However, the voices disentangle on the last beat of m. 48.
2
In m. 19, the alto begins with the first subphrase of CS1 but carries on with the remainder
of CS2 (which, on top of everything else, involves an octave displacement on m. 212). At
the same time, the soprano (after having set off, in m. 19, with notes not belonging to the
primary material) takes over CS1, which the alto had abandoned, and concludes it properly.
In mm. 27-30, the tenor quotes CS1 but reduces its ninth leap to a simple second. The alto
presents the first subphrase of CS2 (see G-C-B-A), continues with notes aping the large
leap missed out by the tenor, and from the middle beat onward finds itself in CS3. The
soprano, after three non-consequential notes, completes the interrupted CS2. In mm. 34-37,
the first subphrase of CS1 is heard in inversion in the tenor, while the remainder sounds in
the soprano. The final trill is prevented by a tied note, the resolution avoided altogether. In
mm. 47-50, CS2 appears in the (alto-positioned) tenor, deprived of its first subphrase and
with a varied ending. Similarly, the CS1 statement beginning in the alto is allowed neither
its ninth leap nor its final trill, which is substituted by a written-out figure-work.
3
There is no way to make the actual notation heard, which has alto and tenor inverted; this
fact would therefore appear inconsequential for the interpretation.
184
WTC I/12
E6 also relies on the same material; this time, the alto takes the lead. (Here
Bach strings the motif to its sequence by filling the phrasing space with an
additional 16th-note; see the D after the middle beat of m. 44.) The imitation is presented in the bass, with soprano and tenor accompanying.
A second episode-type is established in E2. Here the initial impulse is
given in the soprano by a three-note figure that quotes only half of the first
CS1 subphrase. A counterpart enters with the bass, which derives its figure
from the inversion of the same first CS1 subphrase (the extension bends
back upward). The texture is completed by the imitation of the bass figure,
sounding in the alto, with the extension of the inverted CS1 subphrase
continuing downward. As in the first episode-type, this one-measure model
is also sequenced twice (mm. 16-17 . 17-18 . 18-19). This episode recurs
in the first half of E3 (mm. 16-19 . 22-25). Here, the leading voice is the
alto (which is not at all easy to convey under the parallel soprano). Its main
counterpart sounds in the tenor, with the imitation in the soprano. The
second half of this episode is based on the same material but uses it freely.
E7 is another episode to follow the model of this second episode-type. The
texture is very similar to that found in E2; the soprano is in the lead, while
the lowest voice, here the tenor, sounds the counterpart and its imitation
appears in the alto (mm. 16-19 ~ 50-53).
The two remaining episodes, E5 and E8, do not follow either of these
models. E5 features a figure derived from the first subphrase of CS1 (see
bass m. 37, sequenced in m. 38, and m. 39, sequenced in 40). The other
two voices involved in this episode recall the tied-quarter-note-plus-3/16note figure from CS2, matched here with free extensions. E8 features a
measure presenting nothing but this CS2 segment in four-part texture,
followed by two cadential measures.
The role played by each episode in the dynamic outline of this fugue is
determined mainly by their ascending or descending sequence patterns.
Descending sequences are found in E1, E4, and E8. Episodes that embark
on a relaxation before suddenly turning into a final crescendo are E3, E5,
and E6, while in both E2 and E7, the ascending motion dominates.
Both the complexity of the rhythmic pattern in this fugue and the high
degree of intensity expressed in the interval structure indicate a rather calm
basic character. The tempo is confined by features inherent in the composition. On the one hand, the 16th-notes must be calm enough to allow for
the expression of melodic intensity; on the other hand, the quarter-notes
must convey the impression of stringent movement in order to be felt as
the relevant pulse (instead of surrendering this task, as often happens, to
the eighth-notes).
F minor
185
The articulation that corresponds with this character requires legato in

all melodic parts. With regard to the components of the thematic material
this means: All notes in the subject and its three counter-subjects are legato.
Exceptions occur only where phrasing separates a note from the beginning
of the next subphrase (this happens particularly in CS1 and CS2 after the
first subphrases respectively). Non-thematic passages may feature consecutive skips or cadential-bass patterns, as in mm. 34-35, 40-41, and 43 (bass);
these must be taken non legato. In the episodes, octave leaps and cadential
steps in the lower part should be detached (see the bass in E1: C, C, D
etc.; similarly in E6; also in E3).
The relative tempo of the prelude to the fugue depends on which view
is adopted regarding the prelude-type. (a) Performers who feel that the
preludes message is conveyed primarily by the 16th-notes may choose a
simple proportion in which a quarter-note in the prelude corresponds to a
quarter-note in the fugue. (b) Performers who wish to express a meditative
character by focusing on the quarter-note pulse should choose the more
complex proportion of 3:2, where three quarter-note beats in the prelude
correspond to a half-note in the fugue, or six eighth-notes in the prelude
equal a quarter-note in the fugue. (c) For performers who decide to draw
on the inherent structural features and interpret the prelude as an invention,
either of the two proportions is possible, depending again on how one
perceives the prelude 16th-notes.
A last comment on this matter of tempo choice: One should not neglect
that the tempo of the fugue, too, may vary depending on the character of
the preceding prelude. Thus after a metrically oriented prelude in which
the shortest note values served as mere lace-work, performers may want to
give extra weight and melodic quality to the fugues 16th-notes, playing it
slightly more slowly than after a rendition of the prelude in which the short
values were granted full expressive power. (Approximate metronome
settings: (a) all beats = 66; (b) prelude beats = 84, fugue beats = 56.)
The fugue features two kinds of ornaments. One is part of the thematic
material, marking the end of CS1. The others are cadential ornaments
embellishing the final closing formula in mm. 57-58. The trill at the end of
the counter-subject abides by the rules that apply to a trill in a primary
component. As the note it ornaments resolves timely into the subsequent
downbeat, this is a note-filling ornament. Having been approached in
stepwise motion it begins on the main note, shakes in 32nd-notes, and ends
with a suffix. As an integral part of the counter-subject, this trill must be
transferred to mm. 9 (S), 30 (T), 15 (B), 55 (S), and 21 (S).
186
WTC I/12
The two cadential ornaments are also note-filling trills. The one on the
sopranos E begins on the main note, while the compound ornament in the
tenor begins as indicated from below. Both then move in 32nd-notes and
end in parallel suffixes.
The episodes in this fugue end in unusual ways. All of them close on
an imperfect cadence, and several even present these as unresolved chords
(see E1 in m. 13: F is suspended into a chord on C; similarly E2 in m. 19
and E3 in m. 27). This makes the episodes appear ill-suited as section
endings. By contrast, a cadential pattern that conveys structural closure
occurs in the context of a subject statement, at the end of the tenor entry in
mm. 40-43. These observations suggest that in this composition, episodes
are not conceived to round off preceding statements but rather to prepare
ensuing ones. A further factor that may give hints regarding the structure
of this fugue is Bachs use of the ensemble. Two subject statements appear
in reduced ensemble (T: mm. 19-22 and S: mm. 47-50) and are therefore
likely to mark section beginnings.
The harmonic progress through the subject statements is straightforward. The first six statements sound in the tonic area; four of them begin
like the first entry on C, only two in the dominant position on F (tonal
answer, m. 4) or G (real answer, m. 19). The seventh and eighth statements
relate harmonically to the area of the relative major. The former (mm. 34-37)
represents the tonic position of A major but is harmonized with reference
to F minor; the latter reads like the dominant of A and ends accordingly.
In the detailed steps of their harmonic progression, however, both are not
in keeping with the subjects original harmonization. The two final subject
statements return to the F-minor key, representing the dominant and tonic
positions respectively.
The design of the fugue thus presents itself as follows: The two statements in reduced ensemble must be regarded as first statements in their
respective sections. These two sections therefore begin with E2 (in m. 16)
and with E6 (in m. 43). The latter section-beginning is confirmed by the
explicit cadential-bass pattern at the end of the preceding statement. The
two statements harmonically referring to the relative major belong together
and form a section of their own. This section begins with E4 and ends with
the cadential-bass steps in m. 43. Its two entries are further united by their
harmonization of the subject, which deviates from the pattern established
earlier in the fugue, and by the fact that the subject appears abandoned by
its counter-subjects: CS2 and CS3 are entirely missing in both statements
while CS1 occurs only against the first of these entries and is varied at both
ends (see S: mm. 34-37).
F minor
187
Within the first section, the tension rises steadily from one subject
entry to the next. An increasing density of the material, up to a texture that
confronts the subject with three counter-subjects, supports an equally
growing intensity. The episode preceding the fourth statement grants a
transitory change of color along with a short relaxation, which is cut short
by the full force of four thematic voices. The end of the section seems
somewhat unresolved since a suspended G mars the final F-minor chord in
m. 16. This creates an effect as if this section was strung, if only by a thin
thread, to the subsequent one.
The second section begins with an episode of rising tension, causing its
first statement to enter already on a somewhat elevated level. The ensuing
episode generates a decrease of tension in its first half (up to the transitory
188
WTC I/12
cadence to A major in m. 25) but then turns around and prepares for the
next subject statement with a dynamic increase in mm. 26-27. This statement equals the final statement of the first section in the density of its
material and, consequently, in its intensity. Its ending furnishes the longexpected perfect cadence in F minor which, this time, sounds unimpaired
by any suspended voice.
The episode opening section III provides a more drastic change of
color owing to the position of its material: the two motivic parts are in the
bass and tenor, the alto pauses, and the soprano contributes only a parallel,
without independent force of its own. This darker color is enhanced in the
descending sequences. The episode linking this sections two entries is
built similarly to that linking the two statements in the second section. Its
first measures feature descending sequences, followed after a transitory
cadence in E major (m. 39) by a last-minute preparation for the ensuing
subject entry. The two statements themselves create much less tension than
those in preceding sections: with regard to texture, because of the much
lower density in thematic material; with regard to melody, because of the
major sixth that is more neutral in its tendency than the strong minor sixth;
and with regard to harmony, because their dissimilar harmonization omits
all the most striking steps. The second entryand with it the section
closes, as was mentioned before, in a definite cadence.
The fourth section begins with a four-part episode presenting three
measures of high material intensity. This comes as a surprise after the
preceding subject statement, which generated so little thematic motion
around it. The sequences, and with them the tension, descend toward the
middle of m. 46 but are followed by a most powerful, partly chromatic
ascent in the three lower voices. To fulfill the expectations raised in this
sudden crescendo, the soprano with the ensuing subject entry and the tenor
with CS2 begin almost two octaves higher than they ended a little earlier.
With the bass pausing, this creates a most striking shift and propels the
tension up immediately. The linking episode in the middle of this section
combines rising and falling tendencies, leaving the final subject entry as
unprepared as was the last statement in the first section. This statement
returns to the register in the center of the keyboard and to four-part texture
but replaces two of its counter-subjects with syncopated appoggiaturas (see
T mm. 53-54: C-B, A-G and A mm. 54-55: D-C, G-F). At the closure of
this fugue, Bach thus substitutes thematic density with emotional urgency
and creates a new kind of climax. The final episode can do nothing but
gently resolve this cumulated tension.
WTC I/13 in F major Prelude

This prelude is determined by two features: one rhythmic, the other
thematic. The rhythmic pattern is established in m. 2 and continues, with
only minor interruptions, through the entire piece. Its significant trait is a
syncopated right hand figure omitting the beats given by the left hand line
and thus creating a complementary rhythm. The thematic material is based
on three components: (1) a six-note broken-chord figure ending in an
iambic note repetition and regularly followed by an imitation, (2) a treble
figure over a falling bass in the above-mentioned complementary rhythm,
and (3) a cadential bass matched by a rhythmically varied closing formula
in the upper voice (first heard in mm. 57-67). These thematic components
follow one another but never oppose each other contrapuntally, as would
be the case in a motivically determined piece of this period. In fact, the
second and third components are presented exclusively in homophonic
settings. Only the first component is polyphonic in nature.
The first cadence concludes on the downbeat of m. 4, with the lines
continuing in both hands. The subsequent harmonic progression ends on
the middle beat of m. 6. It encompasses a modulation to the dominant key,
which is announced with the first appearance of the B in m. 4. The cadence
comes with an explicit cadential-bass pattern and an embellished version
of a typical closing formula in the upper voice (see the ornamented version
of CB-C in U: mm. 5-6). This close marks the end of the first section.
The F-major prelude comprises six sections:
tonic to dominant
I mm. 1-67
II mm. 67-121
modulation to the tonic relative
III mm. 12-157
modulation to the dominant relative
IV mm. 157-187 modulation to the subdominant relative
V mm. 187-247 modulation back to the tonic
VI mm. 247-30
tonic confirmed
As there are only three components of thematic material in this piece,
several portions sound reminiscent of one another. Genuine analogies,
however, do not occur.
The choice of tempo is based on the understanding that the compound
time signature does not designate the number of beats in a measure but
only aims at facilitating reading. The pulse is in fact felt in 4/8 time, with
189
190
WTC I/13
continuous triplets in each eighth-note. This eighth-note pulse should be

moderately flowing. There are two equally possible solutions for the corresponding articulation. Under the assumption that the flowing quality
prevails over the moderate mood, the non-syncopated longer values (i.e.,
the dotted eighth-notes) would be taken in gentle non legato while all other
notes sound quasi legato. Conversely, performers who feel that this composition exudes calm beneath its ornate surface would want to play mainly
legato. (This second option proves slightly more listener-friendly.) Both
options demand a precise distinction between dotted eighth-notes forming
part of the melodic bass line and others belonging to a cadential-bass
pattern. Moreover, careful phrasing, particularly in the complementaryrhythm patterns, is essential to assure that lines breathe.1
In the right hand, phrasing requires attention primarily in the passages
in complementary rhythm. As each subphrase in these passages ends with
a momentary interruption of the syncopated rhythm, the points of phrasing
are not difficult to spot. But since the final note of these patterns often
overlaps with the beginning of a closing formula, there are instances where
no cut may be possible between the two components.2
The prelude features one regular ornament in connection with the first
thematic component, the broken-chord figure ending in a note repetition.
This ornament is indicated by different symbols throughout the fugue. It
might, however, be preferable to decide in favor of one version and retain
this in all cases. Transferring the trill to structurally analogous notes that
appear unornamented is not appropriate in this piece since the notes to
which this might apply (e.g., m. 21) double as new beginnings. At any rate,
the trill is not note-filling since its resolution does not fall on the strong
1
From these two requirements results the following articulation in the bass line:
legato
mm. 1-3 (E); phrasing before C; legato mm. 3-5 (C-C)
non legato
in the cadential pattern mm. 5-6
legato
mm. 7-10 (E); phrasing before D; legato mm. 10-11 (D-D)
non legato
in the cadential pattern mm. 11-12
legato
mm. 12-14 (A); non legato mm. 14 (A)-15
legato
mm. 16 (first half); non legato mm. 16 (C)-17
legato
mm. 19-20 (F); phrasing before B
legato
mm. 20 (B)-21 (C), phrasing
legato
mm. 21 (D)-22 (G); legato mm. 22 (F)-23 (F) (non legato also possible)
non legato
in the cadential pattern mm. 23 (A)-24 (F)
legato
mm. 24 (F)-27 (tied C); non legato up to the end.
2
The treble is phrased as follows: m. 4, after downbeat A, m. 8, after middle beat F,

m. 10, after downbeat G, m. 21, after middle beat F, m. 29, after downbeat A.
F major
191
beat but is always both delayed and approached indirectly. This thematic
ornament can be rendered as a mordent (as indicated in m. 1), in which
case it includes four orbettersix notes whenever it begins on the upper
neighbor note, as in all cases of note repetition, and five notes when it
begins on the main note as in m. 121, or as a compound ornament (as indicated in m. 7), in which case it comprises eight notes, beginning always on
the upper note, describing a turn to the lower neighbor and back to the
main note, and ending with two trill pairs. The appropriate rhythmic
position of this interrupted ornament is such that the point of interruption
avoids any beat. The example shows two possible solutions for each version:
The initial section features all three thematic components. The brokenchord figure begins in the upper voice and is imitated in the lower voice.
The complementary-rhythm figure consists of two segments with overlapping phrase endings (phrasing in U after m. 41, in L already after the third
dotted eighth-note in m. 3). The closing formula begins in the right-hand
part on the middle beat of m. 5, in the left-hand part on the second dotted
eighth-note of the same measure. Harmonically this section describes a
modulation from F major to the dominant C major.
In the second section, the three components appear with different emphasis. The broken-chord figure is merely transposed; the complementaryrhythm figure much extended. Two descending subphrases in the treble
(see mm. 78-87, mm. 88-101) are accompanied in the lower voice by a single
unphrased descent followed by an additional curve. The closing formula,
however, is only represented by its bass part while the upper voice
continues in the complementary-rhythm pattern. The harmonic progression
leads from C major to D minor, the relative minor key of the tonic.
The third section is considerably shorter but features all components in
compressed versions. It begins at m. 121 with the broken-chord figure in
the bass, imitated in the treble before returning once more to the bass. The
bass then gives way almost directly to an extended cadential pattern. In the
treble, a little link precedes a very condensed complementary-rhythm figure
followed by the complete closing formula. The structure of the fourth
section is similar but even more compressed.
192
WTC I/13
The fifth section begins with an extended imitative pattern of the

broken-chord figure recalling the third section. This leads to an extended
complementary-rhythm figure. This segment is reminiscent of the second
section, especially in the ensuing cadential pattern in which the closing
formula of the upper voice is omitted. The section returns to F major, the
preludes home key.
The final section confirms the home key. The broken-chord figure is
stripped here of both its ornament and its polyphonic reply. Instead, after
a short quotation of the complementary-rhythm component, the brokenchord element is restated (L: m. 26) and followed by a short dominant
pedal with non-thematic figures in the right hand part. The composition
ends with the two subphrases of the complementary-rhythm figure over a
non-melodic bass and the original closing formula, completed in the upper
voice by a graceful, unaccented F-major-chord descent.
WTC I/13 in F major - Fugue

The subject of this fugue spans exactly two measures. It begins in m. 1
after an eighth-note rest and ends at m. 31. This downbeat marks the return
to the tonic after the dominant-seventh chord represented by B, G, and C
in the second half of m. 2. The subjects pitch outline is near symmetrical.
While the first and last intervals are leaps, all notes in between move in
steps. In the subject itself, the rhythmic pattern features eighth-notes, 16thnotes, 32nd-notes in the trill and its written-out suffix, and a quarter-note.
Counter-subjects and episode motifs later add several types of syncopation.
The overall effect is one of considerable variety in note values. The way in
which the various rhythmic values in the subject are introduced is interesting to observe. The initial eighth-notes give way first to 16th-notes and
then to the 32nd-notes of the trill. After this continuous acceleration, there
is a sudden stop on the quarter-notea halt that is even prolonged by the
restbefore the initial eighth-notes return.
This rest in the center raises a question: does the gap in the sound flow
signify an interruption of the tension or is the tension carried through the
silence? The harmonic development reaches its climax at m. 21, which also
marks the center in the symmetric patterns of pitch and rhythm. One may
surmise that the comparatively long note value, enhanced by the ensuing
rest, serves to sustain this climax before the subsequent stepwise descent.
The subject thus consists of a single indivisible phrase.
F major
193
Although the harmonizations of the subject occurring throughout the

fugue feature a number of intricate and varying progressions, it can be said
that in its main steps, the harmonic background of the phrase is that of a
simple cadential progression. The active step from the tonic to the subdominant (or, more frequently,
its relative minor) takes place at
m. 21, followed by the dominant
i
I V I
in the second half of m. 2 and
the tonic on the final note.
The subjects climax is thus ideally supported in all parameters:
representing the subdominant harmony and a sudden rhythmic stop, it is
composed as a kind of axis in a largely symmetric contour where the
preceding pitches emphasize two rises (see the first leap C-F and the step
from the trill C to D), while the ensuing tones perform two descents (see
the falling line D-C-B-A-G and the final skip C-A).
The fugue comprises eight subject statements.
1. mm. 1-3
U
5. mm. 15-17 M
2. mm. 3-5
M
6. mm. 20-22 L
3. mm. 5-7
L
7. mm. 28-30 M
4. mm. 11-13 U
8. mm. 31-33 U
The only modifications in the subject, both very common ones, occur
at its beginning and end. The initial interval is adjusted in the first answer
but never again thereafter, and the final interval is on two occasions (mm.
6-7 and 21-22) enlarged to a fifth to include a drop to the keynote. No
stretto or parallel statements of the subject appear.
The two counter-subjects are of very different character and importance.
CS1 is introduced against the second subject statement (see U: mm. 3-5).
The first of its two subphrases begins with an ascending octave followed
by a circling figure and a halt on a syncopation. The second subphrase is
an extended transposition of the first, beginning with a longer preparation
and closing with an additional dosido formula.3 The climaxes in the two
subphrases could fall either on the analogous syncopations (this results in
a somewhat saucy expression) or on the bottom of the curve before (on E
and A respectively), thereby underpinning the tender mood of the countersubject. The second subphrase, lower than the first, is softer overall.
3
Cf. m. 3: F-E-F-G-D-E-F-G with mm. 3-4: D-C-D-E-B...A-B-C.)
194
WTC I/13
Although CS1 is an almost constant companion to the subject, it undergoes significant modifications on its way through the fugue. Its initial
interval, originally an assertive octave leap, becomes gradually smoother:
reduced first to a seventh (m. 5: C-B), then substituted by a three-note
figure beginning with a fifth (m. 11: G-C-B), it is finally leveled to a
stepwise figure (m. 15: A-G-F; similarly in mm. 20 and 31). Moreover,
CS1 also loses momentum at its end. In mm. 11-13, the last syncopated
figure is replaced by a quarter-note and an appoggiatura that resolves
belatedly after a tie suspension. In mm. 15-17 and 31-33, the second subphrase turns into a cadential-bass pattern, in mm. 20-21, CS1 is shortened
to its first subphrase, and in mm. 28-29, it is entirely omitted.
The second counter-subject is normally expected as a companion to the
third subject statement. However, the line appearing in the upper voice of
mm. 5-7 turns into a rhythmic parallel of the subject. It is thus necessary to
look further, to the ensuing subject entry in mm. 12-13, in order to find the
true second counter-subject. CS2 begins belatedly half a measure after the
subject. It is characterized by a rising scale interspersed with pairs of the
repeated keynote. In keeping with this straight gesture in the ascent, the
development of tension also depicts only a single direction: no subphrasing
interrupts the continuous crescendo to the upper F. This counter-subject is
heard only three times altogether: after its introduction it recurs in U: mm.
20-22 and in U: mm. 28-30.
Another line that accompanies the subject twice and might, at a first
glance, appear as a counter-subject occurs in U: mm. 15-17 and in M: mm.
31-33. Upon closer inspection one detects that its end is composed as a
complement to the CS1 turned cadential bass, is actually making up, in
free style, for the omitted syncopated figure at the end of the first countersubject. This note group should therefore not be regarded as a primary
component in its own right. The following sketch shows the phrase structure
and dynamic design in the thematic material of this fugue.
F major
195
The F-major fugue comprises six subject-free passages:

E1 mm. 7-113
E4 mm. 22-281
E2 mm. 133-151
E5 mm. 30-313
E3 mm. 17-201
E6 mm. 333-35
The cadential close in mm. 22-231 partitions E4, separating a short E4a
from a much longer E4b. Of the remaining episodes, none is further subdivided. The material employed features components of the subject and of
the counter-subjects as well as an independent episode motif.
The subject appears, complete or represented by a segment, in all
six episodes. It is quoted most completely in E2 and E5. In the
upper voice of E2, the first four notes are a transposition of the
subjects head, the fourth note has the value of a quarter-note
followed by an eighth-note rest and is thus reminiscent of the
climax in the subject, and the ensuing six eighth-notes constitute a
free inversion of the subjects second half. The middle voice in this
episode sets out as an imitation, then omits the climax portion and
joins the end of the upper voice figure in parallel thirds. E5 is a
transposition of E2, with the upper and middle voices inverted.
Here the middle voice carries the more complete version of the
subject and is thus leading, while the upper voice is second in importance (which is not easy to play). Another quotation of subject
material appears as a bracket around the first episode: The head
motif opens this episode in the lower voice, stating a rearrangement
of the original pitches but retaining the eighth-note rhythm. The
middle voice concludes this episode with the subjects second half,
retained almost exactly in the shape of the original. The opening
segment in its particular rearrangement of pitches recurs twice in the
fugue: at the beginning of E3 (where it is sequenced; see L: mm.
17-18) and at the beginning of E6, where it is taken up faithfully.
Finally, the subjects head alone in its original shape reigns in the
largest portion of E4 where it is presented first in an imitative
texture of upper and middle voice, then in a pattern of descending
sequences in the lower voice.
Mcs2, a figure derived from CS2, is also found in all six episodes.
However, while CS2 is characterized by a hidden two-part structure consisting of a complete ascending scale over a repeated pedal, the episode
figure mostly plays with two moving lines and creates distinct little
gestures of varying length. Mcs2 occurs as follows:
196
WTC I/13
in E1 (U: m. 7, L: mm. 7-8, M: m. 8, L: mm. 83-10),4

in E2 (L: mm. 13-14),
in E3 (M: m. 17, U: mm. 17-18, L: mm. 18-20),
in E4 (L: mm. 23-253, M: mm. 25-26, U: mm. 26-27, L: m. 28),5
in E5 (L: m. 30), and
in E6 (U: mm. 33-34, L: m. 34, M: mm. 34-35).
M1 is an independent episode motif that occurs only in E1, E3, and E6,
and only ever in the upper voice. It is made up of an ascending fourth leap
followed by syncopated notes in various guises that, more often than not,
feature a descent. M1 is introduced in E1 (mm. 7-8 C-F-E-D-C-B). It
is then sequenced freely in m. 8 (D-G...) to m. 10 (...B) and again in
mm. 10-11 where the syncopations accommodate the dosido closing
formula. In its second version it reappears in E3 (mm. 18-19). Its shortened
third version is taken up in E6 (mm. 34-35).
The two episodes that, as was mentioned above, appear as corresponding owing to their fairly complete quotation of the subject (E2 and E5),
share a further feature: both are interwoven with the ensuing subject statement by an anticipation of the CS1 beginning. At the end of E2 (see L:
mm. 143-151), the note group from G to G anticipates the subsequent
group from A to A; the same recurs at the end of E5, in the first half of
m. 31. Finally, cadential-bass patterns occur at the ends of E1 and E6
respectively, as well as in E4athe only episode segment in this fugue
clearly serving as nothing but a cadential close. While in E1 and E6 the
formulas appear embedded in episode material, E4a features none of the
above-mentioned components or motifs. Instead, both the upper and lower
voices present obvious closing formulas, thus underpinning the concluding
gesture of this measure.
As has surely become obvious from the above analysis, several relationships exist between the episodes of this fugue. E2 corresponds directly
with E5. E1 is related to E3, which presents a shortened version, and to E6,
which repeats its beginning literally. E4b features analogies within itself;
the segment in mm. 23-26 corresponds largely with that in mm. 26-28.
The details unraveled above allow us to discern between three types of
episodes. One type, represented by E2 and E5, is very closely related to the
primary material, as it quotes the largest portion of the subject and is intertwined with the ensuing subject statement through its CS1 anticipation.
4
Interestingly, Mcs2 actually appears earlier than the counter-subject from which it stems.
L: m. 28 belongs to this episode in view of its material, although there is an overlap with
the beginning of the next subject entry.
F major
197
This type serves as a bridge. Another type, represented by E1, E3, and E6,
gains limited independence by using a distinct episode motif. In both E1
and E6, a cadential close is present, whereas this more definite ending is
missing at the tail of E3.6 This episode type provides a color contrast to the
primary material and conveys the impression of structural closure. The
third type is represented exclusively by the five measures of E4b. Utilizing
only segments from subject and CS2 and carrying these through various
harmonic steps, this episode seems purposefully to postpone the ensuing
subject statement.7 The message of this episode type is therefore neither
bridging nor closing but delaying.
The material of this fugue does not express a single character. In the
subject, the first counter-subject, the episode figures derived from these,
and the independent episode motif M1, the complex rhythmic pattern and
the overall stepwise motion clearly indicate a rather calm basic character.
By contrast, the second counter-subject and the episode figure derived
from it (Mcs2) feature a simple rhythmic structure and a pitch pattern consisting exclusively of skips, thus suggesting a rather lively basic character.
The conclusion to be drawn from these observations should accommodate
both facts. The tempo is moderately flowing; swift enough to convey
liveliness in the 16th-notes of CS2, but moderate enough to convey calm in
those of CS1. The articulation should take into account the differing
characters represented by the material. In the subject and the motifs
derived from it as well as in CS1 and M1, all notes are legato, whereas in
CS2 and Mcs2, all notes are quasi legato.8 Non legato articulation is
appropriate in the cadential-bass notes of mm. 11, 16, 22, 33, and 34 as
well as in the octave leap of m. 25.
The relative tempo of the prelude to the fugue is ideally represented by
the compound relationship of 3:2. This translates as follows: three dotted
eighth-notes in the prelude equal two quarter-notes (half a measure) in the
fugue. Approximate metronome settings: prelude beats (dotted eighthnotes) = 96, fugue beats = 63.)
6
It may not seem too farfetched to speculate that this missing close of E3 is made up for by
E4a which, as was shown, presents an explicit cadential formula.
7
It reminds one faintly of that episode portion in the C-major fugue that serves as something like the development section in a sonata form fugue. The reader may wish to
refer back to the chapter on WTC I/3.
8
For the sound balance within these components it is of greatest importance that the
frequent note repetitions do not sound with a greater degree of detachment than the other
intervals; any effect of slurring should be carefully avoided.
198
WTC I/13
The composition comprises two ornaments. The first embellishes the

subject. It should be transferred to all subsequent subject entries, slight
technical inconvenience being no excuse. The difference of symbols in
mm. 1, 3, and 15 should probably not be taken to indicate different executions. In each case, the trill is note-filling. It begins on the main note
since it is approached stepwise, moves in 32nd-notes, and ends with the
suffix that, in all cases but one, is written out. The other ornament appears
in the cadential measure of E4a. It is a mordent launched from the upper
neighbor note G and then touching down on F twice. Its shakes may be in
32nd-notes or slightly faster.
As has been shown above, the fugue features two explicit cadential formulas. One of them marks the end of the first episode. The other, following
the sixth subject statement, was identified as the only subject-free segment
devoid of all the components characterizing the material of the other
episodes, i.e., a typical cadential close. While these cadential formulas
already provide an important clue for the design of the fugue, earlier
observations regarding the episodes and their role in the development of
the composition also contribute to the understanding of structure. E2 and
E5 have been acknowledged as bridging episodes, each stringing together
two statements belonging to the same section. E1 and E6 were found to
exert concluding power, while E3 showed a similar tendency but lacked a
final closure (to be complemented by E4a). Finally, E4b was shown to be
neither bridging nor concluding; one might therefore assume that its
position is neither in the middle nor at the end of a section.
Having said all this, it is time to state one noteworthy analogy in the
fugue. Three separate facts have already been recognized: the correspondence of E2 and E5, the relationship in material of E3 and E6, and the
correspondence of the subject entries in mm. 15-17 and 31-33 (both are
accompanied by a variant of CS1 in which the end appears substituted by
a cadential-bass pattern, and by a third voice that in its second half makes
up for the abandoned tail of CS1, at least in its rhythmic idea). To sum up:
mm. 11m-18
correspond with
mm. 28-35
subject statement + CS2
subject statement + CS2
bridging episode
bridging episode
subject statement
subject statement
with CS1 irregularity
with CS1 irregularity
E3
E6
The overall harmonic development begins after three statements in the
realm of F major with a modulation to the dominant and leads to a close
with a C-major cadence at the end of the first episode. After an immediate
F major
199
return to the home key, the third episode modulates from C major to the
tonic relative minor that, after a subject entry in D minor in mm. 20-22, is
confirmed in the cadential close of m. 22-23. E4b undertakes the return to
the home key before the next subject entry. Thus all but one of the subject
statements appear in F major, either on its tonic or on its dominant.
The following conclusions may be drawn from what has been observed:
The first section encompasses three subject statements and one episode,
with the cadential formula at the end of E1 concluding this section. The
second episode comprises two statements that appear strung together by
the bridging E2, followed by the episode type that, elsewhere in this fugue,
concludes a segment but here lacks the final cadential close. The ensuing
statement appears thus almost as redundant, an impression enhanced by the
fact that it is the only one in the minor mode. The cadential measure of E4a
concludes this second section in m. 231. The third section begins with an
opening (delaying) episode followed by the passage that was recognized
as structurally analogous to the second section without its quasi redundant
final entry. In this section, the third subject statement is now completely
stripped off.
Within the first section, the tension grows from the first to the second
subject entry along with both the growth of the ensemble and the emergence
of the contrasting first counter-subject. However, from here to the third
200
WTC I/13
entry, there is almost no increase as the upper voice does not contribute
independent material but runs a rhythmic parallel to the subject. The
ensuing episode begins with a distinct color contrast. It then allows for a
slight increase of tension as M1 comes in and is sequenced upward, only to
give way to the relaxing cadential close that concludes both the episode
and the first section.
The beginning of the second section can be regarded as the climax not
only of this section but of the entire fugue because the juxtaposition of the
subject with its two counter-subjects creates the greatest degree of polyphonic contrast. The bridging episode sounds softer than the surrounding
subject statements but not necessarily as different in color as did E1. It is
followed by an entry that makes no use of the newly introduced CS2 and
thus sounds less assertive than the preceding one. The ensuing episode
begins again with a color contrast and a slight increase of tension in connection with M1. Its relaxation is interrupted by the quasi redundant entry
that is the most subdued both in this section and in the fugue.
The third section sets off in an almost floating atmosphere. The remaining two subject entries and two episodes repeat the development of
the second section, omitting, however, both the climax and the anticlimax.
The first subject statement of this section is accompanied by CS2 but not
CS1 and thus sounds less polyphonically exciting, and the quasi redundant
final entry is dropped entirely.
WTC I/14 in F minor Prelude

With the exception of very few measures, this prelude exhibits strict
two-part polyphony. The opening idea is imitated in the lower voice and
also recurs constantly in the course of the piece. As the initial imitation is
placed on the fifth, the composition can be identified as composed in
accordance with the principles of a fugue.
The first harmonic progression concludes at m. 21. As it coincides with
the initial statement of the subject, this cadence does not indicate a
section ending. Conspicuous cadential formulas outside subject statements
occur in mm. 12, 18-19, and 21-22. The first and third are perfect
cadences, one in the dominant, the other in the tonic. The third presents an
imperfect cadence (with the subdominant and dominant of F minor).
Structurally relevant analogies are scarce in this prelude. Only the two
measures immediately following the above-mentioned perfect cadences are
composed alike (mm. 123-143 . 221-241, transposed from C to F). Their
three-part texture with a sustained note in one voice and eighth-note + rest
figures in another refers to the beginning of the piece (mm. 12-13 and 22
. m. 1).
The basic character of this prelude is rather lively. This follows both
from the simple rhythmic pattern consisting predominantly of 16th- and
eighth-notes in a straightforward setting, and from the pitch pattern
featuring leaps in the eighth-notes and a mixture of ornamental figures and
broken chords in the 16th-notes. In fact there seem to be hardly any notes
that would require emotional depth; so the tempo, reflecting truth, should
be quite swift. Articulation demands a light quasi-legato touch for the
16th-notes and an effortless non legato for the eighth-notes.
The three ornaments (mm. 12, 18, and 21) in the prelude embellish cadential closes. All appear in a pattern typical for Baroque closing formulas,
a dotted note followed by the anticipated resolution, and are therefore point
darrt trills: they stop short considerably before the succeeding 16th-note,
preferably on the half-beat eighth-note or slightly before. The trills in m.
12 and m. 18 both begin on the upper auxiliary and, moving in 32nd-notes,
comprise four notes. The ornament in m. 21, as it is approached stepwise,
is launched from the main note and features only three notes: two
32nd-notes and a syncopated eighth-note.
201
202
WTC I/14
Since the prelude in F minor is being analyzed under the assumption

that it is built along the lines of a two-part fugue, a crucial question is: how
strict a fugue it is, and where does it deviate from the fugal pattern? The
subject is established in the upper voice.1 This initial subject statement
brings about the first deviance from the modela deviance that is common
in preludes composed as fugues: it is not presented as an unaccompanied
line but comes with two companions. One of these voices moves in
eighth-notes interrupted by eighth-note rests and doubles the peak notes of
the subject in parallel compound thirds, while the other consists of a
sustained tonic keynote and its final octave displacement. In the course of
the fugue the two structurally analogous statements, those after the perfect
cadences in mm. 12-13 and 22-23, are also presented in a sudden three-part
texture, each time with one voice acting as a pedal and the other moving in
parallels to the subject peak notes. (For reference, this line of parallel
eighth-notes will here be called C for companion.)
The following two subject statements allow for two fake countersubjects to make their appearance. Both, as will be shown, derive from the
companion. C1 is introduced in the expected position against the second
subject entry in mm. 2-3. It consists of leaping eighth-notes and, if regarded
as a hidden two-part structure, describes in its lower part the same parallel
in compound thirds to the subjects peak notes as the companion of the
opening entry. (This can be seen most clearly in mm. 9-10.) C2 makes its
first appearance in mm. 4-5 against the third subject statement. It is characterized by broken chords that are filled each with an additional passing
note. The line built by the lowest notes of these chords is, once again, a
parallel to the subjects peak notes.
The example shows the relationship between the subject and its three
companions. In light of the two following regular counter-subjects, the
eighth-notes in m. 1 are revealed as a feature to be taken seriously since
both derive strikingly from it. For easier comparison, all passages are
transposed to the
tonic and reprinted
with the subject in
the upper voice.
harmonizing
companion
1
C1
C2
Whether the A on the downbeat of m. 2 or the third 16th-note F constitute the intended
melodic ending is almost impossible to decide.
F minor
203
A third companion to the subject is added in the second half of the

composition. This companion appears in block chords that strike in eighthnotes separated by eighth-note rests. Its rhythmic pattern is thus that of the
first-measure companion, and its scalar descent derives equally from the
compound-third parallel to the subjects peak notes (see, in the right-hand
part of mm. 14-15, the lowest notes E-D-C-B). This figure thus continues
the development we have so far witnessed, leading from one mutation of
the companion to the next in a logical way, from single notes via hidden
two-part structure and broken chords to block chords. As these chords
make no attempt whatsoever to establish any melodic line, one would not
normally give name them a counter-subject. However, for the sake of
easier identification and because of their striking analogy with the other
companions these chords shall here be referred to as C3. What is truly
surprising is that the inversion of the subject is invariably accompanied by
this C3, which then does not build the usual parallel. As a consequence, the
ensuing subject entry in mm. 15-16, also an inversion, is accompanied by
a compromise: one descending (originally parallel) and one ascending
(now parallel) line. Finally in mm. 19-20, things appear upside down in
that the subject is here returned to its original shape but the scale in the
accompanying chords rises.2 The fourth and final step in the development
of the companion is the explicit parallel in compound thirds. This parallel
occurs in mm. 20-21 but never again.
The subject entries in this fugue and their respective companions thus
present the following picture (Sinv, Spar = subject inversion or parallel):
mm.
U
L
1
S
C
2
3
C1
S
4
5
C2
S
mm.
U
L
12 13 14 15 16
S
C2 C3 S inv
C S
S inv C3
6
7
S
C2
C2 S
9
10
S
C1
11
12
17
19 20 21
C3 S
S
S par
22
S
C
23 24
C2
S
18
The few subject-free passages in this fugue are episodes in the usual
sense. Some take their material directly from the primary features; see,
e.g., the episodes in mm. 3-4 and 5-6, which display a motif (in imitation)
composed of the subjects turn figure and a leap that recalls C1. The lower
voice in mm. 8-9, the upper voice in mm. 10-11, and the lower voice in
mm. 16-17 use the same motif, while mm. 11-12 and 17-19 constitute
2
See in m. 19 the upper notes of the right-hand part. The immediate relationship between
these chords and the chords in mm. 14-15 becomes apparent when one compares the lower
notes here with the uppermost pitches there.
204
WTC I/14
motivically free extended closing formulas. Shorter cadential formulas are

found in the two remaining passages (mm. 21-22 and m. 24).
An attempt to define the structure in this fugue by grouping the
subject statements into sections would probably carry the exercise too far.
The only relevant structural caesuras in this highly virtuoso piece occur at
the cadential closures. Thus performers should try to convey a polyphonic
composition in two halves followed by a coda, and not the (theoretically
possible) six fugal rounds.
WTC I/14 in F minor Fugue

The subject of the F-minor fugue spans four measures. It begins after
a rest with the second quarter-note of the 6/4 measure, thus conveying the
impression of an extended upbeat. Its conclusion is reached at m. 41 where
the melodic line returns to the keynote F after an ornamented G representing the dominant harmony. The length of the final note is exceptional
in m. 4; some later subject statements end with only an eighth-note.
The pitch pattern in the subject and in the entire fugue features mostly
stepwise motion. The rhythm is highly varied, comprising eighth-notes,
quarter-notes, dotted quarter-notes, etc. up to the subjects tied 4/4-note.
Syncopations appear frequently, two of them in the subject itself.
Phrasing in the subject permits two options, which create quite different
effects. On a small scale, there are three similar note groups that can be
interpreted as varied sequences: F-G-A is sequenced in contracted note
values in G-A-B, and sequenced again (after an additional A-G pair) in
A-B-C. It is therefore possible to view the subject as consisting of three
consecutive ascents the last of which is complemented with a descending
motion. In this case, subtle subphrasing after the long notes ending the first
two ascents (i.e., after A and B in m. 2) is adequate. On a larger scale, it is
also possible to imagine that these notes combine to a single gesture: an
ascent through F-G-ABC (or even F-G-A-A-B-B-C). In this
case, the long notes are directed forward instead of serving as transitional
endings, and the tension between one long note and the next is further
enhanced by additional artificial leading notes (A, later B). If this is what
a performer feels, the subject should not be broken into smaller units but
conceived as a single curve. (By the way, the short instances of backtracking during the ascent find their equivalent in the descent, which also
falls in two sweeps.)
F minor
205
The subjects harmonic layout is simple in its large steps but fairly
complex in the small ones. This is caused both by the artificial leading
notes in the subject and by the high degree of chromaticism throughout the
fugue. The main steps of the simple progression are represented by the
three long notes, with A (m. 1) for the tonic, B (m. 2) for the subdominant,
both C and the ornamented G (m. 3) for the dominant and the final F
(m. 4) for the return to the tonic. The more complex harmonic progression,
which Bach uses, e.g., in mm. 15-18, is shown in the example below. The
Roman numerals reveal how Bach in his harmonization of the subject
actually passes through the keys of B minor and C majorthe subdominant and dominant keys of F minor:
The C at m. 31 provides the climax. The highest pitch, it is the goal of

the ascending sequences or, in the alternative interpretation, the target of
the powerful ascent with its enhancing chromaticism. The preparation of
the climax and with it the degree of urgency expressed in the subject is
different in both concepts. In one scenario, the climax is reached in consecutive sweeps each of which begins newly in a relatively relaxed way.
(A string or wind player would decrease volume on the long notes.) In the
second case, the tension grows constantly and thus more powerfully through
all notes. (A string or wind player would increase volume during the long
notes so as to incorporate them into the long crescendo. Keyboard players
can only imagine this.) The character expressed in the two different interpretations has a considerably different impact.
There are nine entries in the course of the fugue:
1. mm. 1-4 T
4. mm. 15-18 S
7. mm. 29-32 T
2. mm. 4-7 A
5. mm. 20-23 A
8. mm. 32-35 B
3. mm. 8-11 B
6. mm. 25-28 S
9. mm. 37-40 S
206
WTC I/14
The subject undergoes hardly any modification. The answer is real,

with all intervals kept intact, and only in one statement is the beginning
slightly varied (see in m. 25 the 3/8 upbeat C-F-D instead of the simpler
2/4 approach C-D). However, a more significant modification, the
inversion of the subject, does occur twice. It is heard in the context of the
alto entry in mm. 20-23 and recurs in the bass statement in mm. 32-35.
Neither stretto nor parallel are found in this piece.
There are two counter-subjects. One is conceived as a fairly regular
companion to the subject, the other materializes only twice. CS1 is introduced against the second subject statement, in mm. 4-7 (tenor). It consists
of two clearly distinct halves. The second segment presents the extended
version of a closing formula (see the typical leaps D-G-C and the
do-si-do figure with its classic syncopated rhythm), while the first half is
characterized by frequent note repetition. On closer inspection one detects
that the pattern actually consists of note pairs: C-B, B-A, A-B, and so on.
The pairs are harmonically conceived either as main note + anticipation
(mm. 4-5 and 13-14) or as appoggiatura + resolution (see, e.g., mm. 8-9).
Whichever their harmonic background, these patterns are well known in
Bachs music under the name of sigh figures. In this counter-subject it is
vital to distinguish between pairs that constitute genuine sighs and others
that do not and should therefore sound different.3 Dynamically, the first
subphrase represents a gradual diminuendo to G. The second subphrase
builds a crescendo up to the half-note C and ends with a final relaxation.
CS2 first appears against the third subject statement, i.e., in mm. 8-11
(tenor). It begins on the second quarter-note (C) and consists of three
subphrases. The first ends on the dotted half-note B. Its ending is then
sequenced in diminution, after which the third subphrase sets out from the
syncopated C with an ascending portion of the melodic F-minor scale
followed by a descent in natural F minor. This subphrase, and with it the
entire counter-subject, should regularly end with the resolution on the
downbeat of the final measure. In this first appearance, however, the resolution note A is delayed. Several dynamic representations are possible for
this counter-subject. The climaxes of the first and second subphrases could
fall either on the long B and on the F in the diminished partial sequence,
3
The two notes in a true sigh are usually neighbor notes; they always represent different
harmonic steps. Leaps, especially if they occur within a single harmony, are therefore not
sighs and should not be played as pairs. Notes that appear in the context of sigh figures
but do not belong are: the octave leap in m. 15 where the lower F is the end of one subphrase, the higher F the beginning of the actual CS1; similarly in mm. 19, 22, 29, and 32.
F minor
207
or on the initial notes in these subphrases. In the third segment, the

syncopated C competes for the climax with the highest pitch of this
subphrase, F. This counter-subject recurs only once in its entire length,
and even then it is heavily varied (compare A: mm. 29-32 with T: mm.
8-11). As if to make up, the third subphrase of CS2 appears separately in
mm. 16-18 (tenor). The example below shows the phrase structure and
dynamic design in the subject and its two counter-subjects:
The F-minor fugue comprises six subject-free passages:

E1 mm. 7-8
E3 mm. 18-20
E5 mm. 28-29
E2 mm. 11-15 E4 mm. 23-25
E6 mm. 35-37
We can distinguish three episode motifs, two of them remotely related
to the primary material: M1 in its simple shape of four ascending notes is
the most prominent episode motif in this fugue. Both its rhythmic gesture,
which consists of three eighth-notes followed by a three eighth-note value,
and its pitch pattern, which contains artificial leading notes, recalls a segment from the center of the subject (compare E1: E-F-G-A, F-G-A-B
and B-C-D-E, C-D-E-F with mm. 2-3: G-A-B-C). The dynamic
shape of M1 is a crescendo. M2, consisting of the first six notes of CS1
with their upbeat, two sighs, and a final note, is prevalent in E3, E4, and
E6. Dynamically this motif takes up the subtle groupings from the countersubject, albeit within an overall diminuendo. M3 appears exclusively in E2
and is only vaguely defined. One could describe it as a concave curve with
the climax in the center, at or close to the lowest point of the curve.
208
WTC I/14
The fourth episode stands out as different from the others because it
emerges from the preceding subject statement without any clear beginning.
The soprano in mm. 23-24 continues the descent in syncopations begun
toward the end of the subject entry (see mm. 22-23 F E, continued in mm.
23-25: D-C-B). At the same time, the bass presents sequences of the
variation of the CS1 beginning that accompanied the end of the inverted
subject (mm. 22-23: D A . 23-24: B-F and, shorter, . 24-25: G-E).
Similarly, the alto also extends the end of the subject statement in varied
sequences (compare m. 23: G-A-B-C with the ensuing groups F-B and
E-F-G-A). This entire episode can thus best be regarded as an extension
of the preceding subject statement.
The structure of the episodes and their interrelationship is interesting.
E1 recurs faithfully (except for the octave displacement) in E5 and thus
creates a distinct symmetry. E1 also reappears, slightly varied and with an
accompanying voice, as the first segment of E2 (see T + B: mm. 11-12) and
of E3 (see S + B: mm. 18-19). The second segment of E3 features a parallel
presentation of M2 in the outer voices (soprano and bass) followed by
imitation, also in parallel (in what seems like the inner voices but is actually
the alto and the bass after an octave displacement, since the tenor is
resting). This pattern is taken up in E6: see m. 35 with a sequence in m. 36.
This creates a second structural analogy.
Finally, there are two cadential formulas apart from the one in the final
measure. One appears in m. 20 where the key of C minor is confirmed in
a perfect cadence. The other formula, with a dangerously similar-looking
bass line, marks the end of the above-mentioned subject extension (see
m. 25). This one, however, is not only an imperfect cadence, but it also
overlaps with the beginning of a new subject statement in the soprano. It
thus does not qualify as a structural caesura.
The features described above also determine the role played by each
episode in the dynamic development of the composition. E1 and E5 both
begin slightly softer than the end of the preceding subject entry, after
which they prepare the subsequent entry in a twofold crescendo. E2 sets
out with a similar buildup. This is followed by a gradual release generated
by the overall descending direction of the M3 appearances. A complete
relaxation is held at bay by the frequent quotations of the outgoing M1 in
the bass of mm. 13-14. E3 begins with the same active gesture followed by
a gradual diminuendo in m. 19 and a complete relaxation in the perfect
cadence. E6 picks up the gradual release. It is a bridge to the subsequent
entry. Finally, E4 continues the diminuendo in the subject ending and thus,
at a moment of particularly low tension, creates an anticlimax.
F minor
209
Both the fugues pitch pattern with its prevalence of stepwise motion
and the rhythmic organization with its variety of note values are unequivocal in determining the basic character of this fugue as rather calm. The
frame for the tempo is set on the one hand by the required tranquility of
character, and on the other hand by the rather long note values, which still
need to be perceived as alive. In other words, the eighth-notes must be
slow enough to convey serenity, but not so slow as to impede the listener
from taking the entire subject on one breath, and thus cause the subject to
fall into pieces. The articulation is generally legato. Only the first half of
the first counter-subject as well as the episode motif derived from it feature
slurring in pairs, following the pattern of the sigh figures. The only notes
to be played non legato are the consecutive leaps in the second half of the
counter-subject (e.g., m. 6: D-G-C) and the cadential-bass patterns in
mm. 20, 25, and 39-40. For the relative tempo of the prelude to the fugue,
a good and feasible solution is: one measure in the prelude corresponds
with half a measure in the fugue. (Approximate metronome settings:
prelude beats = 112, fugue beats = 84.)
The only ornament in this composition is the trill in the subject. As it
is complemented by a resolution, it is note-filling. Approached in stepwise
motion it begins on the main note sustained for the duration of an eighthnote and then shakes in 16th-notes, ending with a suffix. This trill, being
an integral part of the subject, must be transferred to the D in m. 6 and to
the Gs in mm. 10 (where it is indicated in brackets), 17, and 31, i.e., to the
second-to-last notes of all original-shape entries,4 as well as to the B in m.
23 and the E in m. 34, i.e., to the second-to-last note of inverted entries.5
The entering order of the voices and their respective surrounding
texture, in conjunction with the explicit cadence at the end of E3, define
the fugues binary structure: The full ensemble of four voices is reached in
the fourth statement. In the middle of m. 20, a cadential formula closes this
section in C minor, the minor dominant of the home key. The subsequent
entries manifest a protracted suspension before the full ensemble is reached
once again:
4
In the subjects final statement, the Picardy third of the ending impedes execution of the
trill. The ordinary trill on G would shake with A as its upper note. This A, however,
clashes with the A in the middle voice. A trill with A is obviously also out of the question
because of the preceding A. This subject ending must therefore remain unornamented.
5
Students sometimes ask whether trills are also upside down in inverted statements. The
answer is: no, ornaments are not inverted.
210
WTC I/14
inversion, 3 voices, no CS (mm. 20-23),

original, 3 voices, with CS1 (mm. 25-28),
original, 3 voices, with CS1, CS2 (mm. 29-32),
inversion, 4 voices, no CS (mm. 32-35), and only then
original, 4 voices, with CS1 (mm. 37-40).
The two halves of this fugue are largely analogous: both begin with a
subject entry without thematic companion followed by an entry accompanied by CS1 and rounded off by an episode with only M1. Both continue
with a third statement accompanied by CS1 + CS2. Only the ending of the
two halves differs.
The harmonic outline is straightforward. The four subject statements of
the first section are all in F minor (on the tonic, dominant, tonic, and tonic
respectively). The episode concluding this section modulates to the key of
the dominant. The five statements in the second section are again all in F
minor (on the tonic, dominant, tonic, subdominant, and tonic respectively).
In other words, the harmonic succession is the same in both sections, with
only the additional statement in section II on an unprecedented step.
In both sections of this symmetrically built fugue, the tension rises

gradually but constantly from entry to entry. The episodes E1 and E5 have
bridging function within a forward-thrusting direction, while E2 and E6
convey a receding direction, and E4 is conceived as an extension of the
preceding statement without any explicit change of color. Only E3 with its
cadential formula has concluding character.
WTC I/15 in G major Prelude

In the G-major prelude, the overall picture is dominated by 16th-notes.
They appear in triplet groups, with eight such groups to each 4/4 measure.
(Bachs twofold time signature in the first measure makes things look more
complicated and treacherous than they really are since it is obviously not
the treble part alone that features the triplets.)
These 16th-note triplets appear in three different grades of melodic
quality, providing indicators for the basic interpretation of the prelude. A
large portion displays broken chord patterns (see mm. 1-3 and 6-8 as well
as the bass part from m. 11 onward), suggesting a harmonically determined
composition. In other instances, they present a hidden two-part structure
(see mm. 4-5 and 9-10) from which melodically rather neutral peak note
lines emerge. In yet other measures, the 16th-notes appear in one-track
lines that are at least partially melodic (see the treble part from m. 11
onward). These measures allow small motifs to surface. The prelude thus
cannot be described with a single term. It sets out as a harmonically
determined piece from which tentative melodic features soon arise, developing more and more into independent entities.
The first harmonic progression concludes in the second half of m. 2.
As the pattern continues uninterrupted, this cadence should not be regarded
as indicating the end of the first section. This is reached with the subsequent abrupt modulation to the dominant, which brings with it a noticeable
change of pattern from m. 4 onward. (The caesura occurs only in U: m. 3.
The lower part crosses a section boundary at the end of the measure without phrasing). In the second section, a first miniature cadence resolving in
m. 5 does not interrupt the progression of the material, but the subsequent
harmonic close confirms the dominant key on the downbeat of m. 11. The
prelude comprises four structural sections altogether:
I
mm. 1-3
tonic confirmed, modulation to the dominant
II
mm. 4-111 dominant
III
mm. 11-131 dominant
IV
mm. 132-19 return to the tonic
No larger segment of the music recurs. Although the eye is caught by
the obvious similarity between m. 5 and its transposition in m. 9, the position of both measures in the structural context is entirely different: m. 5 is
211
212
WTC I/15
the second in a pair of measures in hidden two-part structure and represents a harmonic resolution, while m. 9 is the harmonically active first in
a pair.
The basic character is easy to determine as rather lively: This is
unambiguously suggested both by the simple rhythmic pattern with only
two note values and by the frequent broken chords (in the triplets) and
leaps (in the eighth-notes). Like the fugue it precedes, the prelude is downright virtuoso and should be played accordingly, in a very brisk pace. It
does not feature any ornaments. With regard to articulation one should
differentiate between the coarse level of simple contrastnon legato for
the eighth-notes and legato for the 16th-notesand the level of subtler
shades. These include: non legato in a light, neutral touch for the accompanying eighth-note leaps, a more expressive non legato for eighth-notes in
hidden two-part structure (where they move predominantly stepwise), quasi
legato with a crisp touch for the 16th-notes in the broken chords, legato in
a moderately expressive touch for the 16th-notes in hidden two-part
structure, legato with some melodic expression in the 16th-notes forming
motivic figures, and legato with melodic intensity for the eighth-notes in
U: mm. 11-13 representing appoggiatura-resolution pairs.
The G-major prelude can be regarded as consisting of two halves. Each
embodies a shorter introduction followed by a longer portion that develops
and complements the material established before. The first half spans mm.
1-111. It is based on a harmonic idea presented in the right hand with
virtuoso broken-chord patterns, accompanied by a rhythmicized pedal note
with octave displacement. In the first simple cadence (mm. 1-2), the
descending pitch direction of the broken-chord pattern is so dominant that
it overruns all considerations for shaping along harmonic lines. In character with this virtuoso figure, the prelude thus begins with a rather energetic
tone color followed by a diminuendo through two measures. The third
measure is composed as a harmonic sequence; it takes up the V7-I pair
from m. 2 and relocates it to D major. The equivalent to this feature in
performance is a sequence also in the dynamic process, i.e., a repetition of
the second half of the previous diminuendo. The question whether the
dynamic level in m. 3 is generally higher or lower than that in m. 2 can be
solved with the help of three observations: the pitch level is slightly lower
in the second half of m. 3 than in the second half of m. 2, the second half
of m. 3 features rests in the right hand part, the thinner texture adding to
the effect of decreasing intensity, and m. 3 was recognized as the end of
the first short section. A definite relaxation serves to underscore this fact in
performance.
G major
213
A new pattern arises from the last eighth-note beat of m. 3 onward.

Retaining the virtuoso quality of the beginning, mm. 4-5 change the
texture into that of a hidden two-part structure. In U: m. 4, a line with
stepwise motion in eighth-note rhythm builds the melodic foreground,
while the background consists of a pedal on D (the tonic of the newly
established key) ornamented regularly by its leading note C. In m. 5, the
foreground line continues while the background figure is transferred to the
left hand. (There should ideally be no audible difference between the two
measures.) The melodic part is further paralleled in the left hand. The subsequent measures return to the texture and material of the first measures
but vary it. Harmonically, mm. 6-7 and 8-9 recall the earlier sequences
with their V7-I. The chords appear extended to whole-measure durations
and feature varied accompaniment figures. It is important for the performance of this section to understand that the harmonic functions in the
leading measures of these pairs represent a sudden and powerful increase
in tension: the chord in m. 6 is vii7 of A (the dominant in D major), the one
in m. 8 follows as vii7 of E (the relative minor to the subdominant). The
next measures, mm. 9-10, return to the hidden two-part structure. The
pedal-note change between the two measures reveals the cadential steps:
E-A or steps ii-V of D major. The melodic parts of both voices set out in
parallels and only separate in the second half of m. 10.
The dynamic equivalent to the processes described above for the first
half of the prelude could be rendered somewhat like this:
mm. 1
2
3
4
5
6
7
8
9
10 11
+
+
+
+
f -mf mp -p mp p mp mp
pf
mf f
mp mp + mp
The second half of the piece, covering the remaining nine measures, is
based on motivic treble figures that introduce a new expressive quality.
Beginning with the first little motif (see m. 112-3, with three sequences up
to the E-D on m. 131), the melodically conceived 16th-notes followed by
the appoggiatura-resolution pair in the eighth-notes add charm and grace to
the prelude. (Note the dynamic shaping with U: crescendo to the appoggiatura, L: diminuendo from each strong beat onward. Dynamic imitations
would be ill-advised here.) The second motif (mm. 132-142) presents a
development of the first. It is twice as long, more elaborate in its triplet
figures, and the appoggiatura is now indirect (D-E embedded in D-E-F-E).
This motif also spawns sequences (mm. 142-152 and 152-162.) The accompaniment, which had contained old material in the first motif, combining
the broken-chord figure and the octave leap from mm. 1-4, is now much
more intricate as it sets out with the varied broken-chord figure from m. 6
and continues it freely, albeit in virtuoso fashion.
214
WTC I/15
The section ends with a figure of only one quarter-note length (see U:
m. 163), complemented in the left hand by a unit that, because of its
rhythm, has to be regarded as overlapping (m. 163-4 C-A-F-D etc.). Both
figures are sequenced many times, the treble right through to the end,
whereas the bass figure gives way to a cadential-bass pattern in m. 18.
The development of tension in this second half is determined entirely
by the direction of the sequences. Thus the first motif, which begins in
something like mf, is followed in its descending sequences by a diminuendo.
The second motif, which sets out with more vigor anyway, undergoes a
further crescendo in its ascending development. As if to compensate this
outburst, the third little figure, launched from a climax of less melodic
intensity than its predecessors, describes a continuous relaxation up to the
end of the prelude.
WTC I/15 in G major Fugue

The subject of this fugue appears confusing with regard to its length.
Asked where it ends one might find three possible answers, yet for one
reason or another, each of them is not entirely convincing. Performers who
regard the return to the keynote G (and with it, possibly, to the tonic
harmony) on the last 16th-note of m. 4 as the end of the subject must admit
that this conclusion is metrically not satisfactory. For performers who
decide to include the downbeat of m. 5, the subject ends with an imperfect
cadence, a solution that is harmonically uncommon. Performers who
interpret the subject as concluding on the downbeat of m. 4 will find it
both harmonically incomplete as it lacks the essential subdominant step
that, substituted here by its relative minor, materializes only in m. 4, and
melodically unconvincing as it ends on the fifth scale degree. Only in the
inversion does this choice of a subject ending sound very persuasive. Bach
must have felt the same, since he states the inversion repeatedly in this
shortened scope (see below).
With this complex answer and no simple solution at hand, it will prove
necessary to give each subphrase a name tag and then state very clearly
which of them appear in the statements of this fugue. With regard to the
first and second of the three concepts mentioned above, three subphrases
must be distinguished. The first subphrase [a] consists of the ascending
turn figures in m. 1, the subsequent eighth-note descent and the syncopated
seventh leap in m. 2. The second subphrase [b], confined to m. 3, features
G major
215
a varied partial sequence of the first subphrase: the eighth-note descent and
the syncopated seventh leap recur, but they are linked by a smaller interval
than before. The third subphrase [c] then consists of the downbeat
eighth-note in m. 4 and the twofold scalar descent (with or without the note
in m. 51). If one follows the third of the above-mentioned concepts and
assumes the closure of the subject already on the downbeat of m. 4, there
would only be two subphrases, [a] and [b], with [b] complemented by a
resolution.
The pitch pattern consists primarily of stepwise motion, with only three
leaps: two in m. 2, one in m. 3. Two of the leaps appear as consecutive
intervals (m. 2: G-D-C). Among the seconds, the longer note values, i.e.,
the eighth-notes in mm. 2-3, have melodic value while the shorter are
either ornamental, as is the case in the two turn figures at the beginning, or
constitute scalar passages, as happens in the two five-note groups at the
end of the subject.
In terms of its rhythm, the subject features three values: 16th-notes,
eighth-notes and (syncopated) quarter-notes. The same three note values
also constitute the predominant rhythmic pattern throughout the entire
composition. A particular attribute of the rhythmic pattern in this subject is
that it displays perfect symmetry in mm. 1-4. Were one to give a name tag
to the rhythmic patterns of each measure, the result would read x-y-y-x.
When analyzing the subjects harmonic background it seems worth
anticipating an important trait of this fugue, i.e., that the subject frequently
appears in inversion. The harmonic outline of this inversion is considerably
different from that underlying the original shape. The following examples
demonstrate this with harmonizations taken from mm. 11-15 and 20-24.
up to this point without subdominant
half cadence
perfect cadence
In our search for the intended climax we may thus not be able to rely
on harmonic features. However, both the rhythmic pattern and the pitch
outline provide congruous guidelines. Concerning the rhythm, the two
216
WTC I/15
syncopations obviously capture special tension. As both of them coincide

with high-tension leaps, there can be no doubt that the quarter-note C in m.
2 represents the climax of the first and the quarter-note E in m. 3 that of the
second subphrase. Between the two, the second is stronger than the first
because it is composed in an ascending sequence. After these two energetic
climaxes, the third subphrase appears as little more than an afterthought.
This may be one of the reasons why the doubt whether or not it belongs
survives. The dynamic tendency follows the descending motion; thus the
D at the outset of this subphrase is the beginning of a decrescendo.
The fugue comprises sixteen subject statements:
1. mm. 1-51
U
9. mm. 51-541
U*
2. mm. 5-91
M
10. mm. 52-541
L*3. mm. 11-151
L
11. mm. 604-63
M*
4. mm. 20-241
M inv
12. mm. 614-641
U*
5. mm. 24-281
U inv
13. mm. 694-731
L inv
6. mm. 28-311
L inv
14. mm. 77-801
M inv-orig
7. mm. 38-421
U
(15. mm. 78-79
L inv**)
8. mm. 43-471
M inv
16. mm. 79-82
U-
Several of the statements are incomplete in one way or another. Only

one is a genuine false start (indicated by parentheses in the table above).
Entries that are shortened at the end are marked with a minus, those that
are shortened in the middle carry an asterisk. Interestingly, the subject
abridgements in this fugue have persuasive reasons. The loss of the third
subphrase arises from the fact that it is perceived as an afterthought and
that it leads to a harmonically imperfect endingwhich, particularly after
the perfect cadence in the inversion, makes it harmonically superfluous. And
the compression of the first and second subphrases into one occurs only in
stretto positions, clearly intended to avoid the octave parallels that would
otherwise result.
There are not many other modifications. In two instances including the
false start, the beginning is delayed by a 16th-note and the subject thus
appears in a slightly altered rhythm (mm. 51 and 78). On another occasion,
a statement sets out as an inversion but later seems to be convinced (or
dragged along) by the entering stretto partner to change its ending into that
belonging to the original shape (mm. 77-80). Finally, one inverted subject
finds its closing note displaced an octave lower (mm. 72-73), and another
features an ornamental variation of the third subphrase (m. 46).
G major
217
There are three stretto passages in this fugue, occurring respectively in

mm. 51-54, 60-63, and 77-82. Parallels of whole statements do not occur.
However, a short but very effective three-fold parallel (two voices moving
in double thirds, another in the same rhythm but in contrary motion) is
created in the final stretto. Here the lower-voice false entry is expanded
with an extra sequence, and the middle voice statement that sets out in
inversion is lured by the upper-voice entry to mold its last subphrase into
the turn-figures known from the subjects beginning.
The fugue comprises only one counter-subject. Its rhythmic structure
allows a division into three segments, yet these structural components do
not function as separate subphrases but only as consecutive portions in the
pursuit of a single purpose. The scope of the counter-subject is slightly
blurred at its beginning, as was that of the subject at its end. Thus the F in
U: m. 5 can be regarded as the counter-subjects first note (if the subject is
understood as ending on the last eighth-note of m. 4) or as the point of
overlapping phrases, belonging at the same time to both elements. It is also
possible to perceive the counter-subject as beginning on the second eighthnote of the measure. Taking a closer look at the segments one finds that,
faithful to the requirements of contrapuntal independence of partners,
their confines do not coincide with those of the subjects subphrases. The
first segment [d] consists only of eighth-notes and ends on the downbeat of
m. 6. It shares with the beginning of the subject the pattern of half measure
model and ascending sequence; therefore its independence is limited. The
counter-subjects second segment [e] spans two measures. Moving exclusively in 16th-notes, it prolongs the previous ascent with an ornamental
figure sequenced at half-measure intervals. The third and final segment [f]
is scarcely one measure long. It begins with a scalar descent that sets out as
a parallel to the subject but then bends backward and reaches the high G,
thus disclosing its nature as yet another section under the overall purpose
of a peak note ascent. From the counter-subjects beginning to this target
note the underlying ascent thus progresses through an entire octave. This
directedness distinguishes it sharply from the subject which, in its concluding portion, returns to the level from which it was launched.
218
WTC I/15
The choice of an appropriate dynamic outline in the counter-subject is easy

since the tension rises perceptibly throughout. The opposite, a continuous
diminuendo, applies in the inversion of the counter-subject.
There are eight subject-free passages within this fugue.
E1
mm. 9-10
E5
mm. 47-50
E2
mm. 15-19
E6
mm. 54-604
E3
mm. 31-37
E7
mm. 64-694
E4
m. 42
E8
mm. 73-76
In addition, the five measures following the last subject statement (E9)
present the cadence that re-establishes the tonic (mm. 82-831) and a coda
on a tonic pedal (mm. 83-86).
The episodes are almost exclusively built from three independent motifs.
While the first two motifs are related and of only moderate melodic intensity, the third motif a truly lyrical quality and is thus the most conspicuous
among the three. M1 is first presented in U: mm. 9-101. Its rhythm is
characterized by uninterrupted 16th-notes that begin immediately after the
first beat of one measure and end on the downbeat of the next. The pitch
pattern is conceived in hidden two-part structure: one of the parts remains,
as an indirect pedal, on a repeated note, while the other part describes a
scalar motion. (The melodic part begins with a 16th-note upbeat, which is
often mistaken as belonging to the background.) In the original appearance
of m. 9 this scale falls, but it rises just as often (e.g., M: mm. 10-111); those
cases will be referred to as M1 inversions. The concluding note on the
downbeat serves as a harmonic resolution. In the original version, the resolution falls on the pitch level of the indirect pedal, whereas in the inversion
(mm. 10-11) it sounds as a continuation of the melody. M2 enters as a
companion to M1, with the original in M: mm. 9-101 and the inversion in
U: mm. 10-111. It is not polyphonically independent but composed as a
parallel to the melodic part of M1. Consequently its rhythm moves in
regular eighth-notes. M3 emerges only in the second episode (see U: m. 17
and sequences). Unlike the two earlier episode motifs it is launched from
the downbeat and ends somewhat indistinctly before the next bar line or on
the ensuing downbeat, in that case overlapping with the beginning of the
sequence. Its lyrical quality arises both from the melodic content of the
16th-note line and from the expressive syncopation on the second
eighth-note of the measure.
As to the dynamic outline of the three motifs, both M1 and M2 represent simple lines that follow the pitch direction: crescendo in all ascending
and diminuendo in all descending scalar motions. The distribution of these
motifs in the eight episodes of the fugue is very dense. Based on the
G major
219
material we can distinguish three episode types: E1, E4, E6, and E9 are all
built on M1 and M2, while E3, E5, E7, and E8 neglect M2 in favor of the
other two motifs. E2 is related to all of the others. It sets out with M1 in
non-motivic surrounding; then combines an M1 inversion with an M2
parallel, and finally presents sequences of the M3/M1 pair, complemented
by a neutral middle voice.
The relationship between the episodes is evident from the table above.
E1 is related to E4 and also, though less closely, to E6. E5 and E7 are
similar, and E8 shows analogy to the second segment of E3. E2 can be
divided to reveal relationships to E1 and E4 in its first two measures, to E5
and E7 in its remainder. None of the episodes exhausts its function as a
cadential close. In one instance, however, a closing formula determines the
material of a final segment: see the do-si-do figure in the upper voice as
well as the typical bass pattern in the first half of m. 69.
The role each of the episodes plays in the development of the composition also stems directly from its material. Those of the first type, based
only on M1/M2, act as bridges, while those ending with descending
M3/M1 sequences have concluding character. The two episodes featuring
the M1 + scale combinations lie in-between; they provide more contrast to
the primary material than the first type but lack any features of conclusive
force. One might thus say that they insinuate a gap that they nevertheless
span.
The basic character of this fugue is best interpreted as rather lively.
This decision is supported mainly by the pitch pattern with its leaps and
ornamental figures. The rhythmic pattern contains four melodically relevant note values, three already in the subject, and additional 32nd-notes in
the variation of M1, but all these values fit smoothly into a generally simple
rhythmic structure. The tempo is swift, playful if not outright virtuoso,
allowing for an energetic and non-superficial touch. The articulation
demands non legato for the eighth-notes and quarter-notes and different
kinds of legato for the 16th- and the few 32nd-notes. Among these, truly
melodic quality seems appropriate for M3 while an equally dense sound,
though for different reasons, suits the hidden two-part structure of M1.
Ornamental 16th-notes like the turn figures and scale portions in the
subject and the longer scales in E3 and E8 are best rendered in quasi
legato. An ideal proportion of tempo between the prelude and the fugue is
founded on the larger pulses of each: one quarter-note in the prelude corresponds with half a measure in the fugue. Approximate metronome settings:
80 for the beats (quarter-notes) in the prelude and the compound beats
(dotted quarter-notes) in the fugue.
220
WTC I/15
The fugue features a number of ornaments, namely in mm. 22, 25-26,

64, 69, and 78. Before pondering any of them, the general question of the
tempo of ornaments in this fugue must be settled. If they are to shake
twice as fast as the shorter note values, the dispute arises as to which of
the values, 16th-notes or 32nd-notes, are to be regarded as the shorter
rhythmic units. In accordance with both the virtuosity of the piece and the
ornamental nature characterizing many of the 16th-notes, it seems advisable
to regard these as the values to be doubled in the trills. The few 32nd-notes
will then appear as additional written-out slides and turns.
The ornaments can be grouped as follows: The mordent symbols in
mm. 22 and 78 designate long trills, as the spelled-out suffixes indicate.
The first, approached stepwise, begins on the main note and allows for five
trill notes (one 16th- + four 32nd-notes) before the suffix. The second
begins on the upper neighbor note and moves in regular 32nd-notes. The
two trills in the inverted subject statement are more problematic. Regarding
both pitch and metric position, the ornamented notes lead into possible
resolutions. Yet the notes providing these resolutions do not belong to the
same subphrases. Performers can draw three kinds of conclusions from this
predicament: (a) to ignore the original phrasing and link each trill, with a
proper suffix, to the subsequent note; (b) to ignore the trill symbol and
play the subject as it was introduced, i.e., unornamented; (c) to respect
both ornament and phrasing by playing a short, mordent-like embellishment of only five notes. The two trills in mm. 64 and 69 share the same
ending: they do not resolve on any appropriate strong beat but end in a tied
note. Both therefore conclude without a suffix and should stop before the
beginning of the tie. Convention suggests that the halt be as late as possible
before the last bar line; however, it is unlikely that Bach had this in mind
in the second case as it is technically impossible to maintain the trill
motion in the second half of m. 70 where the right hand is busy with the
counter-subject. A feasible as well as musically convincing solution is to
play both trills equally long. The one in m. 63, covering slightly less than
a measure, would serve as a model for the one in mm. 69-70, which would
then come to a halt before the middle voice enters. (By the way, these
interrupted trills do not, despite their beginning on the main note, prolong
the first note as they are not note-filling ornaments.)
m. 64
mm. 69-70
G major
221
Several indicators help in determining the structure. The order and

shape of the subject entries define some of the boundaries between sections.
It seems reasonable to assume that the first three statements in identical
shape belong to one section, that consequently the next three statements in
inversion establish the second section, and that the appearance of the first
stretto marks the beginning of a new section. Two subject statements sound
in reduced ensemble, namely those beginning in mm. 38 and 51. These are
therefore likely to constitute section openings. The cadential close in the
first half of m. 69 indicates the end of a section.
The role played by the episodes sheds additional light on the question
of structure. These facts confirm what has already been stated: E1 connects
the second and third statements of section I, and E2 rounds this first section
off at m. 201. E3, spanning a larger gap, connects the three inverted
subject entries of section II with the beginning of section III, marked by
the entry in reduced ensemble (m. 38). Sections II and III, while distinct
for several of the reasons stated above, thus form a superimposed group
owing to the episode that links them. E4 connects the two statements, one
original and one inverted, of the third section. The concluding E5 rounds
this third section off on m. 511, where the fourth section begins with
another entry in reduced ensemble. E6 connects the two strettos of the
fourth section. The concluding E7 rounds this fourth section off on the
middle beat of m. 69, additionally enhanced by the closing formula. E8,
spanning a larger gap, connects the single entry opening section V with
the following stretto. The analogy of E8 and E3 suggests that the fifth
section is to be regarded as structurally corresponding to the joint second
and third sections that, as shown above, form a group.
As these details reveal, structural correspondences in this fugue are
complex. The first two sections feature an obvious analogy in the number
of their entries and the consistency in which these are presented. The
impression is that of a double exposition: first, of the subject in its original
shape, then of its inversion. Yet another exposition occurs in the fourth
section, which establishes the strettos. The second large-scale analogy is
that between the second and third sections on the one hand and the fifth
section on the other.
The harmonic outline of the composition is straightforward. Both the
first and second sections remain on the tonic, the third section is in E
minor (the relative minor of the tonic), the fourth section in D major (the
dominant), and the fifth section returns to G major, approaching it from an
entry on the dominant and crowning it with the unusual statement on the
third of the home key (see U: mm. 79-82).
222
WTC I/15
In each section, the tension rises from the first to the last entry, due
either to an increasing number of participating voices (I, III, and IV), the
enhanced presentation of the subject (V: from single entry to stretto), or
the analogy with the first section (II). Section II is dynamically more
intense than the first since its three entries are launched in full ensemble
and follow one another without interruption. Section III falls back because
of both its reduced range and its minor mode, appearing almost as a softer
appendix. Section IV develops to an even higher dynamic level. Finally,
section V inverts the situation in the pair to which it is analogous (II/III . V)
by setting out from the softer level of the single entry and evolving in a
powerful increase toward the stretto with parallel tail.
WTC I/16 in G minor Prelude

This prelude is determined by three rhythmic figures. They come with
varying textures and pitch curves. The three figures jointly cover every
measure of the piece, with only one short interruption (of one beat length)
and two slight variations in one of the rhythmic patterns.
The first perfect cadence concludes at m. 31. This cadential close should
not be regarded as structural since the bass has so far not taken part in the
harmonic progression but remained on the tonic pedal. Also, Bach generates
a distinctly felt continuity between the patterns in mm. 1-2 and 3-4. The
modulation in the measures following this first cadential close reaches its
goal, a perfect cadence in the dominant D major, on the middle beat of m. 4.
The harmonic resolution is only a 16th-note long here, since the D-major
chord is immediately topped by its seventh and progresses to new regions.
This harmonic closure is thus again not the kind of caesura one is looking
for when trying to determine the preludes structural sections. The search
finally leads to the third cadential close, reached at m. 71 after a second
modulation to B major, the relative major key.
There are four sections in this prelude. Only the first is subdivided by
means of several structurally less definite cadential closes, as has been
shown above:
I
mm. 1-71
i-III
(G minor to B major)
II
mm. 7-111
III-iv
(B major to C minor)
III mm. 11-181
iv-i
(C minor to G minor)
IV mm. 18-19
i-i
(G minor confirmed)
No segment ever recurs, be it varied, transposed, or even in a freer
understanding of structural analogies. Any seemingly recurring measures
are just new manifestations of one of the motivic figures.
The rhythm is complex. In the melodically relevant segments of each
voice, the prelude features three different note values: eighth-notes, 16thnotes, and 32nd-notes. Faster motion occurs in the four long trills. In second
order there are occasional dotted eighth-notes and dotted quarter-notes, as
well as syncopations in passages with complementary rhythm. The pitch
pattern, too, is complex. A predominantly stepwise motion is regularly
interrupted by sudden leaps. Closer inspection reveals that these are cases
of hidden two-part structure, often with a pedal note in the secondary layer.
223
224
WTC I/16
Furthermore there are instances where two voices complement each other
to form one of the motivic figures, thus creating extra melodic intensity. In
view of this complexity, the preludes basic character should be interpreted
as rather calm, and played in a tempo restrained enough to allow for full
melodious substance in the 32nd-notes. The corresponding articulation
requires that all notes except cadential-bass patterns and consecutive leaps
be played legato. Cadential-bass patterns can be found in mm. 10-11, 1718, and 18-19. Consecutive leaps that need to be detached occur in the
lowest voice in mm. 5-6, whereas the seeming leaps in m. 14 form part of
a hidden two-part structure. They require legato between the two layers so
that each of the two voices sounds disconnected.
A tricky question arises in the area of articulation with regard to the
figure introduced in the upper voice of m. 2. The Gs clearly do not belong
to the main melodic line, but they do not form a secondary line of their
own either. Thus they can be interpreted, and played, in two ways. One can
regard them as a pedal background in a hidden two-part structure, in which
case they are not separated from their surroundings by articulation, but only
set apart by a different tone color. Or one can perceive them as charm
notes, the playful escape notes that were to become so common in the
music of the later 18th century. In this case they are treated as integrated
parts of the melodic line that sound just lighter and, because of their
interval gaps at both sides, non legato.
There are five ornaments, all of them marked as trills and all of them
appearing at a moment of harmonic affirmation (i.e., on the tonic chords of
the keys reached by way of modulation in the course of the piece). None of
these ornaments needs, or pretends to aim for, any kind of resolution. This
is most obvious in connection with the trill on the final note, but equally
true in the case of the four other ornaments, which all end in tie prolongation followed by the beginning of a new phrase. Having said this, it
follows that none of the trills concludes with a suffix. As to the beginning,
a launch from the main note is in order in the case of the trills that are
approached stepwise, as in mm. 3, 7, and 19, or begin without any preparation, as in m. 1. And although the context of the trill in m. 11 would allow
a different solution, it should probably match the corresponding ornaments
and also begin on the main note. The speed of the trill motion (twice as fast
as the shorter note values) is in 64th-notes, or four notes to each 16th-note.
In the four trills of one measures duration, the shake comes to a halt on the
last main note before the bar line, i.e., on the final 16th-note that is then
tied over. The duration of the shake in m. 19 is not fixed in any way; it
could be anything from an eighth-note upward.
G minor
225
The first of the three rhythmic figures that determine the piece is introduced in m. 1. Its texture consists of three parts, two of which, the upper
and the lower, are unmoving pitches: a trill in the treble and a repeated
pedal in the lower voice). The rhythmic pattern is even, with three different
values sounding simultaneously (eighth-notes, 16th-notes and ornamental
64th-notes). This figure recurs three times. In m. 3, the key and the order
of the voices are the same, with an octave transposition of the treble. The
eighth-note pedal, however, has been abandoned. A complementary pattern
of the middle and lower voices now provides both the eighth-note and the
16th-note pulses. In mm. 7 and 11, we find the figure transposed and in
inverted voices; the pedal has been redeemed and now appears as an
offbeat eighth-note pulse as part of the upper voice.
The second rhythmic figure defines the playful contour in the upper
voice of m. 2 that has already been mentioned. It takes various accompaniments as it recurs, varied not only with regard to its pitches but also in its
length and in details of its rhythmic design (see mm. 4-6, 8, and 15-16).
The third rhythmic figure first appears in m. 9. Its basic element spans only
a quarter-note. While it starts out in only two melodic guises, later recurrences (see mm. 12-154 and 163-19) show it varied not only in pitch but
also in texture. Thus its only constant is its rhythm.
With regard to the preludes layout one may state: Section I presents
the first two figures and subjects them to a first development. Section II
briefly takes them up in transposition before concentrating on the presentation of the third figure. Section III begins once again with the first figure
but then develops the third, only once interrupted by a reminiscence of the
second figure (mm. 154-16). In the vivid three-part texture of the concluding
segment, which sounds above a newly added tonic pedal, the third figure
reigns supreme.
The dynamic range within this rhythmically determined prelude is not
large. Small-scale dynamics depend exclusively on the overall pitch
direction defining each rhythmic figure. The melodically static first figure
suggests only the slightest increase in tension, while the third figure seems
always composed in superimposed descending pitch lines and thus appears
determined by dynamic relaxation. Only the second rhythmic figure
changes melodic orientation and, with it, dynamic direction, allowing for
diminuendo (as in mm. 2, 5, and 15-16), crescendo (as in mm. 6 and 8) or
a dynamic curve (as in m. 4).
With regard to the overall development of tension, the prelude
describes a soft, largely spaced dynamic curve. The second section implies
226
WTC I/16
more tension than the first but is surpassed by the third, while the coda
returns to the level of the beginning. Two climaxes can be determined,
falling on the downbeat of m. 9 and on the middle beat of m. 12. Both are
motivically connected with the third rhythmic figure but derive their
expressive power above all from their harmonic underpinning.
WTC I/16 in G minor Fugue

The subject of this fugue spans 1 measures. It begins on the second
eighth-note of the first measure and ends on m. 23.1 With regard to the
pitch position in the G-minor scale, it sets out on the fifth degree and
resolves into the third. The beginning with the second eighth-note of a
measure conveys an impression of upbeat. The downbeat of the second
measure, however, is taken up by a rest, so that the possibly strongest beat
is omitted in the melodic development. As a result, listeners are likely to
perceive m.13 as strong and the subject as consisting of three metric
units. In the course of the fugue, the metric position of the subject changes
constantly between a beginning after the downbeat and after the middle
beat. This confirms that at least in this case, Bach regarded the two metric
positions as equally strong. One can thus safely assume a hidden 2/4 time
behind the given notation.
Phrasing in the subject allows for two possible interpretations. The rest
in its middle can be taken for an interruption after which a new melodic
unit is launched. Alternatively, the rest can also be perceived as tensionsustaining; in this case the notes in m. 2 continue the process immediately
preceding the rest.
The pitch outline in the subject consists exclusively of seconds, with
the minor-sixth leap between the second and the third subject notes as the
only exception. This leap constitutes a high-tension interval. The rhythmic
pattern features three different note values: quarter-notes, eighth-notes, and
16th-notes. This rhythmic substance is confirmed throughout the fugue.
Harmonically the subject is very interesting. It does not begin on the tonic
but launches the cadential progression from a point of heightened tension:
1
The G on m. 24 does not belong to the subject. Three reasons support this: Melodically,
this note recurs only in one of the subsequent statements, in m. 6. Harmonically, the tonic
is already reached with B. Metrically, a subject that begins after a strong beat can usually
be expected to conclude on a strong beat.
G minor
227
The climax within this subject falls on

the fourth note (F in the original key).
All features join to support this event:
7
7
7
7
Melodically, the F is approached in
steps of increasing tension. The initial
fifth scale degree represents a higher level of intensity than would be felt
in the common first or third degrees. The ascending semitone reaches the
secondary leading-note and thereby increases the tension. Harmonically,
this note represents chord VI, the step that is most remote from the tonic.
From there the line plunges through the high-tension interval of the minor
sixth to the keynote and, continuing its direction, reaches F, the primary
leading note in G minor. Rhythmically, F is the first longer note value,
and metrically, as was explained earlier, it assume in the function of the
only strong beat.
After this climax, two slightly different dynamic renditions are possible
depending on ones choice for subphrasing. Performers who regard the
subject as consisting of two subphrases follow it with an abrupt and strong
decrease, letting the remainder of the phrase appear as an afterthought,
with a gentle increase toward a secondary climax on C and a final relaxation
thereafter. Performers who regard the subject as an undivided entity will
resolve the tension that was so powerfully built up toward the climax on F
gradually throughout the remainder of the subject. In this case, the C at
m. 22 is integrated into the line and not accented in any way.2
The fugue comprises sixteen statements and one false entry.
1. mm. 1-2
A
9. mm. 17-18
A
2. mm. 2-3
S
10. mm. 20-21
B
3. mm. 5-6
B
11. mm. 21-22
S
4. mm. 6-7
T
12. mm. 23-24
A
5. mm. 12-13
A
13. mm. 28-29
S
6. mm. 13-14
B
14. mm. 28-29
T
7. mm. 15-16
S
(15. m. 29
B)
8. mm. 17-18
B
16. mm. 31-32
A
17. mm. 33-34
T
Although many performers emotionally prefer interrupting rests, the harmonically more
consistent rendition is the one that resolves the tension of the main climax in a single
diminuendo. As the fugue progresses, more evidence arises to support the concept of an
undivided entity.
228
WTC I/16
In the course of the fugue, the subject suffers only minor modifications,
most of them inconsequential. Its answer is tonal, with its first interval enlarged from a semitone to a minor third (e.g., m. 2: G-B). The subjects
beginning is varied in m. 23 where the initial eighth-note is replaced by
two ascending 16th-notes. The final note, being the third scale degree,
appears as a Picardy third in m. 34, a common feature in a minor-mode
fugue. A modification that is more consequential can be observed when, in
the major-mode section from m. 12 onward, the two initial intervals are
changed: the semitone with its particular urging quality is substituted by a
major second and the leap appears as a major sixth that, unlike its minor
brother, is not a high-tension interval. These two changes have considerable influence on the buildup of tension.
In two instances does the subject appear in stretto: mm. 17-18 feature
a combination of bass and alto statements at a distance of four eighthnotes, and mm. 28-29 present a corresponding group of two complete
entries in soprano and tenor, fortified by a third entry at the same distance
in the bass that deviates immediately after the climax.
The fugue features a single counter-subject. CS acts as the subjects
faithful companion throughout the fugue and is only omitted in the final
statement. It begins in a metrical position equivalent to that of the subject,
but half a measure late, after the subjects climax. (This belated beginning
is the reason for the additional G on the fourth beat of mm. 2 and 6, a note
that does not form part of any material but whose function is to support the
beginning of the subject entry when all other voices are resting.) In pitch
pattern and rhythm, the counter-subject is strikingly related to the subject,
particularly to the its answer, which it almost seems to read upside down
and with its halves exchanged.
The counter-subject presents an indivisible phrase. The climax on the

sixth scale degree (B) unites similar characteristics as did the subjects
climax: it combines melodic tension on the secondary leading note of this
entrys D-minor key with harmonic tension on the representative of the
subdominant function and a metrical accent. Yet owing to its preparation
in a somewhat relaxed rhythmic pattern and stepwise motion, this climax
is considerably milder than that in the subject. The remaining four notes
bring a gradual subsiding of the tension.
G minor
229
S
or
CS
The G-minor fugue comprises six subject-free passages.

E1 mm. 4-51
E4 mm. 19-201
E2 mm. 8-121
E5 mm. 243-281
E3 mm. 163-171
E6 mm. 30-313
As the conclusion of the final subject statement marks the end of the
piece, there is neither a coda nor even a concluding cadential formula. All
episodes in this fugue are related to the subject; the only distinction here is
the degree of relationship and the appearance or omission of independent
(i.e., not subject-related) motifs.
It is generally true in all polyphonic compositions that the closest bond
between an episode and the surrounding primary material is achieved either
by way of a sequence or imitation of the preceding entrys ending, or by
way of an anticipation of the subsequent entrys beginning. In this fugue,
the former process is used frequently: E1, E2, and E3 are linked to the
statements preceding them through sequences of the final figure while E4
and E6 pick up this figure in imitation. E5 alone shows a subtle variation
of this pattern insofar as the figure is transformed, by a displacement of its
final note, into a bass line where it no longer appears entirely melodic but
conveys a hint of a cadential pattern (see mm. 24-27).
The only independent motif in this fugue, and the only component in
the entire piece that is not in some way related to the subject, occurs also in
E5. M1 consists of an ascending scalar section in 16th-note motion that is
complemented, after a syncopated halt, by a descent returning in the same
rhythmic pattern to the note of departure. In its original version (see A:
mm. 24-25), the ascent uses the second tetrachord of the melodic G-minor
scale, while the descent quotes the corresponding portion of the natural
G-minor scale. The dynamic outline follows the curve described by the
symmetrical design: a half-measure crescendo complemented by a matching
diminuendo.3 In the course of E5, M1 is followed by two sequences (see
M: mm. 25-26 and 26-27). Moreover, it is imitated, in inversion and
3
M1 has a forerunner that, consisting only of its ascending portion, appears in E2 (see the
soprano in mm. 8 and 9) and in E3 (see the tenor in m. 16-17).
230
WTC I/16
slightly varied, in the upper voice. This imitation should retain the dynamic
curve of the model.4
As can be seen from this overview, none of the episodes serves
exclusively as a cadential close. Only one episode segment fulfills this
purpose. In E2, the home key of the fugue is confirmed (motivically still in
the context of the subject-related figure) with a perfect cadence in G minor
on the middle beat of m. 10. The remaining 1 measures are neither
needed for the closure of the episode, nor do they present any (new or
continued) episode material. Instead this extension functions as a modulation to the related major key.
There is no structural relationship between the episodes of this fugue.
The role played by each episode in the development of the composition is
determined by the direction in which the motif derived from the subject
ending it sequenced. The following details can be observed: In E1 and E3,
the subjects final half-measure is taken up in ascending sequence. Both
episodes thus serve as bridges heightening the tension toward the ensuing
subject statements. E2 begins similarly with an ascending sequence of this
figure but continues thereafter in a generally descending direction (mm. 910: soprano B-A-G, bass E-B-D-G). This descent and the cadential close
it encompasses represent a relaxation. In E4 and E6, the final half-measure
of the subject is not sequenced but imitated. These imitations, though
ascending in pitch direction, appear as a more indirect continuation than
the sequences in the previous examples. Yet while they do not convey the
feeling of anticipation, they succeed in slowing down the decay of tension.
In both cases, however, the episode concludes with an inversion of the
figure and with a plunging final interval that generates a strong decrease.
Thus a gradual diminuendo ends in a steeper drop of tension.
The overall stepwise motion interrupted only for high-tension intervals
suggests that the basic character is rather calm. The rhythmic pattern seems
neutral. On the one hand, three different note values are regularly used
and, in the episodes, supplemented by syncopations; on the other hand, the
rhythmic structure does not seem complex enough to require real tranquility.
The result is a placidly flowing tempo.
4
The frequently heard mistake that renders the descent in the imitation in diminuendo and
the ascents in crescendo actually cuts the motif in two, and each half will automatically cry
for a new partner. The result is usually that middle and upper voice sound as if in a single
track, so that even the original motif in the middle voice is then destroyed. A good way to
avoid this pitfall is to play the leading voice (M) on a slightly higher level of intensity than
the imitating upper voice.
G minor
231
The corresponding articulation requires an overall legato suspended

only in cadential-bass patterns and consecutive leapsnot, however, in the
intervals of the subjects answer and the counter-subject, which derive
from essentially tension-rich intervals. There are no ornaments.
The relative tempo of the prelude to the fugue had best be taken in
complex proportion, both in order to avoid too much uniformity in the
closely related rhythmic patterns of the two pieces, and to bring out the
particular character of each. A good solution is one in which three rhythmic units of the prelude equal one in the fugue; i.e., three eighth-notes in
the prelude correspond with half a measure in the fugue. (Approximate
metronome settings: prelude beats = 44, fugue beats = 66.)
The most obvious indicator for the partitioning into sections is the
number of voices surrounding the subject statements. In two instances, a
subject entry is launched unaccompanied and thus announces without any
room for doubt that a new section is beginning. On another occasion, three
consecutive subject statements sound in reduced ensemble and are thus set
apart from the four-part texture of the entries preceding and succeeding
them. These observations reveal a design in four sections: mm. 1-121,
12-201, 20-281, and 28-34. This is supported by the cadential formula in
mm. 11-12, the consistent use of the major mode in all statements between
mm. 12 and 20, and the nature of E5, which expresses a concluding tendency in its pattern of descending sequences in all three voices.
The harmonic outline confirms this layout. The four statements in the
first section alternate between the tonic and the minor dominant. The five
statements in the second section sound in the major keys relative to tonic
and dominant. The beginning of the third section is marked by two statements representing the subdominant, while the third returns to the tonic.
Finally, all entries in the fourth section are built on the tonic.
Within section I, the tension increases through the first three statements;
the fourth, however, brings a set-back, due not only to the drastic gap in the
higher pitch range but also to the dropping out of the alto and the deceived
expectation of four-part texture. Section II pursues its increase of tension
consistently as the texture develops from two-part via three- and four-part
setting to the stretto statement. Section III retains both its texture and mode
unchanged and seems to feature nothing that would indicate a dynamic
increase. Conversely, section IV sets out with a stretto combining, in the
shortest possible span, the buildup from one to four voices and the overlapping imitation of a complete and an incomplete subject entry. After this
dynamic outbreak, the remaining two statements can only fall back: one
appears in three-part texture and the other without its counter-subject.
232
WTC I/16
The relationship between the sections is complex. Owing to the modification of the interval structure in the major-mode subject, section II begins
in reduced emotional vigor. Its concluding stretto may surpass the final
entry of section I in loudness but probably does not reach its passionate
quality. Section III, while returning to the melodic intensity of the minor
mode, is dynamically static and thereby falls back even behind section II.
After the additional color contrast in very soft shades brought about by the
material change in E5, the beginning of section IV then presents the
sudden climax. The fugue ends in considerable intensity.
WTC I/17 in A major Prelude

The A-major prelude is dominated by a single motif. It appears so
frequently that its absence is ultimately more striking than its presence. Of
a total of 44 measures there are only eight (mm. 16, 17, 33, 34, 39, 40, 43,
and 44) that do not feature the motif. The immediate sequence in m. 2 and
the imitation of model and sequence at the octave in mm. 3-4 give hints
that this prelude may be conceived as an invention.
The first harmonic progression closes at m. 91. The change of accompanying material indicates that this cadence is structurally relevant. As the
repeated appearance of D in subsequent measures signifies, the next
harmonic progression modulates. The target key, E major, is reached after
a cadential formula at m. 181. This cadence again marks the end of a
structural section.
I
mm. 1- 91
tonic confirmed
II
mm. 9-181
modulation to the dominant
III
mm. 18-351
modulation back to the tonic
IV
mm. 35-44
renewed confirmation of the tonic
Several portions in this invention-style prelude recur very faithfully, in
only slight transformation:
mm. 3-51 . 18-201 transposed
mm. 9-181 . 26-351 transposed, voices inverted
(the analogy is slightly blurred in the closing formula)
mm. 20-221 . 41-431 transposed
The rhythmic pattern is simple, based primarily on eighth-notes and
16th-notes. With regard to the pitch pattern, the principal motif consists of
a broken chord preceded by a written-out inverted mordent. Similar
combinations of leaps with ornamental 16th-notes can also be found in
other components of the motivic material. The basic character is thus rather
lively. The tempo should be fast enough to bring out the ornamental
quality of the 16th-notes, but not so fast as to deprive the eighth-notes of
their spirited character. The corresponding articulation is an effortless non
legato for the eighth-notes and quarter-notes, and a quasi-legato touch for
the 16th-notes. The only exception from this pattern of articulation occurs
in the longer notes in U: mm. 43-44 where A-G-A, as one of the typical
233
234
WTC I/17
closing formulas, must be legato. Within the quasi legato, various shades
of touch are required in order to bring out the different textures. Passages
in hidden two-part structure with a melodic line moving before a backdrop
of a repeated pedal (as in U: mm. 15-16, L: m. 32, and U + L: mm. 39-40
where weightier and lighter notes alternate regularly) should be distinguished from those with an ornamented pedal (mm. 13-14 and 30-31 where
the structural notes are interspersed with feathery inverted mordents) and
from the remainder of the 16th-note passages in which all notes take part
equally in the melodic line.
The prelude contains two kinds of ornaments, indicated by the same
mordent symbol. The two closing formulas in mm. 17-18 and 34-35 each
feature the typical interrupted (point darrt) trill with anticipation of the
resolution note. Both begin on the upper neighbor note and shake in
32nd-notes. The other four mordent signs, appearing on the quarter-notes
in mm. 36 and 38 and on the 16th-notes in mm. 41 and 42 respectively, can
be played as just that: three-note mordents that begin on the main note
either because they are approached stepwise, as in mm. 36 and 38, or
because they embellish the initial note of a phrase, as in mm. 41 and 42.
or
mm. 17-18
mm. 34-36
The principal motif dominating this invention-type prelude, M1, spans

one measure. The downbeat rest on which it ends is later often replaced by
a note. Within this metrical structure, the two 16th-notes serve as an upbeat
to the four eighth-notes. The dynamic presentation should outline this with
a little crescendo to the climax on beat 2 and a successive diminuendo to
the end, regardless of whether the downbeat appears as a rest (as in mm. 2,
4, 5 etc.) or as a final note (as in m. 3 and mm. 6-16).
In the course of the composition, M1 appears twice with strictly polyphonic counterparts and twice with accompaniment patterns. Initially it is
counteracted by full chords, double notes, or octave leaps. As representatives
of homophony in an essentially polyphonic composition, all these components should be kept in the background: played in neutral tone color,
considerably softer than the motif and, wherever they appear extended (as
e.g. in U: mm. 3-4), with beat 2 particularly light. The same applies in the
chordal companions variation in mm. 36 and 38. Another homophonic
accompaniment figure is heard in mm. 13-14 and 30-31. It, too, should be
kept in low profile, while the scalar descent in the hidden two-part structure
A major
235
in mm. 15 and 32 creates a diminuendo that contrasts with the dynamic

curve in M1. The other two settings show the principal motif accompanied
by counter-motifs, i.e., polyphonically independent figures. CM1 appears
in mm. 9-12 and 26-29. Its model features a 7/16th-note upbeat followed
by a long downbeat; in the sequences, the downbeat note is replaced by an
ornamental figure. In all cases, the upbeat leads in crescendo to the downbeat and is then followed by a decrease through to the fifth 16th-note. CM2
also begins with an upbeat in 16th-notes (L: mm. 20-22 and 41-43). The
climax on the downbeat is followed by a descending octave leap providing
the relaxation. Yet another companion appears only transitorily. In L: mm.
22-241 and U: mm. 24-261, M1 + sequence are contrasted with descending
zigzag figures in continuous diminuendo.
The preludes design comprises two basically analogous sections, a
short middle section and a coda that recalls material from the middle section.
The analogous sections appear in the harmonic progressions typical for
baroque pieces featuring a recapitulation: the modulation from the tonic to
the dominant is answered by one from the subdominant back to the tonic.
I A mm. 1-9 (I)
III A' mm. 18-20 (V)
B mm. 9-18 (I V)
B' mm. 26-35 (IV I)
M1 against CM 1, then M1 with accompaniment figure,
concluded with a cadential formula
II C mm. 18-26
coda mm. 35-44
material from A,
material from A (varied),
followed by M / CM 2
followed by an accomp.
and M / transitory figure
figure in parallels, M / CM 2,
and a final cadence.
The tension in each section is determined by the extent of polyphonic
contrast, the direction of the sequences, and the harmony.
WTC I/17 in A major Fugue

Beginning on the second beat of the first measure, the subject of the
A-major fugue conveys the impression of starting with an up-beat. The
conclusion after only a single measure therefore comes as a surprise, as
one is still expecting the main content and is now challenged to believe
that this was already it. The principal idea of this fugue comprises seven
regular eighth-notes in all. (In the course of the fugue, the final note alone
comes in a great variety of values, from the 16th-note in m. 29 to the
236
WTC I/17
dotted half-note in m. 6.) The sensation of incompleteness in this subject is

enhanced by the pitch line, which features the rather unusual closure on the
fifth scale degree, and by the abruptness of the harmonic progression.
This short subject is definitely conceived as an indivisible phrase. Its
pitch pattern consists almost exclusively of leaps in broken-chord setting.
The rhythmic structure is of utmost regularity in the subject itself. The
same does not, however, hold true in the remainder of the fugue, where the
eighth-notes appear mixed with 16th-notes, quarter-notes and several
syncopated values.
The harmonic background to this subject
is somewhat peculiar. The subdominant is
reached in metrically weak position on m. 14,
followed in m. 21 by the dominant-seventh
chord that resolvesmelodically inerton
the tied 16th-note of beat 2 (see e.g. the harmonization of the statement in
mm. 5-6). This implies that whenever the final note of a statement is
shortened, the resolution may be cut off from the subjects main body. This
is already the case in the answer, where the bass is way into the partial
sequences following this entry when the E-major resolution of the B7
chord appears.
The dynamic outline poses a problem, despiteor perhaps because of
the subjects short range and straightforward character. On the one hand,
the significant features, which are (harmonically) the subdominant and
(melodically) the highest pitch F, coincide; but they fall on a weak beat.
On the other hand, the final note, which in the original version holds
rhythmic importance through its sudden length, appears so frequently
shortened that it may not be entirely convincing as a climax either.
The fugue contains fifteen subject statements.
1. mm. 1-2 T
9. mm. 23-24 A
2. mm. 2-3 B
10. mm. 24-25 S
3. mm. 5-6 S
11. mm. 27-28 B
4. mm. 6-7 A
12. mm. 28-29 T
5. mm. 10-11 T
13. mm. 29-30 A
6. mm. 13-14 A
14. mm. 30-31 S
7. mm. 17-18 T
15. mm. 33-34 S
8. mm. 18-19 A
A major
237
The subject never appears in inversion, stretto, or parallel.1 Leaving

aside the rhythmic variations in the final note and the adjustment in the
answer (the lowering of the second subject note in mm. 2 and 6), two
interval modifications repeatedly, both with considerable harmonic impact.
In mm. 18, 28, 29, and 30, an initial fourth interval is followed by the
broken triad of m. 1, thus mixing features from dux and comes and
resulting in harmonic ambiguity. More strikingly, the first three of these
entries, all of them composed in the minor mode, borrow several notes
from their major-mode relatives (see m. 18: C + A for C + A; m. 28: D + B
for D + B; m. 29: G for G). Two statements feature a seventh leap instead
of the original sixth between the fourth and fifth eighth-notes. The result is
a seventh chord that re-casts the subject as modulating (see mm. 232-241:
from E7 to A, and mm. 242-251: from A7 to D).
This fugue does not comprise a single consistent counter-subject. The
16th-note figure accompanying the subject in mm. 2-3 recurs frequently
and in much variation throughout the fugue but is not linked as a
companion to the subject. Instead, it provides motivic material that, much as
happens in the E-major fugue from the WTC I, pervades the entire fugue.
The composition encompasses eight subject-free passages.
E1
mm. 3-5
E5 mm. 19-23
E2
mm. 7-10
E6 mm. 25-27
E3
mm. 11-13
E7 mm. 31-33
E4
mm. 14-16
E8 mm. 34-35
The material that characterizes these episodes also occurs outside of
them, overlapping considerably with the subject but never serving as its
dedicated companion. All of it seems, in one way or another, derived from
mm. 2-3, i.e., from the figure sounding against the subjects answer and
from the first episode. Three figures serve as sources for the secondary
material: the eight 16th-notes following the original subject statement,2 the
1
Only once, in an episode where the combination of interval and rhythm pattern suggests
false entries, does a brief parallel occur; see mm. 214-221 alto/soprano.
2
Introduced in m. 22-4 (T), M1 recurs, with the opening note displaced to form an ascending
scale, in m. 4 (B). In rhythmic variation with intersected syncopation it is also found in m.
27 (A, imitated in S). The inversion, M1a, appears in mm. 24-32 (T), 44-52 (B), 153-161 (A),
223-231 (A), and 251-3 (T, with two sequences). M1b, an extended version combining the
ascending scale with the M1 inversion, is found in mm. 53-62 (B), 72-81 (A), 82-91 (B),
92-101 (B), 232-241 (T), and 242-251 (T). An even more extended version materializes in
mm. 291-302 (T). M1c, the (free) inversion of the previous motif, occurs twice, in mm.
16-17 (S) and 18 (B). As a further development from the same root, the scalar ascent is
complemented, after a syncopation, by a new tail in eighth-notes: for M1d see mm. 7-8 (B),
8-9 (S), 9-10 (A; here it gives way to a melodic closing formula), and in mm. 16-17 (B).
238
WTC I/17
four eighth-notes extending the answer in the pattern of a partial sequence,3

and the syncopated half-notes that, with an ornamental variation, sound
against this partial sequence.4 The episodes that are most closely related to
the subject are those containing M2, i.e., E1, E3, E4, and E5. More remote
variants of the motif in E6 and E7 also convey an impression of relatedness.
M1
M1a
M1b
M1c
M1d
M1e
M2
or
M3
The only subject-free passage that serves as a cadential close is E8. In

addition there is one instance within another episode where the motivic
display is suspended and gives way to an extended cadential formula (see
mm. 20-22, at the end of E5). Two further cadential closes appear integrated into the motivic material; both feature the typical do-si-do figure
(see mm. 9-10 at the end of E2 and mm. 15-16 at the end of E4).
The relationship between the episodes is already evident from the
exposure of their motivic content. The first half of E4 and the entire E5 are
both direct variations of E3; similarly, E7 is a (more distant) variant of E6.
Only two episodes maintain the tension, thus serving as bridges between
statements: in E1, a short descent is followed by rising motions, and in E7,
the ascending motion in all voices of mm. 31-32 is complemented by a
relaxation of roughly equal length. All other subject-free passages display
a pattern of gradually decreasing tension, either because of apparent
3
Introduced as a two-fold partial sequence of the subject, this motif is first heard in B: mm.
3-4 above an E pedal (see E-B-G-A, E-A-F-G). It recurs as a separate motif, with
occasional ornamental variation, in mm. 11-13 (B), mm. 14-15 (A) and mm. 19-21 (S).
Remote relatives can be heard in mm. 25-26 and 31-32. A final remote relative combines
the scalar ascent with a subsequent descent. This figure appears several times, with a
varying number of free sequences; see mm. 11-13 (S), 14-15 (B), 19-20 (A) and 31-32 (B).
4
First presented in the context of M2, this motif could, in fact, be read as a parallel to the M2
peak notes. The consecutive syncopations recur in mm. 11-13 (T), 14-16 (S) and 19-21 (T).
A major
239
descending sequences (as in E3, E4, E5, and E6), or in hidden descents (as
in E2 where the soprano features the falling A-major scale).
With the exception of the syncopated half-notes in M3, the rhythmic
pattern in this fugue is simple. Moreover, the pitch pattern is made up of
broken chords alternating with ornamental figures. It is thus easy to
determine the basic character of this fugue as rather lively (or even very
lively). The tempo is confined by the character of its primary material: the
broken-chord leaps in the subject should sound vigorous and thus not too
slow, while the various 16th-notes figures should by no means appear as
mere virtuoso patterns, particularly in view of the manifold motifs Bach
invented and developed, and thus not too fast. The corresponding articulation consists of bouncing non legato in the subject and M2, a fairly
intense quasi legato in all motifs deriving from M1, and a melodious non
legato in M3. The only longer note values that must be played in strict
legato are the cadential formulas in S: mm. 15-16, 22-23, and 34-35, and in
A: mm. 9-10. The fugue does not feature any ornaments.
The relative tempo of the swifter and lighter prelude in triple time to
the more vigorous fugue in quadruple time is best established by equaling
metric, and not rhythmic, values: an ideal proportion is one in which one
measure in the prelude equals half a measure in the fugue. (Approximate
The most obvious indicators that help determine the design of this
fugue are the cadential formulas in mm. 9-10, 15-16, and 22-23. As the
subject entries preceding the first of these closures constitute a perfect
round of all four parts (T B S A), the first section seems thus ascertained.
Between the cadential closes in mm. 9-10 and 22-23, the correspondence
of E3 with the first segment of E4 and E5 creates a larger unit from m. 11
to m. 23. The reduction of the ensemble from four to merely two voices in
mm. 23-24 further enhances the structural importance of this harmonic
closure. Inside this superimposed frame of twelve measures, the closing
formula in the middle of E4 generates a smaller caesura, confirmed in its
structural value by the fact that the ensuing subject statement sounds in
reduced ensemble. As no explicit cadential close occurs between m. 23 and
the end of the fugue, one has to look for other indicators of structural
layout. One hint can be found in the entering order of the voices. The five
final statements (B T A S S) seem to form a group: they follow one another
not only very closely and in a logical arrangement of gradual ascent, but
also lead, after the tension-increasing E7, to the conclusion of the fugue in
a redundant soprano statement. The return to the home key with the bass
entry in mm. 27-28 further ascertains the confines of the final section.
240
WTC I/17
A major
241
The harmonic outline as established in the subject statements and

cadential closes confirms the structural layout deduced above. The entire
first section remains in the tonic key of A major, which is substantiated by
the cadence in mm. 9-10. The second section features entries in the tonic
and its relative minor, and concludes correspondingly with a cadential
close in F minor (mm. 15-16). The third section sets out in B minor and,
after the harmonically ambiguous second statement (B minor/E minor)
reaches the dominant key (E-major cadence, mm. 22-23). The fourth
section is characterized by the two modulating subject entries. Progressing
from E via A to D, this section approaches the home key from the
subdominant region. The beginning of the fifth section, as has already been
mentioned, marks the return to A major which, though weakened in the
harmonically hybrid tenor and alto statements, is not abandoned again.
The five sections of this fugue present very individual faces.
The first section comprises the usual build-up of tension created by
the gradual increase of the ensemble. While the bridging episode
E1 describes a concave curve, with a slight relaxation followed by
a new rise preparing the second pair of statements, the sectionclosing episode E2 with its descending lines brings gradual
relaxation.
The second section begins in the tonic key and in four-part texture,
but its second entry turns to the minor mode. Thus there is a
decrease in tension between the two statements of this section. The
two episodes enclosed in this section enhance this tendency of
relaxation. After the cadential close in the middle of E4 the two
remaining voices trail along somewhat indecisively.
The two subject statements of the third section with their harmonic
ambiguity are not made to bring forth powerful impulses either.
The episode E5, which one more time recalls the descending
sequences already heard in E3 and E4, blends well into this picture
of no news, no emotional features. The explicit cadential formula
thus appears as the closure of a protracted decline, summing up
sections I, II, and III as one large portion.
The fourth section begins with the lowest number of voices found
in a statement outside the initial one and is launched at a correspondingly reduced level of intensity. However, the harmonic
design of the two modulating subject statements, together with the
fact that they bring forth the return to the home key of the piece,
endows this section with considerable urge.
242
WTC I/17
Although mollified transitorily during the episode, this urge continues even more powerfully in the four consecutive statements that
mark the beginning of the fifth section. The ensuing episode E7,
interestingly enough, is the only tension-sustaining one apart from
the very first episode. After its deceptive cadence (m. 332-3), the
final statement sounds in an almost homophonic setting, crowning
the fugue with a triumphant closure.
WTC I/18 in G minor Prelude

The G-minor prelude is composed in strict polyphony. With the
exception of the familiar voice splitting in the final chord and some irregularities of voice leading in the three initial measures, it is devised in
consistent three-part texture. The predominance of a single motif (see U:
mm. 1-21) and the initial imitation at the octave characterize this prelude as
a three-part invention.
Once the home key G minor has been established in two brief progressions (mm. 1-21, 2-31), the first modulation occurs in mm. 4-5 with the
cadential steps IV-V-I of B major, the tonic relative. As both the imitation
of the motif and the first episode have by now been introduced, we should
regard this cadence as the conclusion of the first structural unit. Corroboration comes with the subsequent motivic statement which, reminiscent
of section beginnings in a fugue, appears in reduced ensemble, i.e., with a
rest in the lower voice. At the same time, the fact that listeners expect the
third entry of the principal motif in the middle voice and that this expectation is met at the beginning of the next harmonic progression threads the
two passages together to a larger unit, i.e., section I.
The question where the second section ends is trickier. One could tie
this decision to the ensuing return to the tonic, and thus to the perfect
cadence concluding at m. 101. This interpretation can be supported by the
observation that the subsequent motivic statement is the first in the piece to
appear in inversion. Alternatively, one could base ones reading on the
cadential close in m. 91, and thus take the progression as leading to the
subdominant area. This interpretation can be supported by an observation
on the structural level: The next section would then begin, as did the first,
with two statements followed by two measures laid out as a variation of
mm. 3-4. As this is the only instance in the prelude where a passage is taken
up with recognizable resemblance albeit with exchanged voices, it seems
important enough. The correspondence sheds light on the analogy between
the beginnings of the first and second sections and, if one decides to take
this as significant for the interpretation of the preludes structural design,
even invites a similar conclusion for the beginning of the third section.
Depending on the alternative explained above, the layout of this threepart invention can thus be read in two quite different ways, as the
following tables show:
243
244
WTC I/18
I
II
III
IV
mm. 1-101 (1-51, 5-101)

mm. 10-181 (10-131, 13-181)
mm. 18-271 (18-221, 22-271)
mm. 27-29
tonic, relative major, tonic

tonic, minor domin., subdom.
subd., tonic, tonic confirmed
tonic
or:
I mm. 1-91 (1-51, 5-91)
II mm. 9-171 (9-131, 13-171)
III mm. 17-29
tonic, relative major, subdom.

subdom., minor dominant, tonic
tonic confirmed
The rhythmic pattern of this prelude is basically simple. This suggests

a rather lively character. At the same time, the slightly subdued spirit
inherent in many minor-key compositions is reinforced here by the specific
melodic structure of the principal motif, which peaks on the minor sixth
and falls back to the minor third. To accommodate both traits, the tempo of
the prelude should be fairly swift, without conveying the impression of
energetic activity in the eighth-notes. Instead, a desirable effect is that of a
gentle swing in half-measure pulses. An articulation that matches this basic
character demands legato for the 16th-notes and non legato for the longer
values. In order to express the nuances hinted at above, the non legato in
the principal motif and all figures immediately deriving from it should be
very delicate. In the larger leaps (M: mm. 3-4 etc.), the separation can be
slightly more pronounced, and a distinctly detached style is required in the
cadential-bass patterns. Exceptions occur in the three dosido formulas
demanding legato (L: mm. 1-2, M: mm. 13-14, U: mm. 2-3).
The score comprises only one ornament, which is represented by a
mordent symbol (see U: m. 13). As the note to be ornamented resolves
duly onto the following downbeat, this ornament must be interpreted as a
note-filling trill. It begins accordingly on the upper neighbor note, shakes
in 32nd-notes and concludes with a suffix (D-E).
The prelude builds entirely on the principal motif and its inversion.
What appears as a kind of counter-motif (see, e.g. M: mm. 1-21) recurs
only in m. 2 but never again thereafter. The principal motif is thus solely
responsible for the character and development in the prelude.
The episodes (or motif-free passages) also derive their material from
the same source. The first episode type makes use of the 16th-note group
from the motifs head, transposing its second three-note group down a
fourth and sequencing this newly-assembled figure twice in descending
direction (see U: mm. 3-4, and L: mm. 11-12). The second episode type
G minor
245
presents the same figure (L: mm. 22-23) along with the inversion of the
original 16th-note group (M: m. 25, U: m. 26) and a version of the group
in its first shape but with a broken triad at the end (L: mm. 19-21). In
addition to all these transformations of the motifs head, this episode also
recalls the motifs tail (mm. 19-21, in imitation between U and M).
In order to determine the dynamic processes in this prelude, one may
wish to distinguish a varying intensity among the statements of the motif.
Intensity is created here not only, as is generally true for all polyphonic
compositions, by the number of voices surrounding the statement or the
mode in which it is set, but also by the nature of the accompanying material.
In this prelude, Bach seems to use parallels as a characteristic means. The
following hierarchy can be deduced from this observation:
The lowest level of intensity is expressed in statements quoting
the motif without any doubling (mm. 9 and 10, 15 and 17, 27
and 28).
Slightly more emphasis is created by the parallel of the motifs
tail (mm. 1, 2, and 7); less, where the doubling is set in contrary
motion (as in mm. 8 and 16).
The fortification of the 16th-notes alone generates even greater
density (mm. 5, 6, 14 and, somewhat more indirectly, m. 13).
This last level of intensity is exceeded only once where both
halves of the motif are doubled. In m. 18, the 16th-notes appear
in parallel and the eighth-notes are matched in contrary motion.
Relating these observations to the structural features of the prelude one
detects that
the endings of all sections appear in lessened intensity;
in the first and second sections, the climax occurs immediately
after the diminuendo resulting from the episodes descending
sequences and the mid-section cadence;
the first section begins in moderate intensity, with two statements, while the second section, which introduces the motifs
inversion, sets out with only one statement of low intensity;
the third section contains only one complete statement of the
motif that, appearing at the beginning of the section, exhibits
the highest level of intensity in the entire piece;
the coda returns to a very relaxed state.
Here is a graphic representation of the invention, in the second of the
two readings offered above, highlighting the appearance of structurally
analogous sections.
246
WTC I/18
WTC I/18 in G minor Fugue

This subject encompasses two measures. Beginning on the second beat
of m. 1, its upbeat character is attenuated by the length of the first note,
which seems to stand still rather than lead anywhere. After a modulation to
the dominant key, the subject ends with a figure very unusual for a melodic
line: a cadential-bass pattern (see G-G-A-A-D).
There are two subphrases. The first, ending on D (m. 22), is characterized by rhythmic variety (a quarter-note, several eighth-notes, and two
16th-notes) and an emotional pitch line (two leading-notes: F-G and
C-D, as well as one high-tension interval: G-C). The second subphrase
comprises only regular eighth-notes in a line that is melodically very low
key. Phrasing is thus not determined by structural features (like sequences)
or pitch level in this fugue, but by a drastic contrast in melodic intensity.
The pitch outline in the subjects first measure displays only stepwise
motion, circling around the key note. The return to G at m. 21 is followed
immediately by a high-tension interval, represented here by a tritone leap
G minor
247
to the artificial leading-note of the fifth scale degree. The resolution of this
leading-note marks the end of the first subphrase. The ensuing major-sixth
leap is thus not a melodic interval but rather a split between the two subphrases. The second subphrase comprises whole-tone steps and a perfect
fifth. Conspicuous note repetitions on the fourth and fifths scale degrees of
D minor enhance the impression of a cadential-bass pattern. Having
observed this one understands why these notes seem to convey so little
melodic message. They are harmonic notes (as representatives of the chords
they imply), and their significance is vertical rather than horizontal.
The subjects harmonic background is most intriguing in a segment
where it might be least expected: the eighth-note-descent B-A-G. The
melodic return to the keynote at m. 21 is not matched by a similar return in
the harmonic progression. On the contrary, after an initial alternation
between tonic and dominant (G minor and D major, with or without their
seventh) in m. 1, this downbeat represents the harmonically most active
step in this phrase, i.e., that to the secondary dominant triggering the
modulation. The melodic resolution of the artificial leading-note C thus
coincides with the harmonic resolution of the diminished chord into the
new tonic. The freshly established key is then confirmed in a cadential
progression.
The subjects main climax occurs in the first subphrase. The melodic
and harmonic developments both favor the first beat of the second
measure, where G-C constitutes a high-tension interval while at the same
time representing the pivot chord of the modulation. The end of the first
subphrase on the measures third eighth-note provides the resolution to
both the harmonic tension and the melodic leading-note, thus generating a
steep decrease after the gradual increase during the first measure. By
contrast, the dynamic outline of the second subphrase is very gentle. The
first note (F) serves as an upbeat to the subdominant representative G,
which is followed by an even relaxation. Although the subjects two subphrases could thus hardly be more contrasting, they nonetheless complement one another symmetrically with regard to tension: the gradual rise
through most of the first subphrase is answered by a gradual decay through
most of the second, the sudden release after the expressive first climax
finds its counterpart in the concise upbeat to the secondary climax.
248
WTC I/18
The subject appears altogether twelve times.
1.
mm. 1-3 T
7. mm. 17-19
2.
mm. 3-5 A
8. mm. 19-21
3.
mm. 5-7 S
9. mm. 24-26
4.
mm. 7-9 B
10. mm. 26-28
5.
mm. 11-13 T
11. mm. 32-34
6.
mm. 15-17 B
12. mm. 37-39
T
A
S
B
T
S
Apart from the alteration of the first interval in the answer, no modifications of shape or length occur in the subject throughout the fugue.
Inversions, parallel statements, or strettos are not used.
Bach invented two companions to the subject that appear repeatedly.
However, just as the subject itself displays a somewhat unusual melodic
line in its second subphrase, the counter-subjects also do not quite conform
to ordinary expectations of polyphonic counter-parts. CS1 is introduced
against the second subject statement in mm. 3-5. It begins with an upbeat
(which is later occasionally dropped or varied). The following long note
and particularly the ascending groups on the sixth and seventh eighth-notes
of the measure sound so much like a parallel to the beginning of the
subject that they may hardly pass as contrapuntal. Only the middle segment with its sequences of the ascending figure leads a polyphonically
independent life (T: m. 3-4 F to B), while what follows thereafter qualifies
as a very traditional closing formula (G-A-F-G). Notwithstanding this
limited independence, CS1 accompanies almost all subject statements, with
exceptions occurring only in the entries in mm. 24-26 and 26-28, where
CS1 is omitted altogether, as well as in the final entry, where its beginning
and end are considerably varied. CS2 is also first presented in its expected
place against the third subject statement (T: mm. 5-7). Its characteristic
features are the syncopated fourth leap and the descent in quarter-notes
concluding in a tie. The beginning as stated in m. 5 with upbeat eighthnote and quarter-note (this value again in rhythmic parallel to the subjects
initial note) is later dropped or varied, just like that of the first countersubject. The suspension created by the tie at the end of CS2 redefines the
harmonic surroundings of the subjects final note as still awaiting resolution. Statements accompanied by this counter-subject are thus unlikely at
section endings. CS2 recurs four times, in mm. 11-13 (A), 15-17 (S), 19-21
(B), and 32-34 (S).
G minor
249
The phrase structure in the two counter-subjects is worthy of closer

inspection. While the way in which the material is connected parallels the
subject: a melodious segment followed by a traditional formula, the
segments in CS1 are strung together in such a manner that they build a
single curve without any need for partitioning. One step further, CS2
consists of an indivisible unit with regard to both structure and material. In
terms of dynamics, neither counter-subject engenders a tension process S
independent of the one
found in the subject;
CS1 meets the subjects CS1
second (weaker) climax,
while the peak of CS2
coincides with the sub- CS2
jects first (stronger)
climax.
There are six subject-free passages in this fugue.
E1
mm. 9-11
E4 mm. 28-32
E2
mm. 13-15
E5 mm. 34-37
E3
mm. 21-24
E6 mm. 39-41
The material of these episodes comprises two salient features. The first,
very unusual in Bachs fugues, is a homophonic formula (see mm. 9-101),
made up of the cadential-bass pattern from the subjects second subphrase
in the bass, a slightly extended version of the second segment from CS1 in
the soprano, and chordal filling notes in alto and tenor (which will later,
from m. 13 onwards, be substituted by the last two notes from CS2). The
other conspicuous feature (see mm. 21-23) is polyphonic in design; it is a
motif arising in an imitative setting.
The distribution of these two components of secondary material in the
episodes is very straightforward:
E1
formula + sequence (ascending)
E2
formula + sequence (descending)
E3
motif in B/A + sequences (ascending)
E4
motif in A/S + sequence (ascending) = E4a
formula + sequence (descending) = E4b
E5
formula (varied) + sequences/imitation
E6
formula (varied) + cadential closing formula
As can be seen from the table, there are several relationships among
the episodes. E1 serves as the model for E2 and E4b as well as for E5 and
250
WTC I/18
the first half of E6, while E3 is taken up in E4a. The role played by each
episode in the development of the composition also stems immediately
from the material employed. The episodes that are based on the formula
give a concluding impression due to the salient cadential pattern, while
those displaying the imitative motif are either accompanied by a similar
cadential-bass line (as in mm. 28-30) or followed by an explicit cadential
close (see the perfect cadence in A major in m. 24).
Triggered by the two-faced subject, the character of the entire fugue is
ambivalent. It changes constantly between calm, melodious lines of high
expression and cadential patterns of almost neutral tone quality. As a
result, a definition of the basic character must be attempted separately for
the two subject segments and the material resulting from them. In the first
subphrase, the pitch pattern with its poignant leading-notes and tritone
interval as well as the corresponding rhythmic variety suggests a rather
calm basic character. In the second subphrase, by contrast, all melodic
expression seems aborted in favor of a non-committal formula.
The tempo is moderate in order to accommodate the leading-notes
(F-G) and high-tension intervals (G-C) appearing in eighth-note (m. 2)
or even 16th-note rhythm (mm. 21-23, 28-29). The articulation should
reflect the ambiguous character depicted above. In the subjects first
subphrase, the appropriate articulation demands legato throughout. In the
second subphrase, however, non legato articulation is needed to convey the
cadential character. Both the pitch pattern with its note repetitions and
perfect-fifth interval and the regular rhythmic pattern support this interpretation. Correspondingly, the remaining material also divides into two
fields. All cadential formulas maintain the detached style, while the motif
and the counter-subjects tend toward a melodious character. CS2 is
basically legato; only the fourth leap may be detached. The episode motif
with its complex rhythm and semitone as well as high-tension intervals is
entirely legato. Only CS1 shares the ambiguity of the subject: its first
segment is legato while in the second the non legato intention is enhanced
by the written rests.
The most straightforward tempo proportion between the prelude and
the fugue is probably also the best. It is achieved by equating half a
measure in the prelude with a quarter-note in the fugue. (Approximate
metronome settings: 60 for the compound beats in the prelude and the
quarter-notes in the fugue.) No ornaments need to be considered in this
fugue.
When trying to determine the structure of the G-minor fugue, only the
first section is unambiguous as it is distinguished quite clearly by the
G minor
251
entering order of the voices. Four subject statements, presented in uninterrupted succession by the four parts involved in this fugue, are followed by
the concluding first episode. The next statement appears in reduced ensemble and thus confirms the beginning of a new section at m. 111. In the
absence of any assistance from explicit cadential formulas and structural
analogies in the further course of the fugue, the reduced number of voices
involved in a subject statement as an indicator for the beginning of a
section must be complemented by observations concerning the material.
Reductions of the ensemble appear two more times. In mm. 19-21, the
soprano is resting during the entry of the alto, and in mm. 24-26, the bass
is temporarily suspended during much of the soprano statement, although
the first two beats of the entry sound in complete four-part texture. The
recurrence of the tenor statement in m. 17 distinguishes this subject entry
as a redundant one, thus signifying the imminent closure of a section.
As the final statement of the first section (mm. 7-9) and the final
statement of the fugue (mm. 37-39) both appear polyphonically less
intense than the statements preceding them, it may make sense to look into
the appearance of counter-subjects throughout the fugue. If one assumes,
as a hypothesis, that Bach may have composed lessening density of
contrapuntal material towards the end of each section, the following
groups can be established:
mm.
S
CS1
CS2
1
T
3
A
*
5
S
*
*
7
B
*
11
T
*
*
15
B
*
*
17 19
T A
*
*
24
S
26
B
32 37
T
S
* (*)
*
The harmonic outline of this fugue describes a very simple curve. The
minor dominant serving as the secondary key is reached in m. 11. However, as the subject itself in its original version also modulates to this key,
this D-minor cadence does not seem to establish a truly new tonal center.
Moreover, the ensuing statements still remain very closely linked to the
original tonic: already the tenor entry in mm. 11-13 modulates back to G
minor, the bass entry follows with a subdominant-tonic version, and the
redundant tenor statement closes once again on the minor dominant. Only
the third section leads into new harmonic fields. Its first episode modulates
from G minor to A major, the bass statement begins in F major (the
dominant relative) and concludes in B major (the tonic relative), and the
second episode (E4) modulates back to the tonic G minor.
252
WTC I/18
This fugue seems singularly static with regard to tension. With the
exception of the three initial subject statements, all increases in texture in
the course of a section are annihilated by a simultaneous decrease in
polyphonic intensity. The dynamics of this fugue thus live entirely from
the contrastparticularly in the third sectionbetween subject-dominated
passages and episodes.
WTC I/19 in A major Prelude

This prelude is a polyphonic composition in consistent three-part
texture. Its most prominent thematic unit spans 2 measures. Owing to its
complex structure, it appears as a subject rather than a motif. It is imitated
on the dominant and recurs frequently throughout the composition, in
consistently alternating voices. In other words, this prelude is a fugue.
The initial harmonic progression concludes on the middle beat of m. 3.
As it coincides with the end of the first subject statement, this cadential
close does not indicate a sectional caesura. The first cadential close of
structural importance, i.e., one that occurs outside the confines of a subject
statement, emerges in mm. 11-12. It comes as a typical closing formula,
complete with a cadential-bass pattern in the lower and a do-si-do figure in
the upper voice. This cadence concludes the first section at m. 121 in F
minor.
There are only two sections, each spanning twelve measures. The first
leads from the tonic to the relative minor key, the second from the tonic
relative back to the home key. The sections correspond with one another in
great detail. Apart from a small rearrangement in the order of the components, occasional slight variation, and an extended cadential close at the
end of the piece, the second section seems like the answer to the first.
Compare
with
mm. 1-3
mm. 12-14 (U/M/L corresponds with U/M/L)
mm. 4-6
|
mm. 7-8
|
mm. 15-17
|
mm. 83-111 |
mm. 173-201 (U/M/L corresponds with M/L/U)
|
> mm. 20-223 (U/M/L corresponds with L/M/U)
mm. 11
mm. 223-24 (both are cadential closes)
Looking at the pitch pattern and the rhythm for guidance with regard to
the compositions character, one finds large leaps both in the longer note
values and in the 16th-notes along with a rhythmic pattern that gives the
impression of simplicity and regularity. (The frequent syncopations that
combine with the quarter-note motion to form a complementary rhythm are
not truly perceived as metric displacements.) The basic character of this
three-part fugue should therefore be interpreted as rather lively. This
character, however, accommodates a strong lyrical element.
253
254
WTC I/19
The ideal tempo is moderate with gently paced quarter-notes and rather
swift 16th-notes. The articulation is legato for the 16th-notes and non
legato for the eighth-notes and quarter-notes. Among the quarter-notes,
those with distinct melodic quality (like the four chromatic steps in L: mm.
1-2) are only softly detached, while notes of cadential character (see, e.g.,
L: mm. 2-3, where notes proceed in sequencing descending fifths) sound
more clearly separated. Among the eighth-notes are several appoggiaturas
(U: mm. 7-8 and 15-17), which must be perfectly linked to their resolutions.
The only other longer notes that require legato appear in the trebles
closing formulas in mm. 11-12 (F-E -F) and 24 (A-G-A).
The A-major prelude features only one indication for ornamentation,
the cadential mordent in m. 3. Since it is approached stepwise, this mordent
begins on the main note and can do with only three notes. In m. 14, the
upper voice features a similar figure. It also appears in the context of a
cadential close and equally consists of a dotted-note group on the dominant
before its resolution into the tonic. Although Bach did not indicate ornamentation here, it may safely be assumed that such a typical closing formula
would have been decorated with the same cadential mordent (in this case
beginning on the upper neighbor note and featuring four notes).
The subject of this fugue consists of two subphrases. The first ends
with a 16th-note rest at m.14; the second, consisting exclusively of regular
16th-notes that descend in half-measure sequences, concludes with the C
at m. 33. The fact that the rest in the middle of the subject is of structural
importance and not tension-sustaining follows from the relationship among
the preceding notes. The F at m. 13 is composed as an appoggiatura that
resolves indirectly (i.e., through an artificial leading-note) into the eighthnote E. This resolution concludes a dynamic curve and thus marks the end
of the first subphrase. The scalar ascent after the rest picks up the tension
and prepares a second climax at m. 21. The 16th-note figures in the second
subphrase contain an interesting feature. Stripped of what may be recognized as ornamental splitting, the underlying quarter-note pattern performs
a descent in consecutive fifth that parallels the ones heard in the middle and
lower voices in simple intervals. The example below shows this simplified
melodic line.
The subject recurs five times. The only modification it undergoes
occurs at the beginning of the second section where the four initial 16thnotes sound an octave higher than the remainder of the phrase. (This is
certainly due to the confines of the keyboard in Bachs time: continuing in
the high register for even the first subphrase alone would not have been
possible).
A major
255
Faithfully throughout the entire fugueincluding the initial, normally

unaccompanied entrythe subject is supported by two companions. CS1
is introduced in the lower voice. As is revealed by all its later statements,
it begins on the second beat of m. 1; the A on the first beat is a harmonysupporting note of no melodic relevance. Like the subject, CS1 also consists
of two segments. The first, a chromatic descent through four quarter-notes,
recurs regularly, while the second, introduced as cadential steps, is subjected
to constant variation but always preserves its cadential character. CS2 is
first presented in the middle voice. It begins belatedly, after the end of the
subjects first subphrase. The first two of its prominent syncopations together with their falling fifth remain essential material of this counter-subject
throughout the piece, whereas the closing formula undergoes constant
variation. Both counter-subjects represent decreasing tension as their
overall dynamic design. In their variations, performers will want to modify
only the details but never the principal outline.
In addition to this primary material, two of the subject-free passages
feature in their treble a three-note episode motif, M1, presenting Bachs
famous upbeat + appoggiatura-resolution figure. This motif is remotely
related to the end of the subjects first subphrase. In mm. 63-83, M1 is underscored by a parallel and accompanied by 16th-note figures in descending
sequences; in mm. 143-173, the middle and lower voices create a complementary pattern of descending scales, also in sequences. The reason why
this episode is longer than the first lies in the difference of the harmonic
processes: while the first episode only links the dominant (reached at the
end of the subject answer in m. 63) with the tonic on which the third
subject statement is to begin, the episode in the second section modulates
from the tonic relative F minor back to A major.
The structure of this fugue is straightforward. Each of its two
sections comprises three statementsone in each of the three voices,
connected with half-measure linksas well as a bridging episode and a
cadential close.
256
WTC I/19
mm. 1
2
3
U subject
M
CS2
L
CS1
mm. 12 13 14
U subject
M
CS2
L
CS1
5
6
CS2
CS1
subject
7
8
9
10 11
12
episode
CS1
cadent.
with
subject
for9 motif A
CS2 9 mula A
15 16 17 18 19
episode
CS2
with
CS1
9 motif A subject
20 21 22
subject
CS1
CS2
9
23 24
cadent.
formula A
WTC I/19 in A major Fugue

The subject of the official fugue in this pair is in several regards
quite unusual. It begins with a single eighth-note on the downbeat of the
first measure, followed by a 3/8 rest. Such a brusque beginning appears
almost as a conscious attempt to avoid all melodic characterization. This
impression is reinforced by the continuation after the rest. A sequence of
five perfect fourths leaping upward in regular eighth-notes mock listeners
hope for a line while also blurring the 9/8 time signature with their paired
grouping. This pattern is interrupted only after 1 measures. Interestingly,
the change happens on three levels at once. Rhythmically, the D on the
sixth eighth-note of m. 2 suspends the regularity, if not monotony, of the
eighth-note pulse with a syncopation. Melodically, the descending semitone
D-C on the seventh and eighth eighth-notes of m. 2 represents the first
interval to express some emotional warmth. Harmonically, the suspension
results in an appoggiatura, i.e., a note that is not part of the underlying
harmony and therefore requires resolution. The concurrence of these three
features allows a satisfactory conclusion of the subject only after this
resolution, on beat eight of m. 2. This reveals yet another unusual feature
of this subject: the initial statement and its answer already appear in stretto,
a device normally reserved for entries outside the exposition.
With regard to the question of subphrasing this subject is probably best
understood as consisting of two segments: the initial note with an implied
subsiding of the tension during the rest, and a curve from the fifth eighthnote to the end. (Another concept, in which the rest would serve to heighten
the tension from the initial downbeat to the continuation of the line, is
theoretically possible but in practice somewhat artificial.) The tension
within this subject climaxes on the syncopation. The dynamic outline thus
A major
257
consists of a moderately attacked initial note followed after the rest by a

gradual crescendo that sets out from a fairly soft level, leads to a peak on
the syncopation, and is complemented by a short relaxation.
Bachs harmonizations of the subject in the course of the fugue show
a variety of small changes. They all have in common that the first
representative of the subdominant appears with the three-note group at the
end of the first measure, followed on the downbeat of the second measure
by the dominant (see, e.g., mm. 5 and 24). This leads then into an
interrupted cadence (in most cases
on the second eighth-note of the
subjects second measure), and the
return to the tonic only occurs with
another S-D-T progression:
The A-major fugue comprises fourteen subject statements.
1. mm. 1-2 U
8. mm. 23-24 L
2. mm. 2-3 M
9. mm. 25-27 U*
3. mm. 4-5 L
10. mm. 27-29 M
4. mm. 6-7 L
11. mm. 31-32 M
5. mm. 9-10 U
12. mm. 33-35 L
6. mm. 13-15 L
13. mm. 42-43 M
7. mm. 16-18 L
14. mm. 44-46 L
(*Note the voice crossing in mm. 25-27: M sounds temporarily above U.)
In the course of the fugue the subject undergoes a great number of

changes. In fact the only feature that is reliably stable is the interval
structure of the sequenced fourths in its center. The beginning appears
modified in various ways. In two instances the initial note appears in a
different octave from the remainder of the phrase (see m. 13 where it
sounds an octave lower, and m. 27 where it lies an octave higher). In two
other instances it belongs to a key that is subsequently given up. This
modulation causes the second subphrase to seem harmonically displaced.1
In a fifth case, the curious one-note subphrase is completely altered. In m.
31, the downbeat-note has become a rest and the ensuing rest is replaced
1
See m. 16: the initial B is harmonized in B minor but the statement continues as if in
F minor, later modulating to E. See also m. 42: the initial F is harmonized in F minor but
the statement continues as if in A major, with only the final resolution returning to F minor.
258
WTC I/19
by notes. As these notes appear in the form of an additional ascending

fourth (enhanced by a syncopation), they give the impression to extend the
second subphrase, and thus change the subject in this statement into a
single indivisible unit.
The subjects center with its consecutive fourths is also modified in
detail and in length. Some modifications are tonal. In the answer statements
of mm. 2 and 6, the first fourth interval of the second subphrase is placed
a whole-tone lower. In m. 13, the first leap appears as a diminished fourth.
This is due to the minor mode of this statement and must thus be regarded
as another tonal adjustment. In mm. 16-17, the third interval among the
consecutive fourths is augmented. This interval modification is an artificial
one; it marks the pending modulation to E major. Similarly in mm. 33-34,
the augmented fourth E-A and the diminished fourth A-D are witnesses
of the modulation in this statement from D major to B minor. Furthermore,
three statements feature alterations in the extension of the sequencedfourth pattern. In two instances there is an inner extension: mm. 25-26
contain two and mm. 33-35 three additional eighth-notes in the fourth
pattern. By contrast, in mm. 16-17 the consecutive fourths are shortened by
two eighth-notes.
The syncopation, seemingly so characteristic in this subject, is not
exempt from changes either. It is displaced in various ways or even omitted
entirely. In m. 34, it is delayed by three eighth-notes but basically retains
its position (with tie suspension and weak-beat resolution). In m. 26, owing
to the subjects extension by two eighth-notes, the syncopation appears
later, in a different metrical position, and with an unusual strong-beat
resolution. In m. 17, after the abbreviation of the consecutive-fourths
pattern, the longer note falls on the fourth eighth-note in the measure
which in the compound time of 9/8 is not a weak beatand thus, despite
its tie prolongation, forfeits the character of a syncopation. Finally in mm.
44-45, the second subphrase progresses in uninterrupted eighth-note
motion without any longer note value.
And as if all this was not enough, the subjects conclusion also suffers
various modifications. In three instances, the resolution after the syncopation is delayed. In m. 28, this occurs without any further changes in the
harmonic pattern. In m. 17, the modulation from B minor to E major has
already taken place earlier in the subject statement, and the delayed resolution arrives on the expected pitch. In m. 14, however, the deviation leads
from G to an A that then resolves onto Binstead of the expected A.
This alteration is not merely one of melodic detail but entails a modulation
A major
259
from F minor to B minor in this final extension of the subject. The fugues
last subject statement (see mm. 44-46) features a freely varied ending.
Oddly, the only true stretto in this fugue is the one between the initial
statement and its answer. Further in the course of the composition, subject
entries appear neatly separated from one another. There is one instance,
however, where the rhythmic pattern of the voices accompanying a subject
entry contains a hint of stretto. In mm. 25-26 the middle and lower voices,
accompanying the upper-voice entry, feature a single eighth-note on the
second of the compound beats, followed by a 3/8 rest and a sequence of
eighth-notes in leaps of fourths. Although neither the intervallic connection
between the two subphrases nor the end conforms to the subjects original
shape, the entry of this rhythmic group gives the impression of a stretto.
Owing to the many sequences within this subject, this impression soon
turns into one of parallel motion or homorhythmic contrary motion
respectively (see m. 26).
In the main body of the fugue, the voices accompanying the subject
statements display hardly any characteristic or even recurring features, and
true contrapuntal lines are non-existent. The small figures that do appear
repeatedly are all derived from the subject in a more or less direct way.
Only with the introduction of regular 16th-note motion from m. 23 onward
can one distinguish voices to such an extent as to follow more than one
longer melodic idea at a time. These 16th-notes create a subject companion
that recurs once. In its original appearance, CS spans M: mm. 231-248.
When taken up in mm. 27-29, the counter-subject is extended along with
the subject; its end now falls on m. 291. This counter-subject does not seem
to comprise any features that would distinguish its dynamic outline. The
absence of any rhythmic hallmarks within the constantly flowing
16th-notes and of particular intervals or otherwise highlighted melodic
features classify it as a passively jingling accompaniment rather than a
competing partner. However, in order to distinguish CS from the many
other 16th-note figures yet to come, it is possible to stress the hidden line
inside the runs.
In the passages of the fugue that are devoid of 16th-note motion one
can discern three little motifs that appear repeatedly. M1 first emerges in
U: mm. 3-4. With its fourth leap, tied note, and weak-beat resolution it
260
WTC I/19
sounds like a free sequence of the subjects final five notes. This motif
recurs at the beginning of mm. 5, 6, and 7, each time in the middle voice.
Moreover, variations of this figure can be found in U: m. 43 and M: m. 44.
M2 is introduced in U: mm. 58-65. Here an initial tied note and weak-beat
resolution are complemented by two notes in zigzag motion (a third or,
later, a fourth leap up / a fifth leap down). This figure imitates the end of
the subjects answer plus the two notes following it (see M: mm. 36-41).
M2 recurs in mm. 7-8 (U/M in parallel), 8-9 (M), 10-11 (M, imitated in U),
mm. 11-12 (L, imitated in U), 17 (M, imitated 17-18 in U, then sequenced
in M), 18-19 (U), 22 (M), 43-44 (U), 46 (M), 47-48 (U), 51 (U), 52-53 (M).
If one allowed for interval modification, many more recurrences could be
observed. M3 appears as an emotionally even more intense combination,
consisting of two tied-note-plus-resolution pairs in ascending sequence.
This figure occurs three times: in mm. 12-13 (M), 13-14 (U), and 16 (U).
There are seven subject-free passages in this fugue.

E1 m.
8
E5 mm. 29-30
E2 mm. 11-12
E6 mm. 35-41
E3 m. 15
E7 mm. 46-54
E4 mm. 18-22
Several of the subject-free measures use no recognizable melodic
material. They fulfill no function other than that of a cadential close. This
is true for the extended cadential formulas in B minor in E3 and E6a (mm.
35-361), and for the last measure of E6, which presents a closing formula
in F minor. More frequent are episodes or segments thereof that play with
one of the motifs before they give in to the conclusion: E1 begins with an
M2 parallel followed by a closing formula in A major. E2 includes two M2
imitation patterns before ending in F minor with M3 accompanied by a
cadential bass pattern. E4a and E7a (the first segments of E4 and E7) both
feature M2 with free imitations in the lower voice before leading into a
cadential close in E major and A major respectively. E7c (from m. 51) also
presents M2 with free imitation, now accompanied by 16th-note motion in
the lower voice that also concludes in a cadential-bass pattern.
Three episode segments reveal a particularly close relationship to the
primary material by using larger fragments of subject and counter-subject:
A major
261
E4b (mm. 20-22) exposes in its upper voice a little 6/8-note figure that is
imitated in stretto (see U/M: mm. 20-21) before moving on to a rhythmic
variation of the subject, also with imitation (see mm. 21-22; owing to the
pattern of consecutive eighth-notes, both imitations soon form parallels).
E5 features the first half of CS (U: m. 29, imitated in L) as well as freely
leaping eighth-notes (L, imitated in U). E7b transposes this combination
(mm. 49-50), while E6c recalls it in free variation (mm. 39-411).
The only episode portion to feature a longer motif is E6b, the central
segment of E6 (mm. 36-394). This is particularly noticeable since all three
voices establish patterns of 12/8 length, thus momentarily weakening the
metric order of the 9/8 time.2 Both motifs are then launched again in their
original position, where they break off (see mm. 388-394 along with the 16thnote chain of equal extension in L: F-C, sequenced as G-D).
As the above-listed material reveals, several of the episodes are related:
E6a
.
E3
E7a
.
E4a
E7b
.
E5
In a wider sense, E6 and E4 are conceived in structural analogy: both
begin with a segment providing a cadential close, after which they proceed
to present some genuine episode material that is duly imitated. Both then
continue with a variation of primary material (a variation of the subject in
E4, a variation of the counter-subject episode in E6), and both conclude
in another cadential formula.
The role these episodes play in the development of tension within this
fugue is determined, on the one hand, by the frequent use of extended
cadential formulas and, on the other hand, by the relative independence of
the material presented. There are episodes or segments thereof that complement the preceding subject statement with a straightforward relaxation;
this is the case in E1, E2, E3, as well as in E4a and E6a. Other episodes
serve as bridges. This is true of E5, but less of the passages corresponding
with it. Then there are those episode segments that, after a cadential close
but before the beginning of the next subject entry, attract attention for their
own sake. They are the ones requiring the most radical change of color in
order to be fully appreciated. E4b and E6b are obvious examples. The only
subject-free passage not to fall into any of these patterns is the final
episode (E7). Its three segments, which all allude to different models heard
before, seem primarily to express a reluctance to end the fugue.
2
In mm. 362-374 see U: D-B, imitated in M: E-C, and M: F-E, imitated in U: G-F.
262
WTC I/19
The pitch pattern with its predominance of fourth leaps, and the
rhythmic pattern with its regular eighth-notes accompanied in the center
sections of the fugue by equally regular 16th-notes, both suggest a rather
lively basic character. The tempo may be fairly swift. Bachs 9/8 time
should be interpreted as a choice of notation rather than of pulse, to be
rendered with the idea of a compound triple time.
Articulation in this piece needs careful planning if it is not to destroy
essential musical details. The basic attitude corresponding to the character
encompasses distinct non legato for the eighth-notes (in a crispness that
comes close to the staccato touch of later musical eras), non legato of more
extended duration for the quarter-notes, and a quasi legato of almost
classical leggiero quality for the 16th-notes. These categories, however,
include important exceptions: Within the non legato, all appoggiaturas
must be linked to their resolutions. This applies to the appoggiaturas in the
subject as well as to those in M1, M2, and (twice) in M3. Moreover,
several notes in the cadential formulas demand legato (see mm. 15-16 M:
D-C-D; mm. 19-20 M: E-D-E, mm. 22-23 U: A-G-A; mm. 48-49 M:
A-G-A; mm. 53-54 U: A-G-A). Within the quasi legato of the 16th-notes,
a distinction is desirable between the hidden melodic notes in the
counter-subject and the remaining notes of more directly virtuoso quality.
The tempo proportion between prelude and fugue can be rendered in
two slightly different ways. Performers wishing to emphasize the contrast
of characters, with a more lyrical prelude preceding a more virtuoso fugue,
will opt for a 16th-note in the prelude that corresponds to an eighth-note in
the fugue. Conversely, performers seeking to diminish the contrast and to
present a prelude in Allegretto motion followed by a fugue in Allegro ma
non troppo tempo will choose a proportion in which a quarter-note in the
prelude corresponds with a dotted quarter-note in the fugue.
The fugue comprises two ornaments. One of them, that in m. 8, is a
typical attribute of cadential formulas. As such it should be transferred also
to the corresponding formula in m. 41 (on the dotted G). The second ornament is designated in m. 26. All ornaments begin on the upper neighbor
note and move in 32nd values. The one in m. 8 comes with a written-out
suffix, while its structural counterpart in m. 41 is followed by a weak-beat
resolution. The latter thus represents an interrupted trill that stops short on
the fourth note of the shake, which is then tied over and sounds as a
syncopation to the measures final eighth-note. The trill in m. 26, like that
in m. 8, is a note-filling trill comprising eight 32nd-notes including the
suffix notes.
A major
263
mm. 8-9
41-42
26-27
The fugue contains several features that indicate its structure. These
appear, in the first instance, in the change of rhythmic pattern, the choice
of the episode material, and the texture, and in the second instance, in the
order of subject entries and the harmonic design.
The emergence and disappearance of the 16th-note accompaniment
results in three differently colored passages. The first ends at m. 231, the
second at m. 421. The two structural caesuras are preceded by episodes
that, as was shown above, are conceived in structural analogy: E6 and E4
both consist of a segment with a cadential close followed by genuine
episode material and a variation of primary material. The fact that the
subject statement in mm. 23-24 appears in reduced ensemble further corroborates the first caesura.
Within the first larger passage, the redundant lower-voice entry in mm.
6-7 announces the imminent end of the first section, confirmed by E1 with
its cadential formula. The second section, beginning in m. 9, also features
a redundant lower-voice statement (a repetition of the preceding entry).
Within the second larger passage, the only internal episode, E5, concludes
the third section after a complete set of three statements, one in each of the
voices. The fourth section then begins in m. 31 and contains two subject
entries as well as the closing episode. The relationship between the statements of these sections is also confirmed by the harmony: while the three
statements of the third section appear in the keys of tonic and dominant,
the two entries of the fourth section belong to the subdominant (D major
and its relative B minor). The third passage comprises only two statements
followed by a long episode. No further subdivision into sections is possible
here. However, three facts demand a slight modification of this notion: the
final episode consists of several structurally distinct segments, the first of
which ends with a cadence in the home key (mm. 48-49). Above all, the
16th-note motion that is characteristic for the fugues central passage but
abandoned in the fifth section is taken up again from m. 49 onward. All
this indicates that one should regard the fifth section as completed at m.
491 and followed by a coda in the final six measures.
264
WTC I/19
A major
265
The harmonic outline confirms the structure described above. The four
initial subject statements are in A major, the key in which the first episode
concludes the first section. After another entry on the tonic, E2 modulates
to the relative minor key. The following two statements each modulate
(F minor to B minor and B minor to E major), but E4 closes the section
again in the home key. The third section resembles the first one in that its
statements all relate to the tonic. The fourth section, just like the second,
leaves the home key area, this time for the subdominant and its relative
minor. The episode closing this section re-establishes not the home key but
its tonic relative F minor, leaving the final return to the second, modulating statement in the fifth section.
As a result of their descending order, the four initial entries of the
fugue create the deceptive impression of a four-part fugue. The tension
grows throughout and is only released in the concluding short episode. The
second section features a reverse of this process. Its first entry is the only
harmonically stable one in the original major mode, while its second statement begins in minor and ends in a modulating extension. The third entry,
conceived again as a redundant statement, is varied to such an extent that
it appears much weaker than the preceding one. This section comes to a
transitory close in very soft shading on the downbeat of m. 20, after which
the remaining segments of E4 create an independent little tension curve.
As the third section sets out in reduced ensemble and, what is more, as its
three entries in L, U, and M are arranged in such a way that their actual
pitch position (particularly that of the initial notes) sounds in ascending
order, a gradual increase of tension similar to that in the first section is
created. The fourth section recalls the second one insofar as it also
proceeds from the major to the minor mode in decreasing tension, and also
comes to a transitory close in soft color after the first segment of its
concluding episode, after which the remaining segments of E6 attract fresh
attention for their motivic material. The fifth section begins in the minor
mode. Its second subject statement, although returning to the major mode
and even, in its extension, to the tonic, remains comparably inconspicuous.
In the coda, however, the prolonged ascent of the leaping eighth-notes in
mm. 49-51 engenders a final climaxone that is virtuoso rather than
thematic.
The relationship between the sections is an important feature of this
fugue and should by all means be conveyed in performance. The increase
of tension in the first section and the complementing decrease in the
second section, followed by the small tension curve in E4, all find a
faithful correspondence in the increase of tension in the third section, the
266
WTC I/19
complementing decrease in the fourth section and the following small

tension curve in E6. In the final portion of the composition, the subject
statements appear merely as an afterthought of lesser importance, whereas
the coda provides an unexpected additional climax.
As can be seen from all that has been said, the A-major fugue features
all the structural ingredients of a real fugue. Nevertheless, owing to its
unusually non-melodic subject, orientation for listeners remains extremely
difficult. Only the most exquisite color shading and careful articulation
(see particularly the appoggiaturas) can guide them through this apparent
jungle of leaping intervals and render this composition as a fascinating
experience.
WTC I/20 in A minor Prelude

This prelude displays a contrapuntal texture. The two-part structure is
not strictly observed. It thickens occasionally when supporting block chords
are added as in mm. 5, 6, 7 and 9, 10, 11 or a legatissimo is spelled out by
letting one sound overlap with a number of consecutive pitches as in mm.
13-15, 16+18, 17+19, and 26-28. In terms of its thematic material, the
composition builds entirely on a single motif, the slightly ornamented
broken-chord figure introduced in U: m. 1. The prelude can therefore be
regarded as conceived along the lines of a two-part invention.
In conjunction with its two sequences and a complementary figure in
m. 4, the principal motif establishes a four-measure phrase that constitutes
a complete harmonic progression. The cadential close in m. 4, which
coincides with the end of the larger melodic idea, does not qualify as an
indication of a section ending. The second harmonic progression concludes
in analogy to the first after the statementnow in the lower voice and in E
minorof the four-measure melodic idea (m. 8). Listeners have to wait
until the end of the subsequent development, again four measures long, for
the relaxation and cadential close indicating a break in the preludes
layout. The harmonic conclusion comes after a modulation to C major and
falls on m. 131.
Three structural sections can be distinguished in this prelude:
I
mm. 1-13 tonic minor dominant tonic relative
II mm. 13-22 tonic relative tonic1
III mm. 22-28 tonic confirmed
The predominant features of pitch and rhythm are broken chords in
eighth-notes and ornamental 16th-notes, including written-out trills as in
L: mm. 1-3. The basic character of this prelude is thus rather lively. The
1
The harmonic conclusion of section II is very strange. We expect a return to the tonic, and
the chords in mm. 20-21, D minor and G diminished-seventh, seem to announce this with
their ivvii7. At the last moment, however, the lower voice sidetracks with a leap to C instead
of a step from the leading note to A. The tonic chord thus appears in inversion and fails to
satisfy the listeners demand for a resolution. Even more strangely, the process is repeated
similarly three measures later, where the left-hand progression G-A-B (A) diverts to C and
an unusual major-seventh leap adds to the harmonic irregularity. It is only after yet another
cadential measure that the downbeat of m. 26 finally presents the true tonic bass.
267
268
WTC I/20
tempo should be fairly swift, with the 9/8 time signature read as a convenience for 3/4 time with triplets. The appropriate articulation includes non
legato for the eighth-notes and quasi legato for the 16th-notes. The score
does not contain any ornament symbols.
The one-measure principal motif M1 consists of an A-minor chord
broken in zigzag, changing direction with each of the three compound
beats: up, down, up. Only the first interval is filled with a passing note.
This motif is sequenced twice, one step up in m. 2 and another three steps
up in m. 3. Thereafter, the gradual ascent is complemented with the upper
keynote on m. 41. The subsequent passive descent consists of a run and a
falling A-minor chord in a complementary pattern between both hands.
This descent appears more like a link than an integral part of the fourmeasure phrase, and later modification can easily be anticipated. Only the
very last eighth-note in m. 4 announces the shift from the tonic to the
minor dominant with a D, representative of the chord B-D-F-A, the
dominant-seventh of E minor. In this key the four-measure phrase is then
imitated by the lower voice.
The link following the imitation in m. 8 features the run and broken
chord in ascending direction. The final eighth-note of the measure, A,
seems once again about to initiate a harmonic shift, this time to the key of B.
However, the motifs ensuing entry in L: m. 9 ignores the modulation and
begins once more on E. The shape of the four-measure phrase is modified
here. The two sequences of M1 are descending in steps and the linking
measure, instead of displaying a complementary pattern, splits the original
ingredients. With simultaneous runs in the upper voice and broken chords
in the lower, it ends the first section of the piece.
Throughout the entire initial section, the contrapuntal companion is
characterized by a long note on the first compound beat in each measure
followed by a trill figure on the second beat. Within the initial phrase, both
the long note and the trill represent a tonic pedal, while in the imitation and
the developing phrase, block chords and moving trills accommodate the
steps of the harmonic progression. The final 3/8-note group in each
measure serves as a small-scale link, consisting of broken chords in the
initial phrase and of varying runs in the second and third phrases.
The second section sets out in m. 13 with M1 and its original fourmeasure phrase back in the upper voice, the sequences once more moving
upward. There are, however, two new modifications. The counter-motif
omits the trill figure and instead complements the long notes at the outset
of each measure with runs. Moreover, the link at the end of the phrase
A minor
269
combines the idea of a run with that of the broken chord by retaining the
chord notes from within the run in a split-voice texture. This phrase ends
in C major, without any sign of a modulation even in the final eighth-note
of m. 16. The next measure thus presents a surprise both in pattern and in
harmony. Over a repeated pedal note on the pitch a very mindful listener
may have anticipated for the pivotal eighth-note (F, of D-F-A-C), the
upper voice fits the remaining chord notes into a new complementary
pattern. The resolution of this modulating chord follows immediately with
the newly established pattern, in the final measure of the C major phrase
(compare m. 18 with m. 16). A sequence of the pivot-chord measure
then leads to D minor in m. 20. The last two measures of the second
section come full circle with the principal motif, now in the lower voice,
accompanied by a free variation of the modified counter-motif (compare
U: mm. 20-21 with L: mm. 13-15). The unusual harmonic ending of the
section was already mentioned.
The beginning of the third section gives the impression of a recapitulation: the upper-voice motif begins as did the original in m. 1, and the
counter-motif also takes up the trill-figure familiar from the first section
(compare m. 22 with m. 1). The third beat of the measure, however, features
a variation in the motif, the second measure is harmonically diverted
(compare m. 23 with m. 2), and the expected third segment of the
sequential pattern is substituted by a closing link (similar to the one that
concluded the initial section in m. 12). This leads to the reiteration of the
cadential-bass deviation (see mm. 24-251). The remainder of m. 25 continues the scalar descent and, more importantly, corrects the cadential close
by reiterating it with the proper bass steps and melodic leading note.
Finally, mm. 26-28 are built upon a tonic pedal. The lower part with its
voice splitting is reminiscent of the beginning of the second section (mm.
13-15), while the upper part displays a free variation of the principal motif.
The trill figure, added at the very last moment in a split-half of the upper
voice, recalls once more the characteristic feature of the original companion and thus ends this invention with a five-part A major chord.
The dynamic presentation of the composition should be designed to
reflect the structural outline as well as the varying thematic density. The
climax of the first section falls on m. 9 where, after two four-measure
phrases in continuous crescendo, the developing phrase sets in with a sudden major mode, both voices sounding in a fairly high register. The second
section reaches its dynamic peak at the end of its initial phrase (m. 16).
Thereafter, the tension recedes in gradual waves. (Careful: The G at m.
270
WTC I/20
211 often blurts out because of the preceding large leap but should actually
be very soft, in keeping with the pending cadential close.) In the third
section, the dynamic curves are very gentle as long as tonal stability is still
awaited. Then, however, in the final three measures with their multiplying
parts, a powerful crescendo leads to a luscious completion.
WTC I/20 in A minor Fugue

This subject spans exactly three measures. It begins on the second
eighth-note of m. 1 with the keynote A and concludes with the same note
at m. 41. The eighth-note rest in the middle of the phrase appears between
a strong-to-weak descending third causing a momentary relaxation and a
new upbeat; it must therefore be interpreted as an indicator of phrasing
within the subject. The two subphrases divided by this rest are almost
equal in length. The pitch pattern presents two long stretches of smooth,
stepwise motion, both of which end in consecutive leaps (see m. 2: F-G-E
and m. 3-4: B-E-A.) The rhythm consists exclusively of eighth- and 16thnotes in a simple pattern. Syncopations, tied notes, or other outstanding
rhythmic features do not appear.
Yet while this description of the subjects attributes is correct, it does
not convey the entire truth. Two additional observations shed light on the
particularities of the melodic line. One regards the groups of 16th-notes.
All of them are ornamental figures embellishing an underlying simpler
line. The first two groups, e.g., display the well-known inverted mordent.
The other observation regards the assumed simpler line that is being
ornamented. An examination of the pitches shows that two of the subjects
eighth-notes are of secondary relevance with respect to the melodic
structure. In m. 2, the low E on the second half of beat 3 provides a
harmonic foundation to the unusual diminished-seventh leap; melodically,
however, it is an escape note between the high tension captured in the G
and the resolution brought about by the A after the rest. Similarly, the E at
the end of m. 3 serves to keep the eighth-note pace going: melodically it is
another escape note interspersed into the descending tetrachord D-C-B-A.
The result of this short melodic analysis is a skeleton that enhances the
comprehension of the subject:
A minor
271
In the course of the fugue, Bach harmonizes this subject with slight
variations. In mm. 8-9, e.g., the dominant already appears on the downbeat
of the second measure, while in mm. 4-5 the tonic is retained much longer
and the dominant is only reached on the middle beat. In the second half of
the same two statements, Bach harmonizes m. 101 as a subdominant
representative, while the analogous m. 31 returns to the dominant after only
an eighth-note resolution to the tonic. Beyond these differences, the basic
harmonic outline affirms a progression in which the two relevant steps are
the short-lived resolution of the dominant to the tonic on the last eighthnote of the second measure and the perfect cadence that underlies beats 3
and 4 of the subjects third measure and resolves on the final downbeat:
This fugue features the proud number of thirty-nine subject entries.

Particularly in the works final portion, many statements are considerably
shortened. In the list below, statements completely lacking the second
subphrase are marked with an asterisk.2
1. mm. 1-4 A
14. mm. 37-40 A
27. mm. 67-70 Sinv
2. mm. 4-7 S
15. mm. 43-46 S
28. mm. 68-71 Ainv
3. mm. 8-11 B
16. mm. 43-46 B
29. mm. 73-76 Binv
4. mm. 11-14 T
17. mm. 48-51 Ainv 30. mm. 73-76 Ainv
5. mm. 14-17 Sinv 18. mm. 49-52 Tinv
31. mm. 76-77 Tinv*
6. mm. 18-21 Tinv 19. mm. 53-56 Binv 32. mm. 77-78 A
7. mm. 21-24 Binv 20. mm. 53-56 Sinv
33. mm. 77-78 S*
8. mm. 24-27 Ainv 21. mm. 57-60 Sinv
34. mm. 80-82 A
9. mm. 27-30 S
22. mm. 58-61 Ainv 35. mm. 81-82 S
10. mm. 28-31 T
23. mm. 62-63 Binv* 36. mm. 83-85 Binv*
11. mm. 31-34 A
24. mm. 62-63 Tinv* 37. mm. 84-85 Tinv*
12. mm. 32-35 B
25. mm. 64-67 B
38. mm. 84-86 S
13. mm. 36-39 T
26. mm. 65-68 T
39. mm. 85-86 A*
Note that this fugue, while basically in four parts, introduces a fifth voice after the general
pause in m. 80. This new voice establishes a tonic pedal in mm. 83-87. It seemed practical,
however, to retain the counting of the original four voices as soprano, alto, tenor, and bass
during these measures of five-part texture and use bass 2 for the additional voice that
does not participate in the polyphonic play.
272
WTC I/20
In the course of the fugue, the subject undergoes a number of

variations, harmonic modifications, and abbreviations. The most frequently
occurring change of shape is the inversion: nineteen of the thirty-nine
statements, i.e., almost exactly half of the total number, are upside down.
The second-last subject note, earlier described as of secondary relevance
for the melodic structure, is omitted in m. 70 (A) and substituted by an
ornamental figure in mm. 17 (S) and 26 (A). The ornamental substitution
is followed by a displaced final note in mm. 20-21 (T), 23-24 (B) and, with
a slightly different ornamental figure, in m. 60 (A). As all these variations
happen in inversions, one might get the impression that they are due to the
interval direction. However, the four inverted entries in mm. 48-56 appear
without any such adjustments. The four final eighth-notes are substituted
by longer note values and harmonic deviations in mm. 60 (S) and 75 (B),
whereas in m. 76 (A) they are simply cut off. Even shorter versions of the
subject appear toward the end of the fugue where the second subphrase is
reduced to two notes (A: mm. 80-82) or even only one (A: mm. 77-78,
S: mm. 81-82 and 84-86). In two other instances, the subject statement
breaks off after the eighth-note rest (see T: mm. 76-77 and A: mm. 85-86).
Even less substance is left where not only the second subphrase is
completely dropped but the end of the first subphrase is already varied. In
two statements the last two eighth-notes of the first subphrase are replaced
by a quarter-note, thus supporting the interpretation that this final
eighth-note is another structurally non-essential note (see S: m. 82 and A:
m. 86). In mm. 83-84 (B), only the first eleven eighth-note beats are intact;
in mm. 77-78 (S), this is reduced to ten and in mm. 84-85 (T) to only eight
eighth-note beats.
All these entries appear, however, as structurally relevant components
of strettos. The shortest of the abridged subject statements are of equal
length as the episode motif that often quotes the subjects initial bar; this
creates confusiona confusion Bach apparently intended. Thus in mm.
62-63 the stretto between bass and tenor, featuring incomplete first
subphrases in both voices, does not immediately reveal whether its loyalty
lies forward or backward. Only in-depth analysis will show whether these
are false entries or structurally relevant ones.
Harmonic alterations of considerable impact occur frequently; see, e.g.,
mm. 10 (B), 19-21 (T), 44 (B), 58-59 (A), 69-70 (S), and 74-75 (A).
Strettos are a characteristic feature in this fugue; in fact, from m. 27 onward, not a single subject statement remains uncontested by an overlapping
entry in another voice. Toward the end of the fugue, strettos even involve
three voices (as in mm. 76-78: T/A/S) or four (as in mm. 83-85: B/T/S/A).
A minor
273
Given that the imitation distance in the strettos is usually four eighth-notes,
i.e., the same length as the sequence at the beginning of the subject, this
results in a short parallel. By contrast, parallels of any larger segments of
the subject do not occur.
In a fugue that is so essentially based on stretto work, it is certainly not
surprising that we will look in vain for a characteristic and independent
counter-subject. There are two short accompanying figures that deserve to
be mentioned since they recur. Their common feature is the fact that both
establish a short parallel to one fragment of the subject. They are thus not
quite polyphonically independent. In mm. 4-5 the alto, having just introduced the subject, builds a counterpoint to the beginning of the subject
answer. It begins with two descending tetrachords (see the eighth-notes
A-G-F-E and C-B-A-G), continues with a parallel of the first half of the
second subject measure, and ends, at the latest, with the B on the middle
beat of m. 5. This counterpoint is taken up in mm. 8-9 where it is shared
between soprano and alto, and in mm. 11-12 in the alto (where the
tetrachords are reinforced by a parallel in the soprano). Later in the fugue,
the sequencing tetrachords alone, without the ensuing parallel turn-figure,
recur frequently as a companion to a subject entry; see particularly mm.
14-16 (B, T, A), 28-29 (B), 53-54 (T), 57-58 (T+B), and 73-74 (T). In mm.
18-19, the soprano creates a short parallel to the second segment of the
subjects first subphrase. Similar brief parallels occur in mm. 21-22 (A)
and mm. 24-25 (T). As these parallels can also be read as rudimentary
imitations of the subject beginning, an interesting interpretation is that they
are embryonic forerunners of the stretto entries to come.
This fairly long fugue encompasses altogether sixteen subject-free
passages; most of them, however, are only very short.
E1
mm. 7-8
E9
mm. 52-53
E2
m. 14
E10 mm. 56-57
E3
mm. 17-18
E11 mm. 61-62
E4
m. 27
E12 mm. 63-64
E5
m. 31
E13 mm. 71-73
E6
mm. 35-36
E14 mm. 79-80
E7
mm. 40-43
E15 mm. 82-83
E8
mm. 46-48
E16 mm. 86-87
In the material that makes up these subject-free passages, two facts
stand out: many episodes are simple cadential formulas of one-measure or
even only half-measure length, and several others feature the subjects
head motif. E2, E3, E4, E15, and E16 present variants of the cadential bass
steps, E2 and E4 additionally feature the dosido formula, E3 and E4 the
274
WTC I/20
typical melodic turn with dotted leading note and anticipated resolution. In
three other episodes, the same components make up only a portion of the
subject-free passage; see the second half of E8 as well as the conclusion of
E12 and E13. In the latter case, the sopranos cadential formula overlaps
with the beginning of the subsequent statement in the bass. The subjects
head motif is found in seven of the episodes, often accompanied by a
scalar passage (see E1, E3, E6, E7, E10, E12, and E16). Finally in E5, E8,
and E9 the scalar segment appears without the subjects head motif
Only three of the episodes feature figures that are immediately
sequenced and thereby attain a minimum of independence: bass and alto in
E7 mm. 40-41, the trill motif in alto and tenor of E9 as well as in the
soprano of E10, and the imitations of soprano and tenor including their
sequences in the alto of E13. E8 is the only episode presenting an imitation
of the subjects tail (see T: mm. 45-47). The 16th-notes in all episode types
form ornamental figuration or broken chords.
There are few structural analogies among the episodes. One can detect
a correspondence between two cadential and two non-cadential passages
(E2 and E4, E1 and E5). Moreover, E9 and E10 share the same prominent
motif and scalar descent. The role each episode plays in the development
of tension within the fugue is determined by motifs and closing formulas.
E1 bridges between two consecutive statements but does not create
a color contrast since the Ms figure appears as a sequence of the
previous measure. Conversely, E2 gives the impression of a close,
with a definite relaxation. E3 is ambiguous: the bass pattern speaks
of conclusion, the false entry in the alto seems to point forward. E4
follows with an unequivocal closure and a definite tension release.
Like E1, E5 connects two entries. E6 with its ascending sequences
leads to the subsequent stretto. E7 introduces the first color contrast
owing to its novel material and, thereafter, a decrease of intensity.
E8 presents the long-awaited closure with its extended cadential
formula. Surprisingly, it is succeeded by what must be interpreted as
a preparation for the ensuing subject statement.
E9 and E10 serve as dynamic bridges. E11 like E7 presents genuine
episode material with a concomitant color contrast but no conclusion (the stretto in mm. 62-63 is structurally analogous to that in
mm. 43-46), and E12 with its cadential close corresponds with E8.
E13 also suggests a color contrast because of its genuine little motif.
Its ending with the overlap of a cadential close with the beginning of
an ensuing subject statement recalls E2 and distinguishes this as
another bridging episode. E14 differs from all its predecessors by its
A minor
275
abrupt halt in a six-part chord. This unexpected halt anticipates the

chord on the middle beat of m. 82, also in six-part texture, with
which it corresponds in its harmonic distinction as an inverted
dominant-seventh chord and its rhythmic feature with ensuing
general pause. In other words: E14 is not so much a fugal episode as
a dramatic passage and may be interpreted with a matching rise of
dramatic intensity. E15, following after the works second general
pause, provides for some relaxation. The concluding E16, set in
seven-part texture, has no choice but to crown the fugue with a
triumphant climax.
Despite the apparent predominance of stepwise motion in this piece,
the ornamental character of the 16th-notes and the overall simplicity of the
rhythmic pattern indicate a rather lively basic character for this fugue. A
moderately swinging quarter-note pulse is a good choice to accommodate
the pensive side of the piece. Both the surface pattern with its ornamental
16th-notes and the assumed underlying simpler line are marked by a strong
metrical component and express a dance-like character. In order to convey
this quality, the natural metric order in four-four time, with weak second
and fourth beats, should be observed. The articulation appropriate for this
character demands non legato for the eighth-notes and all longer note
values and quasi legato for the 16th-notes. In the frequent cadential closes,
the dosido figures as well as the dotted-note formulas require genuine
legato. Many exceptions occur, however, in the context of notes (often
suspensions) that constitute appoggiatura-resolution or closing formulas.3
A good choice for the relative tempo of the prelude to the fugue is one
that reflects the interpretation of the compound meter: a dotted quarternote in the prelude corresponds with a quarter-note in the fugue. Approximate metronome settings: 80 for both the compound beats in the prelude
and the quarter-notes in the fugue.
The A-minor fugue features both cadential and melodic ornaments.
Beyond these, some cadential ornaments may be added where they would
have been so self-evident for Bachs contemporaries that marking them
explicitly did not seem imperative. In mm. 21 and 64, the dotted-note
3
For appoggiatura-resolution see mm. 6-7 [alto: C-B, E-D, D-C], m. 11 [soprano: D-C], m.
26 [soprano: B-A], mm. 40-41 [tenor: G-F, F-E, E-D], m. 62 [alto: F-E]; m. 64 [alto: G-F];
m. 70 [tenor: A-G]. For do-si-do formulas see mm. 13-14 [alto: E-D-E], mm. 16-17 [bass:
G-F-G], m. 23 [soprano: C-B-C], mm. 26-27 [soprano: A-G-A], m. 32 [tenor: E-D-E], m.
34 [alto: E-D-E], mm. 47-48 [soprano: C-B-C], m. 60 [bass: A-G-A], m. 64 [tenor:
D-C-D], m. 73 [alto: F-E-F], mm. 75-76 [bass: G-F-G]. There is only one closing formula
in the bass that requires partially legato articulation; see mm. 47-48 [D-E-F-G = legato].
276
WTC I/20
figures in the soprano carry the mordent that is so typical for this formula.
The brackets indicate that the composer himself did not find it necessary to
specify this obvious ornamentation, but one of his students took it down.
The same brisk mordent should be added in m. 17 on the dotted F and in
m. 73 on the dotted G. Among these, only the mordent in m. 17 begins on
the main note and therefore makes do with only three notes; the remaining
ones are launched from the upper auxiliary and comprise the regular four
notes.4 The trill in the motif appearing in E9 and E10 represents a
note-filling ornament. Starting each time on the upper neighbor note and
shaking in 32nd-notes, it comprises eight notes including the suffix that is
spelled out in all cases.
Much headache has been caused by the apparent impossibility to play
the fugues final five measures with their sustained bass pedal and the
texture with up to seven voices. The following excerpt attempts to help by
suggesting one possible execution, fingering, and use of the middle pedal.
In this A-minor fugue, Bach seems to play with transformations of the

subject in a way that reminds one of the D-minor fugue, also in Book I.
Conclusive evidence for the structural design can therefore be derived
already from an observation of the order in which the subject statements
enter, from the change of shape found in a number of consecutive entries,
and from the presence or absence of strettos. The data that can be collected
in this respect are particularly indisputable in the first three quarters of the
fugue.
4
In m. 27, the figure appears in the alto and thus makes ornamentation much more difficult
to execute. However, for performers with a good technique it is possible to play the
mordent in the left hand, then take over its sustained final note in the right thumb, the
ensuing 16th-note A again in the leftquite without interrupting the legato required for the
upper voice.
A minor
277
The first four entriesone in each of the four voicesappear

without any overlapping (see mm. 1-14).
There follow four inverted statements, also one in each voice, also
neatly after one another (see mm. 14-27).
The next eight entries retrieve the original shape of the subject but
are grouped in pairs: every other statement enters already half a
measure after its predecessor and thus overlaps for two and a half
measures (see mm. 27-46).
This group is followed by eight entries in inversion that feature the
same kind of stretto at the same half-measure distance (see mm.
48-63). Note that the correspondence between the two groups of
four-fold stretto gives the two abridged statements in mm. 63-64
more importance than their radically shortened shape would suggest.
Analysis reveals that they constitute indeed the fourth, though
incomplete, group of this section.
Each of these four groups ends with one of the episodes that were
earlier recognized as cadential formulas: E2 in m. 14, E4 in m. 27, the
second segment of E8 in mm. 47-48, and E12 in m. 64. The confines of
sections I through IV can thus be regarded as confirmed.
The remaining subject statements, from m. 64 onward, comprise a
stretto in original shape followed by two strettos in inversionall using
the complete phrase. Immediately afterward there are three strettos comprising, respectively, three entries (mm. 76-78), two entries (mm. 80-82),
and four entries (mm. 83-86). All these statements are drastically abridged;
in fact, in each of the strettos, only one of the voices progresses at all into
the second subphrase. In order to determine the intended grouping of these
six strettos one had best again look for guidance in the episodes. There are
two subject-free passages ending with a perfect cadence (E13: belated
cadential close in m. 73, and E15: cadential formula in m. 83). These
should therefore be interpreted as designating section closes. The fact that
in the former case, the typical closing formulas in the soprano and alto
overlap with the beginning of the subsequent bass statement indicates that
the structural partition between the fugues two final sections is not as
clear-cut as the earlier section-endings were. The cadential close of E15 in
A major, on the other hand, appears strong and definite; it determines the
final measures with their pedal note in the added second bass as a coda.
The use of texture in the various sections may give further clues for
Bachs conception of the fugues design. The first stretto of section III
(and, with some irregularities at the beginning, also that of section IV)
begins accompanied by only one voice, i.e., in reduced ensemble. Even
278
WTC I/20
more striking is the fact that the first stretto in section V is at first entirely
unaccompanied. By contrast, the beginning of section VI with three active
voices gives the impression of reduced ensemble but actually encompasses
four parts; the indirect pedal C in the soprano suddenly comes to life again
in m. 75. This substantiates the interpretation that sections V and VI are
more closely connected than any other consecutive sections. The fugues
overall design thus contains two pairs of corresponding sections (sections
I/II and III/IV), followed by the only minimally structured larger body
containing sections V/VI and a short coda.
The harmonic outline supports the structural correspondences.
The first section comprises entries on the tonic and dominant of A
minor before ending on an E-major cadence. Section II begins on
the dominant and returns to the tonic with an A-minor cadence.
I/II = tonic - dominant - tonic
During its three initial strettos section III is still firmly rooted in the
home key. After the longer interruption of E7, however, the redundant entry-pair presents the relative major key, which is confirmed,
at the end of E8, with a cadential close in C major. Section IV
returns to the home key but ends, both in its second stretto and the
final abridged pair, in the subdominant region, which is confirmed
in E12 with a cadential close in D minor.
III/IV = tonic - its relative major - tonic; end on subdominant
The two final sections, earlier ascertained as forming a larger unit,
further underscore the subdominant field. Section V begins with a
stretto of entries in A minor and D minor, and were its final cadence
not placed across the beginning of the sixth section, it would conclude in F major, the subdominant relative. Section VI, after several
harmonic adventures,5 ends with a perfect cadence in A major, the
Picardy-third version of the tonic (m. 83). In the coda, however, the
final four-part stretto is led by a soprano statement in D minor, a last
reminder of the subdominant key set against an overwhelming
assertion of the tonic (three A-minor entries and the A pedal). The
final cadence, as expected, confirms the A major conclusion.
V/VI = subdominant - (its relative major) - major tonic
coda = subdominant/(minor) tonic - major tonic
5
A striking number of pedal notes mark section VI as concluding. See mm. 73-75: C
(soprano), mm. 76-77: G (bass; beginning at m. 753 but escaping momentarily to its leading
note), mm. 79-80: D (bass, interrupted), and mm. 83-87: A (additional second bass voice),
as well as the two inverted dominant-seventh chords that, enhanced by the voice-splitting
into six parts, precede the reiterated general pause in mm. 80 and 82.
A minor
279
The development of tension in the first section of this fugue is both

normal, insofar as it describes a steady dynamic rise, and unusual, insofar
as this rise seems least caused by contrapuntal complexity. Similar observations can be made with regard to the second section: owing to the frequent
280
WTC I/20
parallel motion in partial phrases, the impression of polyphony remains

relatively weak. The third section is the first to present considerable polyphonic contrast. As the connecting episode E6 with its fake stretto entries
keeps the tension from dropping, an almost continuous crescendo results
through these three paired statements. E7, however, introduces the first
pronounced color contrast in this piece. The ensuing fourth stretto of this
section is not only redundant due to the entering order of the voices (its
leader, the soprano statement, is a repetition of the soprano-led stretto
that opened this section); it also fails to regain the intensity interrupted by
E7, owing both to a rather passive fourth voice that consequently drops out
and to the less dramatic character of the subjects major-mode variant. The
extension of this entry, created by the imitation of the second subphrase,
concludes the section with a prolonged diminuendo. The development is
similar in the fourth section. The second stretto is the most powerful from
a contrapuntal aspect, while the third, with its many parallels of partial
phrases, its color contrast in E7, and the often passive voices brings about
a relaxation.
The unaccompanied stretto launching the fifth section constitutes the
beginning of an entirely new dynamic development. The tension rises
despite the color contrast in E13 all the way to the general pauses and the
subsequent cadential close. A powerful coda, appropriately interpreted in
crescendo allargando, concludes this extraordinary fugue.
WTC I/21 in B major Prelude

The B major prelude is in its texture essentially non-polyphonic. The
performers two hands act either in alternating motions or join forces for
many-voiced chords. Never does one of them establish a pattern without
the other.
There are three conspicuous figures. Two of them are characterized by
an aspect of keyboard technique rather than by particular melodic or rhythmic shapes. The first appears as a pattern of open-position broken chords,
executed in a complementary play between the two hands. This figure
prevails in mm. 1-7 and 9, after which it recurs only in the final m. 21. The
second figure is a single-voiced wavy run, also kept in virtuoso style, i.e.,
in the performers own choice of a complementary use of both hands. This
figure is introduced in mm. 3-4 and taken up in m. 8 and in parts of mm.
10-20. The third and last component creates a contrast to the two one-track
figures described before insofar as it engages both hands simultaneously,
creating a dotted-note chordal pattern (see mm. 11, 13, 15 and, in modified
rhythm, mm. 17-18). The composition thus recalls several Baroque genres:
the toccata, the fantasy, and the French overture. As there are no real
melodic features and, except for the short dotted-note figures, no prominent
rhythms, the aspects that solely characterize the piece are, on the one hand,
the virtuoso layout, on the other hand, the harmonic progressions and the
sequential patterns within it.
The first harmonic progression concludes in the first half of m. 3, with
the perfect cadence in the home key B major. Yet while this B-major
chord provides the harmonic resolution to the preceding F7 chord, the pitch
position of the bass with its leap of one and a half octaves in m. 31 counteracts any true relaxation. Instead, the downbeat of m. 3 doubles as the beginning of a new structural unit. The second harmonic progression is more
extended. The music modulates to the dominant F major, which first appears
by way of a plagal cadence in mm. 53-63. The seventh E, however, added
at the last moment, converts the F-major chord into a V7 and thus once
again refers back to B major. The dominant key is only established with
the authentic cadence that, complete with a traditional cadential-bass
pattern, concludes at m. 101 (with a possible passive extension throughout
the broken F-major triad in the first half of m. 10). This cadential close
281
282
WTC I/21
must be regarded as a structural caesura. Understanding this is particularly

important since the virtuoso 32nd-note pattern continues uninterrupted. It
therefore happens easily that the structural partition is either overlooked or
misplaced after the D on m. 111, which already represents the first step into
the subsequent harmonic and structural unit. The prelude thus comprises
only two sections: mm. 1-10: tonic to dominant and mm. 10-21: dominant
to tonic.
Both the very regular rhythmic structure with almost continuous 32ndnote motion and the broken-chord patterns indicate a rather lively basic
character. As for the tempo, this virtuoso piece should give a very brilliant
impression and may be played almost as fast as is technically possible.
(What is swift and brilliant, however, is the surface pattern, not so much
the actual pulse of the piece, the quarter-notes.) Whether this tempo and
character are valid throughout the entire prelude is a question frequently
discussed. In one of the manuscripts, the introduction of the dotted-note
rhythm pattern in m. 11 is marked Adagio. What are we to make of the fact
that this indication did obviously not form part of the original layout but
was added later? One possible interpretation is that Bach could have
decided to add this heading on the spur of the moment because a performer
(student?) played in a character that he deemed too light to be appropriate;
in this case the indication may refer to a contrast in touch and expressivity
rather than to any considerable change of tempo. Or, basically along the
same lines, even if Bach originally estimated that an intended change of
character was expressed clearly enough by the rhythmic hint at the style of
the French overture and did not need any explicit verbal invitation, he may
have added Adagio to suggest a freer treatment of temporubato. (The
two extreme interpretations, either to play through the entire piece without
any change of attitude, or to play the second half twice as slowly as the
first, are the least likely solutions.)
The appropriate articulation includes non legato for the eighth-notes
and quasi legato for the 32nd-notes. Touch should vary with the character
of the components: lightly bouncing non legato combined with a very crisp
quasi legato in mm. 1-10 and 19-20, a heavier detached style (possibly
combined with a denser legato) in mm. 11-18.
The only ornament in this piece is demanding. The downbeat of m. 19 is marked
with a compound ornament that, in correct
execution, would encompass eight notes and
end in a tie-prolongation, a density not at
all easy to realize in the desirable speed.
B major
283
The chief processes in the prelude rest on the facts described above.
The first phrase is determined entirely by figure 1. It appears in a pattern of
descending half-measure sequences and is rounded off by a half cadence.
The descending sequences are counter-balanced in the following phrase by
two ascending sequences built from combinations of figures 1 and 2. The
third phrase picks up the ascending motion, from B through an octave to
A, beginning in straightforward chromatic steps and continuing in a diatonic passage with octave displacements. While the ascent presents itself in
figure-1 pattern, figure 2 in m. 8 returns in a double curve to the low E, the
leading-note of the long-anticipated dominant. A one-measure cadential
pattern with figure 1 confirms this key and with it the preludes first half.
In m. 10 Bach subtly creates the impression of both rounding off and,
at the same time, announcing something new. While the first half of the
prelude is made up entirely of the two figures with open-position broken
chords and scalar runs, the tail of its final cadence introduces a combination of the two figures: a pattern that, like figure 2, uses one-track texture
but, like figure 1, moves in broken chords. This combination (here referred
to as figure 1+2), turns out to frame the second half of the piece.
The first statement of the dotted-note figure 3, presented in m. 11, is
linked to its two descending sequences by extensive statements of figure 2
whose dive and soar respectively span almost three octaves. Despite the
inverted direction of the run in the second statement of the figure, the
listener can still perceive the continuing sequential pattern. After the second
sequence, the recurrence of figure 1+2 seems to announce the impending
close of this section. Yet once again the addition of a seventh redefines the
tonic triad and thus diverts the cadence. Another statement of figure 1+2
leads to the compound ornament before one measure each of figures 2 and 1
complete the preludes larger frame in symmetrical fashion.
WTC I/21 in B major Fugue

The subject sets out after an eighth-note rest, extends through four entire
measures, and ends at m. 51, where the D represents the return to the tonic
after the F7 chord in m. 42-3. A closer look at the phrase structure reveals
three subphrases in the pattern [a a' b]. These are easily distinguishable
once the sequential pattern is recognized behind its veil of ornamentation:
compare the initial subphrase, F-G-F-B-D-C, with its sequence, an embellished G-A-G-C-E-D, followed by a consequent around A-C-E-D.
284
WTC I/21
Phrasing in the subject therefore occurs after the downbeats of mm. 2 and 3,
with exactly one measure length for each of the two subphrases, while the
subsequent two measures represent a single unit.
The interval organization in this subject includes sixths, fourths, and
seconds in the eighth-notes and predominantly small intervals in the essentially ornamental 16th-notes. As the sixths occur in weak metrical positions
(i.e., after the beat, not toward it), these large leaps do not convey emotional tension but an energetically bouncing temperament. Corresponding
with this lively interval pattern, the rhythmic pattern throughout the fugue
is very simple, consisting almost exclusively of eighth-notes and 16thnotes. Even the regular syncopations in the main contrapuntal voice (see
U: mm. 5-7 etc.) cannot diminish this effect.
The subjects harmonic background is characterized by two active
steps within the first two subphrases: both times a simpler chord underlying
the 5/8 upbeat is followed by an incomplete minor-seventh chord. The
third measure concludes the cadence with a straightforward V7-I. The
harmonic progression thus ends already at m. 41. Yet for the dance-like
character Bach intended for this fugue, he prefers phrases with an even
number of measures. The additional eight eighth-notes, consisting of an
exact repetition of the ones just heard, thus appear as a metric complement
with neither melodic nor harmonic information.
I ii7
vii7 iii7
V7
V7
The subjects dynamic design should convey the phrase structure, the
metric organization, and the harmonic features. A distinctly virtuoso aspect
enters with the ascending sequences and their growing amount of embellishment. Within the first two subphrases, the climax falls on the respective
downbeatsboth because this is the point of harmonic emphasis and
because any other accent would blur the metric structure. (If, as happens
easily for lack of attention, the peak notes B and C are accented, this will
necessarily give listeners the mistaken impression that the subject begins
with three upbeat eighth-notes followed by the first downbeat on B.) By
the time the third subphrase is approached, the meter is firmly established.
Therefore, the fact that this incomplete sequence comes without what was
previously the climax and instead just relaxes from the E (V7, m. 33) to the
final D (I, m. 51), suggests that in this subphrase, highest pitch and
dynamic peak coincide. With regard to the relation between the three
B major
285
subphrase climaxes C, D, and E, the works virtuoso character favors the

high E. However, in case a performer prefers an interpretation that seeks
more serious nuances in this fugue and its subject, it is also possible to
regard m. 31 as the strongest climax and the two final measures as a less
intense tail to the more substantial main body.
This fugue comprises only eight complete and subject entries as well as
an incomplete one.
1. mm. 1-5 U
5. mm. 22-26 M
2. mm. 5-9 M
6. mm. 26-30 L
3. mm. 9-13 L
(7. mm. 35-37 M)
4. mm. 13-17 U
8. mm. 37-41 U
9. mm. 41-45 M
Except for the interval adjustment in the tonal answer, the subject
remains untouched throughout the fugue in detail as well as in shape. No
stretto or parallel statement ever materializes. But ever after its first entry,
it comes escorted by companions that remain faithful to the end. They are
not, however, quite as independent from their leader as true contrapuntal
technique would require, particularly in the case of the second companion,
which runs largely in parallel to the subject. Although the term countersubject might thus not seem appropriate in its strictest sense, the usual
abbreviations are employed here for easier reference. CS1 is introduced in
U: mm. 5-9. Its three subphrases coincide exactly with those of the subject.
As in the subject, the first and second subphrases are related by ascending
sequence (although the first subphrase appears more elaborate than the
second here). Also as in the subject, the third subphrase consists of a onemeasure model followed by its repetition. (This subphrase encompasses a
four-note parallel to the subject; see m. 81.) In terms of the dynamic
outline, the three climaxes fall slightly earlier than those in the subject. The
prominent features within the first two subphrases are the final syncopations, which capture the peak of tension. In the third subphrase, the short
upbeat-like ascent preceding the protracted note repetition indicates a soft
crescendo, which is followed by a long diminuendo. CS2 seems more like
a filler than a polyphonic partner. Introduced in U: mm. 9-13, it consists of
four short gestures interrupted by rests. The first two, again conceived as
ascending sequences, complement the two 16th-notes missing in the
rhythmic patterns established by the subject and CS1. Their second note
sounds in unison with the other companion, and their final eighth-note,
286
WTC I/21
rather than fulfilling any melodic purpose, complements the downbeat

harmonies. The (identical) third and fourth figures blatantly double the
subjects segments in parallel sixths. To look for a meaningful development of musical tension within this highly dependent element would mean
to overrate its importance.
The fugue comprises only three subject-free passages: E1 = mm. 17-22,

E2 = mm. 30-35, and E3 = mm. 45-48. All of them are directly related to
the primary material and can be divided into only two patterns, both
established in E1. In mm. 17-19, the subjects third subphrase is sequenced
a whole-tone higher while the accompanying companions are imitated in
inverted voices. A similar process occurs in mm. 45-47. Here the middle
voice repeats the second half of the previous subject entry an octave lower,
while the second half of CS1 moves from upper to lower voice (and thus
two octaves down) and that of CS2 from lower to upper voice (one octave
up). In mm. 19-22, E1 is reduced to two-part texture and the material
employed derives entirely from the subject. The upper voice continues in
sequences of the measure-3 pattern (now descending by steps), while the
lower voice takes up the first subphrase in inversion (compare L: m. 19
with U: m. 1), which is then also sequenced in descending steps. This
combination recurs similarly in E2, where mm. 30-33 feature the inverted
subject head in the upper voice (the descending sequences are underpinned
only at their very ends by doubling in the middle voice), the lower voice
recalls the descending sequences of the measure-3 pattern, and mm. 33-35
continue this process in inverted voices. The only difference is that the
filling voice is now more complete.
The only episode measures that remain are the two cadential measures
at the end of the composition. These feature a cadential-bass pattern and a
traditional closing formula in the upper voice. As all these passages are
designed in sequential patterns, the ascending or descending direction of
B major
287
each sequence determines the dynamic tendency and, with it, the role each
episode plays in the course of the fugue. Thus E1 momentarily surpasses
the tension expressed in the fourth subject entry but subsequently recedes
with each measure. E2 is conceived, despite its two-fold structure, as a
single decreasing line, and the first segment of E3 resembles an echo from
which the cadential close resurfaces.
One can thus state that this fugue is only rudimentarily contrapuntal in
texture and entirely determined by its subject. At the same time, the subject
remains virtually unmodified throughout the composition. This may
suggest that what matters in this fugue it is not the material, and that the
chief task fulfilled here by the subject is to establish the vigor and mood
upon which the entire piece depends.
The basic character is lively. This is supported both by the simple
rhythmic structure and by the pitch pattern with its frequent leaps and its
ornamental 16th-notes. The tempo may be very swift, particularly since the
demands of polyphonic playingand hearingin this fugue are almost
negligible. The articulation combines an energetically bouncing non legato
for the eighth-notes with a crisp quasi legato for the 16th-notes. Ornaments
do not occur in this fugue. The relative tempo of the prelude to the fugue
uses the larger pulse for a proportion: a half-note in the prelude translates
into a dotted half-note in the fugue. Approximate metronome settings:
prelude beats = 72, fugue beats = 108.
In the absence of changes in texture or conspicuous cadential patterns,
an analysis of the structure must rely on the episodes and the harmonic
development. As has been shown, the episodes are related with one another
in that two patterns introduced in E1 return in reverse order in E3 and E2.
This may be interpreted as suggesting an axis symmetry (a+b b, a) and
can thus hint at a design in which the first section would find its correspondence in the joined second and third sections. The tonal organization
supports this view. The four initial subject entries remain in the sphere of
B major. The first episode then modulates via the relative key G minor to
its dominant D Major. The fifth and sixth subject statements represent the
sphere of the relative-minor key (with G minor in mm. 22-26 and C minor,
the subdominant relative, in mm. 26-30). The ensuing episode confirms the
key of C minor. Thereafter, the incomplete statement leads back to the major
mode. Thus the next complete statement can enter in the subdominant E
major, followed by the return to the tonic in the final statement.
A big dramatic development is clearly not the purpose of this virtuoso
fugue. Once the subject and its two companions have been established,
there is little distinction between successive entries. The interpretation of
288
WTC I/21
tension will thus have to concentrate on two features: the differentiation of

color between the (outer) sections in the major and the (middle) section in
the minor mode, and the gesture expressed in the episodes.
WTC I/22 in B minor Prelude

This prelude is determined by its gently swinging pulse, which is conveyed by an overwhelmingly uniform rhythmic pattern. The entire piece
builds upon only two rhythmic
models. As the second is a close
variation of the first, these can
Rh1
Rh2
even be regarded as two versions
of a single archetype. (The only
two exceptions, featuring six consecutive 16th-notes, occur in cadential
context, in mm. 194 and 234). The effect of this rhythmic uniformity is a
meditative atmosphere in which transitory emotional upheavals seem of
subordinate importance.
The first harmonic progression concludes at m. 31. Up to this point, the
music develops over a tonic pedal, thus giving the impression that the
piece has not quite begun. Moreover, the cadence supports the melodic
motif presented in the upper voice. This harmonic close should therefore
be interpreted as marking the end of a phrase rather than of a structural
section. The second harmonic progression concludes, still in the tonic key,
on the middle beat of m. 7. Both the way in which this cadence is prepared
and the subsequent development designate this closure as a structurally
relevant caesura.
There are altogether four sections in this prelude. Two of them show
harmonic subdivisions.
I
mm. 1-73
tonic
(mm. 1-31, 3-73)
II
mm. 73-132 tonic dominant
III
mm. 132-201 dominant dominant relative dominant
(mm. 132-152, 152-201, ending in an imperfect cadence)
IV
mm. 201-24
dominant tonic
The only analogy within this structural pattern is obvious as it suggests
a recapitulation: mm. 20-21 take up mm. 1-2, transposed to the dominant
key and with slight variations in the middle voices.
The preludes basic character is rather calm. The pitch pattern consists
to a very large extent of seconds. Another striking feature that influences
the character appears in the frequent pedal notes (mm. 1-31: B, mm.6-71:
F, m. 7: B, mm. 93-101/11-121: B, mm. 20-221: F, and mm. 233-24: B).
289
290
WTC I/22
The tempo in which this character can best be expressed is fairly slow. The
quarter-notes of the time signature can be imagined as a solemnly swaying
movement, and the eighth-notes perceived as the continuous pulse should
be completely without haste. The articulation required in such a piece is a
dense legato.1 The score does not contain any ornaments.
The pulse in this metrically determined prelude is thus two-fold: the
main swinging motion is in quarter-notes, whereas the consistent pounding
is experienced in the eighth-notes. This ambivalence is established in m. 1
where the unremittingly throbbing pedal note contrasts not only with the
rhythmic motif but also with the mid-texture chordsand these, in their
iambic pattern, underpin the larger quarter-note pulse.
rhythmic model:
chordal background:
repeated pedal:
As the example shows, the preludes initial phrase displays a constantly changing number of voices (from four to seven, though mostly
five). Thereafter, the texture from m. 3 up to the downbeat of m. 20 is in
consistent four-part writing. The recapitulation of the initial phrase then
resumes the five-part texture, with extra voice splitting only in m. 22,
where beat 2 features seven and beat 3, nine voices.
While the initial phrase might easily pass for the introductory line of a
homophonic composition, there are several passages that may be described
asmodestly yet very consistentlycontrapuntal. A plausible interpretation of these confusing data is the following: the prelude can be regarded
as a work in four-part writing whose opening phrase and its recapitulation
1
As this prelude contains such a large number of repeated notes, the help of the damper
pedal will probably be inevitable to produce the desired smooth result. This needs careful
pondering. Using the pedal only where there are repeated notes would create an uneven
tone color. Thus regular pedaling throughout is the only alternative to not using the pedal
at all. The cleanest solution is to depress the pedal after the last 16th-note fraction of each
eighth-note beat, and to release it with the onset of the following eighth-note. The nature of
this meditative piece, however, admits also a slightly less transparent execution in which
the pedal is simply changed on each eighth-note, thus giving the second 16th-note in each
rhythmic figure a somewhat murky coloring. Eighth-notes, however, must be clear.
B minor
291
are enriched with a bass 2 as well as with a few irregular voice

splittings.2 Such an interpretation contributes to a better understanding
both in listening and performing.
A more detailed analysis of the development of melodic figures and
changing textures is nonetheless worthwhile. Supported by the alto in
double thirds, the soprano motif establishes the first rhythmic pattern
(Rh1) in mm. 1-31; it will here be referred to as M1. In its varied imitation
by the bass in mm. 3-51, the motif is accompanied by a figure in the highest
voice that retains Rh1 for half a measure and then modifies it to Rh2. The
middle voices lengthen their earlier iambic pattern by an extra weak beat.
In m. 4, alto and soprano launch a complementary pattern that soon
pervades all four voices: A+T move in double thirds, like the two melodic
voices in mm. 1-2, while S+B are joined in rhythm but not in pitch outline,
like the two melodic voices in mm. 3-4. These two pairs alternate in such
a way as to produce Rh2. In mm. 6-7 one of the voices abandons the
surface activity and retreats into a treble pedal while the others continue
with Rh2 in free style. As will be shown, these materialsthe melodic
figure M1, the rhythmic patterns Rh1 and Rh2, the homophonic texture
with weak/strong figures in the accompanying chords, and the modestly
polyphonic texture in complementary-rhythm patternare the building
blocks from which the entire prelude is constructed.
The second section begins with M1 in a varied and extended version
(S: m. 73-101). The bass, while moving in the same rhythmic pattern,
creates an independent pitch line, and the middle voices take up the
lengthened weak/strong/weak pattern. With the bass descent from B to B,
the harmonic development seems to remain in the tonic sphere. But the
final chord of this phrase is an inverted V/V and represents an imperfect
cadence in the key of the dominant. The sections second phrase begins by
combining the idea of the complementary pattern in double notes with a hint
of imitation (A+B: mm. 10-11, freely imitated by S+T). There follows a
cadential close conceived as a free adaptation of the rhythmic pattern. The
final measure of this section features a fifth voice, this time composed as
soprano 2, which contributes a traditional do-si-do figure.
The third section plays with the complementary pattern in double notes.
Its first phrase presents this pattern in simple imitation, without stretto
overlapping (S/B, A/T: mm. 13-15) while the second phrase places it in the
denser texture introduced in mm. 10-11, this time with soprano and bass as
2
In m. 22, bass 2 splits first to the third C/E, then to the triad A/C/E, only to resume its
former single-voiced sounding with the low F on the fourth eighth-note of m. 23.
292
WTC I/22
the two leading voices (mm. 15-17). The section tail, which like those of
the other sections is freer in style, sets out in m. 18 with a joint four-part
realization of Rh2 and ends by combining features from the first and
second sections (compare the sustained soprano and the descending bass in
the first half of m. 19 with m. 6, and the bass repetitions leading into an
imperfect cadence in mm. 19-20 with mm. 9-10). In the fourth section, the
varied recapitulation of the initial phrase accompanies the leading soprano
motif with parallels first in the tenor, then in the alto, and finally in both. It
gives up Rh1 for Rh2 at the end of m. 21 and ends in a many-voiced
fermata on a vii7 chord. The ensuing cadential formula once again alternates
the two rhythmic patterns while returning, both in the sustained soprano
and in the repeated bass 2, to the tonic pedal and to the double-third texture.
The dynamic rise and fall in such a meditative prelude should be much
smoother than in other pieces. It develops as a secondary feature and must
never lure listeners away from the all-pervading pulse. A rendering that is
both natural and unobtrusive follows the course of the ascents and descents
of pitchboth those in the overall design and in sudden leapswith very
gentle increases and decreases. As a result, the first section undergoes a
gradual tension increase (approximately from p to mp) up to m. 5, followed
by a short release, while the second section presents a gradual release
(approximately from p + to p ), followed by a mild rebuild of the original
piano color in the cadential formula. The third section begins with very
delicate touch in the sparse texture of its first phrase, but in its second
phrase engenders the strongest and steepest build-up of tension in this
piece. This leads to the first overall climax (preferably no stronger than mf)
on m. 162, from which the remainder of the section descends in gradual
diminuendo. The fourth section contains the second overall climax. Despite
its compact chord, it should sound introverted and intense rather than extraverted and loud. The prelude closes on a soft note, giving the impression
that it returns to the level of the beginning.
WTC I/22 in B minor Fugue

The question where the subject of the B minor fugue ends permits two
answers. The subjects trunk, which contains all the relevant harmonic
steps, ends at m. 31. The four quarter-notes that follow could be regarded
as an extension were they not used with such consistency throughout the
fugue. It is therefore appropriate to speak of a shorter and a longer version
of the phrase. Following a terminology used originally for distinguishing
B minor
293
metric endings in Greek poetry, these different shapes of the same trunk
are often referred to as male and female. The male version appears
condensed, concluding with the last essential harmonic step without any
softening addition. The female ending adds, in poetry, an unaccented
syllable after the final accent. In music, it extends the tonic reached in the
perfect cadence with notes passively remaining in the same harmony.
This regular extension is conceived as a smooth continuation, both in
rhythm and pitch pattern, of the second half of the main body: mm. 2 and
3 of the subject contain exclusively quarter-notes in stepwise motion. This
gentle curve appears in sharp contrast to the subjects beginning where two
half-notes followed by a downbeat rest create rhythmic tension. At the
same time, the falling perfect fourth, in itself a relaxed interval, is followed
(across the rest) by a minor ninth. This interval, rarely used for horizontal
progressions in Bachs time, expresses an emotional gesture of the highest
intensity. The question thus arises whether or not this exceptional melodic
jump is really meant to be perceivedand playedas an interval, i.e.,
whether the consecutive notes belong to a single structural unit.
Thus not only the ending but also the phrase structure of this subject
can be interpreted in two ways. In the first option the rest, together with the
unusually large distance between the two notes enveloping it, is regarded
as an indication of structural interruption, comparable to a comma in
verbal language. As a result, the subject consists of two subphrases that
differ considerably in all features: the first subphrase begins on a strong
beat and contains only two long note values in a falling fourth, while the
second subphrase sets out on a weak beat and features only shorter note
values in stepwise motion. In the second option, the rest is interpreted as
tension-sustaining, as a silent continuation of the melodic flow, and the
interesting interval is thus fully exploited. This minor ninth, whose tension
is enhanced by the fact that it spans across a suspended downbeat, then
becomes the subjects primary melodic focus.
The subjects harmonic background
can be depicted as shown in the example.
Modifications are frequent. In several cases,
i V i ivV i - - - - the male ending is raised a semitone and
harmonized not as a tonic but as a major
chord with seventh on the tonic root. This chord is heard as a dominantseventh and resolves correspondingly to the subdominant. The melodic
line supports this modulation by lowering the entire female ending one
whole-tone step, thus leading to the root of the target harmony. The fairly
frequent occurrence of this harmonization in the fugue provides additional
294
WTC I/22
proof that the extension of the subject, initially appearing as harmonically

redundant, is actually an intrinsic part of the whole: in these modulating
statements, the male ending does not provide the harmonic resolution but
needs the extra measure to convey a satisfactory sense of closure.
The subjects climax falls on G in m. 2. Even disregarding the melodic
power the preceding interval might exert, the G combines many tensionenhancing features. It is highest in pitch, harmonized as a subdominant,
and placed on the minor sixth of the scale, one of the natural leading-notes.
While this climax is followed, in whichever interpretation of the rest
preceding it, by a gradual diminuendo (with a steep decrease up to the
male ending and a gentler further release in the female ending), the shaping
of the notes preceding the G depends on the notion of the phrase structure.
Performers assuming two subphrases will conceive the F in the middle of
m. 1 as a conclusion of the first structural unit, thus playing it relaxed and
soft. The ensuing G will then appear as a suddenly powerful new beginning.
In this reading, the subject comprises two diminuendos: one, launched by
the initial note, short and steep; the other, beginning on the powerful second
climax, in gently lessening intensity. Conversely, performers reading the
subject as a single, indivisible phrase will take the initial descending fourth
to fulfill a different task: as the anchor of the high-tension interval it will
be regarded as geared forward and incorporating a much higher degree of
intensity than the initial B. In this interpretation, the subject thus consists
of a powerful dynamic build-up from B through F and the rest to the high
G, followed by the gradual release in the quarter-notes.
The fugue features 22 subject statements, presented in five voices;
v1 to v5 refer to voice 1 to voice 5 respectively, counted from top
to bottom. Entries containing the above-mentioned variation of the female
ending and thus effecting a modulation are marked with an asterisk, while
statements omitting the female ending altogether appear with a minus (-).
1. mm. 1-4 v1
8. mm. 29-32 v4* 15. mm. 53- 56 v4
2. mm. 3-6 v2
9. mm. 32-35 v5* 16. mm. 55- 58 v2
3. mm. 10-13 v3 10. mm. 37-40 v2* 17. mm. 55- 58 v3
4. mm. 12-15 v4* 11. mm. 46-48 v3- 18. mm. 67- 70 v1
5. mm. 15-18 v5* 12. mm. 48-50 v5- 19. mm. 68- 71 v2
6. mm. 25-28 v1 13. mm. 50-52 v1- 20. mm. 682-71 v3
7. mm. 27-29 v2- 14. mm. 502-52 v2- 21. mm. 69- 72 v4
22. mm. 692-72 v5
B minor
295
The subject appears both in stretto and in parallel. The stretto settings
involve two essentially different cases of premature succession. There are
entries in which only the female ending of the leading statement overlaps
with the subsequent beginning (see mm. 3, 12, 27, 55). In other cases, the
distance is as close as possible, i.e., the second subject note in one voice
becomes the initial note in the ensuing statement (see once in m. 50, twice
in m. 68 and twice in m. 69). In the light of the latter, the parallel entry in
mm. 55-58 might appear as an even closer stretto.
In addition to the modifications caused by modulation, the subject
undergoes changes of melodic shape in connection with its tonal answer,
where the original falling fourth is augmented to a fifth and the ensuing
large interval spans a minor tenth (see, e.g., mm. 3-6). While this tonally
determined lowering of the subjects second note remains without effect on
the close on the tonic, some cross-combinations of these intervals occur
that do in fact alter the harmonic background of the statement. In mm.
25-28, the combination of an initial falling fourth with an ascending minor
tenth places the subjects tail one note up and causes a modulation from D
major to E major; conversely, in mm. 53-56 the falling fifth is combined
with the ascending minor ninth, thus engineering a modulation from B to
E. A last-minute tonal adjustment can also happen by way of a variation
in the female ending (see the final ascent to E in mm. 12-13).
The B-minor fugue features no counter-subject. This may be partly due
to the unusual female ending frequently serving as a contrapuntal accompaniment to the beginning of the second or later subject entries in a group.
The fugue comprises six subject-free passages.
E1 mm. 6-9
E4 mm. 40-46
E2 mm. 18-24
E5 mm. 58-67
E3 mm. 35-36
E6 mm. 72-75
Two of the episodes, E2 and E5, develop from the preceding statement
by sequencing the subjects second and third measures (see v5: mm. 18-20
and, in a more modified version, v2/v3: mm. 58-60). E1 and E4 contain a
motif that consists of an inverted-mordent figure ascending to a dotted
half-note on the downbeat. A genuine episode motif, M1 is sequenced and
imitated. It not only fills the entire subject-free passage but even extends
into the beginning of the next statement (see v1/v2: mm. 11; v5/v2: mm.
42-47; note that v2 is the uppermost of the three voices involved in these
measures). The two remaining episodes, E3 and E6, are essentially nothing
but extended cadential closes. E3 features a short cadential-bass pattern and
a longer dosido formula in v1, while in E6 these figures are of opposite
length (see v5 in whole-notes: mm. 73-75, v1 in half-notes mm. 74-75).
296
WTC I/22
Moreover, the same combination of cadential traits can also be found at the
end of E2 (mm. 23-25). The structurally analogous E5 does not show this
cadential close. Instead, this close appears in the context of a full-fledged
subject statement, in mm. 53-55, where it is easily overlooked.
As these details reveal, the episodes in this fugue are arranged in a
symmetrical pattern of structural correspondence: E1 relates to E4, E2 to
E5, and E3 to E6. With regard to the dynamic outline in each, E1 and E4
with their ascending sequences of M1 describe an increase in tension and
thus lead from the preceding subject statement onward into the subsequent
one. E3 and E6, on the other hand, conclude a structural section. In the two
intermediate episodes the overall pitch tendency is very pronounced. In E2,
the initial sequence from the subject tail begins lowerand therefore
probably softerthan the end of the preceding entry. At the same time, the
first voice completes a descending line begun immediately after the disappearance of M1 and continued through two entire entries, in ever larger
note values (v1, mm. 11-20: C-B-A-G-F-E, D-C-B, A). This long
descent is countered by sudden ascents in all five voices (see particularly
mm. 20, 21) that discontinue their build-up only near the cadential formula
at the end of the episode. In the structurally corresponding E5, these lines
seem reversed. The initial sequences from the subject tail begin in an
elevated position (mm. 58-59) but are then followed by descending lines.
The most prominent of these is the one in the uppermost voice which, as in
E2, concludes in whole-notes (v1 mm. 62-67: G-F-E-D-C-B-A-G-F).
This episode thus places its climax at the beginning and subsequently
conveys an almost complete release, so extended that the anticipation for
the events to come is heightened all the more.
The basic character of this fugue is rather calm, not only because of the
complex rhythmic pattern with its many syncopations and tied notes, but
even more so because of the combination of stepwise motion with high
melodic tension. Yet the tempo should not be slow since Bachs time
signature indicates a pace in half-note beats. The appropriate articulation
demands legato in all note values, with the exception of the cadential-bass
steps mentioned above. There is one detail, however, that depends once
again on the interpretation of the subjects phrase structure. For performers
who conceive the subject as consisting of two subphrases, the falling fourth
is the only larger interval; it may therefore be played legato, particularly
since it comes in strongly ebbing dynamics. By contrast, performers who
regard the subject as indivisible are faced with two consecutive large leaps
both of which express an active increase in tension. These should then be
detachedthe first by articulation, the second by the written rest.
B minor
297
Owing to the fugues alla breve indication, the relative tempo of the
prelude to the fugue is most convincing if rendered in the complex
proportion of 3:2, where an assumed eighth-note triplet in the prelude
corresponds with a quarter-note in the fugue. (Approximate metronome
settings: prelude beats = 40, fugue beats = 60.) The fugue contains only one
ornament, the trill in bar 49. It begins on the upper neighbor note, shakes
in three pairs of sixteenth-notes and ends with a two-sixteenth-note suffix
leading into its resolution on the ensuing downbeat.
Many of the observations made above in the discussion of the episodes
encourage an exploration of a possible structural symmetry in the fugue.
This symmetry does in fact exist; it is complex but striking. Here are the
relevant details:
first half (mm. 1-36)
second half (mm. 37-75)
S
in v1, v2
S
in v2
E1 with ascending M1
E4 with ascending M1
S
in v3, v4, v5
S
in v3, v5, v1/v2, v4
E2 with gradual ascent
last entry with cadential close
ending in cadential close
S in v2+v3
E5 with gradual descent
S
in v1, v2, v4, v5
S
in v1/v2/v3/v4/v5
E3 with cadential close
E6 with cadential close
Within each of the two structurally corresponding halves, the entering
order of the voices, together with the sense of closure conveyed by some of
the episodes, allows two sections to be distinguished. In the first half, the
layout is straightforward: section I comprises five subject statements linked
by a bridging episode (E1) and rounded off by the cadential formula at the
end of E2. Section II consists of four entries and the short E3 with its
cadential close. As to the structural boundaries in the second half of the
fugue, two interpretations are conceivable.
In the first concept (which accords with the rules derived from
other Bach fugues), section III, like section I, contains five subject
statements, one of which is a stretto. They are linked by a bridging
episode (E4) and closed by the cadential formula accompanying the
last entry. Section IV then consists of a parallel statement, an episode
(E5), the impressive five-part stretto and a short final episode (E6)
with cadential close. The textural density supports these findings:
sections II and III both begin with only three of the five voices
taking part (see mm. 25-28: v4 and v5 resting; mm. 37-45: v1 and
v3 resting), and at the beginning of section IV, after a full five-part
cadential close in mm. 53-55, v1 pauses for twelve measures.
298
WTC I/22
The second concept builds on the admittedly strong tendency of
relaxation in E5, on the structural analogy of this episode with E2
in the fugues first half (which serves as the final segment of the
first section). Moreover, the fact that the v4 statement in mm. 53-56
overlaps, in its female ending, with the ensuing parallel entry makes
it more difficult for both performer and listener to perceive a
section ending and new beginning here. In terms of the dynamic
development, this concept is much easier to realize. In structural
terms, however, it not only leaves the fourth section with a single
group entry as opposed to two (one parallel, one stretto), but also
results in a somewhat unbalanced overall picture.
B minor
Concept I: sections
I
mm.
1-24
no. of bars 24
299
II
25-36
12
III
37-54
18
IV
55-75
21
Concept II: sections

I
II
III
IV
mm.
1-24
25-36
37-67
67-75
no. of bars: 24
12
32.5
8.5
The harmonic outline of this fugue underpins the structure with its four
sections in two halves, and with the third and fourth sections as indicated
above in concept I. The initial five subject entries all relate to the home
key B minor, in the usual alternation of tonic and (minor) dominant. The
second section establishes the subdominant region in its first and last entry;
the intermittent statements in B therefore function not as a return to the
tonic but as the dominant of the subdominant. The third section introduces
the major-mode version of the subject in its initial statement but soon
returns to the subdominant and its dominant. Finally, the fourth section
returns to the home key, with the parallel entry in the minor dominant and
the five-part stretto in the tonic. (The five subject entrances in the stretto
pick up the five initial entries in the fugue.)
Within the first section, the tension grows steadily. This is due, on the
one hand, to the increasing number of voices from one subject statement to
the next, and, on the other hand, to the above-mentioned rising pitch
outlines in both the bridging E1 and the concluding E2. Similarly, the
texture in the second section expands from three to five voices, adding
lower registers as the section progresses. The process, however, is more
restricted as regards the growth in the number of voices (3 to 5 vs. 1 to 5)
and much shorter in duration (12 vs. 24 measures). Thus the impression of
increase is not quite as powerful as in the initial section.
The third section not only begins, as the previous one, with two voices
resting, but is also transposed to the major mode. The resulting expression
is much gentler, particularly since the large leap appears now as a major
tenthan interval with no exceptional tension. The bridging episode, the
two single statements and the subsequent stretto build up new intensity.
But the structural climax in the stretto is offset by the still incomplete
texture in these measures, and the final statement (v4: mm. 53-56),
although now in five-part texture, is not a convincing climax either, both
because of its weaker position as a single entry and because of the relaxing
tendency in the accompanying closing formula. Spanning across the
confines of this section, this statement is easily heard as conveying the
anticipation of something significant yet to arise. It is therefore important
300
WTC I/22
to notice that Bach postponed the full five-part texture in this section until
this entry, and that, by contrast, the subsequent parallel statement sounds
in four-part setting only, as does the long E5, with the upper voice resting.
(This gives rise to a very sensitive question for performers: how does one
convey the absence of the highest voice when one is in fact playing high
pitches?)
The fourth section is thus, under several aspects, the weakest one. It
comprises only two (group) entries, its build-up of texture is smallest (only
from four to five voices), and its episode is the only one in the entire fugue
to constitute a protracted decrease of tension. In this regard, the fourth
section continues the gradual decline of dynamic power from one section
to the other. When the final five-part stretto counter-balances this overall
decline by erecting a powerful super-climax, this should come as a true
surprise to listeners.
WTC I/23 in B major Prelude

The material of this prelude derives almost entirely from a short
ornamental figure introduced in the first half measure. There is only one
instance in the piece where this motif is momentarily absent. The prelude
thus belongs to the motivically determined compositions. Its texture is
polyphonic, with three parts to which Bach adds a fourth only in the final
measures (after a splitting of the upper voice in m. 163).
The first harmonic closure occurs at m. 23 where an F7 chord, suggested
in the upper and middle voices above a sustained tonic pedal, resolves onto
the tonic. The tonic root is silentfading in the sustained lower voice
and implicit in the middle-voice rest as a melodic continuation of the rising
line. This cadential close is not structurally decisive as the harmonic
progression coincides with the first melodic phrase built by the motif, its
two sequences, and a concluding note. Moreover, neither of the other two
voices has yet taken part in the polyphonic display of material, nor has the
bass even begun to participate in any development. The first structural
caesura appears at the subsequent perfect cadential close. As the frequent
Es from m. 3 onward indicate, the prelude modulates to its dominant F
major, where the progression concludes at m. 61. This cadential close is
firmly established in mm. 5-6 with a cadential-bass pattern (ii-V-I) and a
melodic dosido formula. Similar cadential features mark the next structural break, which occurs, after a modulation to the tonic relative (G minor),
in the middle of m. 10, as well as the return to the tonic in m. 15. (The bass
patterns in m. 4: C-F-B and in m. 13: F-B-E, both appearing to support
V/V-V-I progressions, are not conclusive since the upper voice remains
unresolved in each case.)
The prelude thus comprises three sections and a short coda.
I
mm. 1-61
tonic to dominant
II
mm. 6-103
dominant to tonic relative
III
mm. 103-151 tonic relative back to tonic
coda mm. 15-19
confirmation of tonic
There is only one short structural analogy: mm. 4-61 correspond with
mm. 13-151. Both include the above-mentioned melodically thwarted
cadential closes (in B and E major respectively, tonic and subdominant)
followed by full cadential closes in F and B major, dominant and tonic.
301
302
WTC I/23
The preludes basic character is lively, owing both to the ornamental

character of the principal motif and to the very even overall rhythmic
pattern. The tempo allows for a brisk pace in the quarter-notes. The 16thnotes should sound ornamental, i.e., without emphasis on each single note,
but by no means hasty. The articulation includes legato for the 16th-notes
and non legato for all other note values. In detail: a gentler, dynamically
shaped detached style in the melodic quarter-notes (mm. 1 and 6) and
equally melodic eighth-notes (mm. 5 and 103-12) should be distinguished
from a more neutral non legato touch in the cadential-bass patterns in
quarter-notes (mm. 3-4, 5-6, 13, 14-15, 18-19) and eighth-notes (m. 10).
Longer note values that demand absolute legato appear in the do-si-do
formulas (see U: mm. 52-61: F-E-F and 184-193: B-A-B, M: mm. 94-103:
G-F-G, and even the truncated one in U: mm. 142-144: B-A-rest).
The question whether or not phrasing before or after the principal motif
should be expressed by a slight cut in the sound flow arises mainly in two
cases: if the motifs after-beat beginning is preceded by an on-beat note
belonging to a melodic line (as in L: m. 43 and M; m. 53) and if the principal motif, originally only seven 16th-notes long and omitting the strong
beats, is extended to include a final on-beat note before it is sequenced (as
in L: m. 63 and U: m. 73). Stark leaps between notes are certainly
inappropriate in this piece, but for performers with good skills in subtle
articulation, slight interruptions constitute a more plausible solution than
simple legato continuation.
The principal motif consists of seven 16th-notes circling around a
central pitch (B in the first half of m. 1, etc.). The circling motion
resembles an inverted turn and is followed by a repetition of the three
rising notes, thus ending one note above its center. Harmonically, this
ascent is already established in the middle of the motif, on beat 2, since the
step to the higher pitch, accompanied by a step upward in the middle
voice, represents a progression also on the harmonic level. The motif is
thus not just self-centered but comprises an active step. In terms of melodic tension, this active step is expressed in a delicate crescendo through
the initial notes up to the middlethe note that is also metrically the
strongest in this group. The three notes completing the motif provide the
complementing diminuendo, so that the end is as soft as the beginning.
Delineating the motifs tension-curve clearly is essential as it otherwise
risks coming out as mere finger work due to its short duration in relatively
fast tempo. Moreover, the strong-beat 16th-note Bach later occasionally
adds traps performers into an awkward accent on this extra note, thus
distorting the motifs original shape. Apart from this passive extension, the
B major
303
principal motif undergoes only two modifications. Four times, a slight

change in one of the intervals (see U: mm. 22 and 122 as well as L: mm. 51
and 141) eliminates the ascent characterizing the original shape. Moreover,
inversions convey a resolving tendency. They are almost exclusively found
in the coda (six times in mm. 15-18, only once already in m. 12).
The principal motif is complemented by two memorable note-groups
that recur and thereby form part of the thematic material. One is the ascent
in quarter-notes introduced in M: m. 1. It describes a gentle dynamic rise
that strings the principal motif together in one overall direction. This
quarter-note motif recurs in the first measure of section II (M: m. 6) and in
double notes in the coda (m. 17). The other secondary motif graces only
section III. It moves primarily in eighth-notes (the three motifs thus each
favor a different note value) and is conceived as a question + answer pattern
of the outer voices (mm. 10-11, U: D-G-F-E and L: C-B-A-G-F;
sequenced in the subsequent measures). One very convincing way of
molding this motif pair is to choose complementing dynamics for the
complementing segments: crescendo for the upper-voice question and
diminuendo for the lower-voice answer.
The development of tension in each of the sections is fairly straightforward as it largely follows the overall pitch outline. In section I, an initial
increase of tension (mm. 1-23) is followed by a long decrease (mm. 23-6).
In section II, proportions are reversed with a much more extended increase
(mm. 6-93) and a short relaxation (mm. 93-103). Section III begins in the
minor mode and with a new motifboth reasons to create some color
contrast (mm. 103-123), followed then by a long decrease that corresponds
to that of the first section (mm. 123-151). In the coda, the tendency to relax
is thwarted by three features: the replacement of the principal motif by a
figure with an ascending-scale component (m. 153-161), the denser texture
after the upper voice has split, and the recurrence of the ascending
quarter-note motif, now in double thirds. The prelude thus ends in an
assertive mood.
WTC I/23 in B major Fugue

The subject of the B-major fugue is two measures long. It begins after
an initial eighth-note rest with a long upbeat and ends after an ornamented
C, the representative of the dominant harmony, with a return to the
keynote on the downbeat of m. 3.
304
WTC I/23
The pitch pattern consists almost exclusively of seconds, interrupted

only for the interval CF. This leap of a perfect fifth is not of expressive
quality, and thus does not entirely match the mood of the stepwise motion
around it. Closer inspection of the subjects pitch pattern provides an
explanation: the fifth represents a change in pitch level rather than an
interval in a continuing line. The initial ascent, launched from the keynote,
breaks off with a quarter-note on the first strong beat, only to start afresh
from the lower F and climb even higher. The subject thus consists of two
subphrases that relate to one another in such a way that the first appears as
an abandoned attempt of what the second then completes in a more
powerful format. A look at the rhythmic design confirms this assumption.
Considering that the longest note value, the half-note C, is ornamented by
a trill and thus sounds very animated, the quarter-note C does in fact
constitute the most powerful rhythmic interruption, marking the point
where the regular motion comes to a halt before a fresh start.
The subjects harmonic background is difficult to determine as Bach
harmonizes it differently in almost every statement. It seems safe to claim,
however, that there is a twofold progression, with an interrupted cadence
either on the downbeat of the
second subject measure (as,
e.g., in m. 6) or slightly later
(as, e.g., in m. 30), and the
perfect cadence taking place
in the two final notes.
In a subject with two subphrases one must obviously expect two
climaxes. As it was already established that the second subphrase completes the aborted efforts of the first, the balance between the two subphrases does not require further decisions. The first climax is easy to
determine since the fourth note C is supported by the melodic rise from an
ornamented B, the harmonic movement from the tonic to the dominant, the
rhythmic value of a quarter-note, and the metric position on a (relatively
strong) middle beat. Determining the climax of the second subphrase is not
quite as easy. The contour reaches its peak on E, a note that is neither
harmonically nor rhythmically or metrically supported. By contrast, the
ornamented C represents two important steps of the cadence, apart from
being in a rhythmically and metrically stronger position than the off-beat
eighth-note E. As a result, a climax on E will give the subject (and, with it,
the entire fugue) a more virtuoso touch by stressing a superficial feature (a
high note), while a climax on the trilled C gives the subject more depth by
emphasizing its structural traits.
B major
The fugue comprises twelve subject statements:
1. mm. 1-3
T
7. mm. 18-20
2. mm. 3-5
A
8. mm. 20-22
3. mm. 5-7
S
9. mm. 21-23
4. mm. 7-9
B
10. mm. 24-26
5. mm. 11-13
T
11. mm. 29-31
6. mm. 16-18
A
12. mm. 31-33
305
Sinv
Ainv
B
T
A
S
In the course of the fugue, the subject undergoes two kinds of modifications, one at the beginning and the other at the end. The initial intervals
of both subphrases are enlarged from seconds to thirds in all tonal answers
(mm. 3, 7, 31). In the inverted answer, only the opening step of the first
subphrase is enlarged (m. 20). The final resolution is delayed (mm. 7, 31),
diverted (mm. 18, 20), or omitted (m. 22). In one instance, the subject
ending is varied but arrives in time at the proper resolution (m. 26). In
other words, only four entries are quite original: nos. 1, 5, 9, and 12.
Parallel statements do not occur; nor do true strettos in which a substantial segment of one subject entry overlaps with another. The only
instance where an entry begins at less than two bars distance from the
beginning of the previous one occurs in m. 21, i.e., in connection with the
one entry in the fugue that omits the final resolution. Thus the process one
is hearsas opposed to what one sees in the scorecan be said to be
concluded by the time the subsequent statement enters.
Bach invents one counter-subject for this fugue. It is introduced against
the subject answer in mm. 3-5. Beginning two eighth-notes later than the
subject, the counter-subject also displays two subphrases separated by a
change in pitch level (see m. 34). The phrasing occurs one eighth-note after
that in the subject. Moreover, in its first appearance the counter-subject
even ends with a metrically delayed resolution, i.e., one eighth-note after
the subject (m. 51 E-D). Besides this similarity in phrase structure, the
counter-subject is also
related to the subject
in its pitch pattern as it
shows an overwhelming participation of
seconds and a scalar
ascent in the second
subphrase. Its dynamic
306
WTC I/23
independence, too, is limited. Neither performers who outline the descent

in the first subphrase with a diminuendo, thus continuing the tension
decline across the cut between the subjects two subphrases, nor
performers who express an active gesture in a downward crescendo, thus
imitating that in the subjects first subphrase, will achieve much individuality, for in the second subphrase, both pitch and rhythm favor the
syncopated highest notewith the result that this climax very nearly
coincides with that of the subject. To conclude: there is in this fugue little
challenge for the subject. Moreover, the counter-subject is not a very
faithful companion. In its complete range it recurs only three times (see
A: mm. 5-7, S: mm. 7-9, and A: mm. 31-33). In addition, the second subphrase appears once without the first (see A: mm. 12-13), as does an even
more truncated second half of the second subphrase (see S: m. 17).
The fugue comprises five subject-free passages.
E1 mm. 9-113
E3 mm. 233-241
E2 mm. 133-161
E4 mm. 26-291
E5 mm. 33-34
The first episode introduces the listener to several motifs that must be
regarded as genuine episode motifs as they are unrelated to both subject
and counter-subject. Closer inspection reveals that they all derive from a
single common source. The different versions share the shape of a concave
curve in which the longest note, providing both the harmonic and the
dynamic climax, falls on the lowest pitch. The example below shows three
versions of this motif, all found in mm. 91-111, 133-153, and 261-281:
M1
M1a
M1b
In the three more substantial episodes, M1 invariably provides the

opening. In each case, it is accompanied by M1a in such a way that the two
run in parallel thirds or sixths throughout their tail. Also in each of the
three cases, M1 is sequenced by the variation M1b, which is then followed
by an extra tail. Before this sequence and in stretto with the original M1,
there appears a variation in which the quarter-note is replaced by two
eighth-notes in octave displacement. (In E2 m. 14, this octave leap even
involves a swap of voices from tenor to bass.) All three episodes then end
with separated tails and heads of the motif. Both the short E3 and the
final E5 only feature the head of M1b (see B: mm. 23-24, and T: mm.
33-34). In the former, the two upper voices continue with sequence and
B major
307
imitation drawn from the sopranos fragmentary counter-subject figure, so

that this episode appears more like an extension of the preceding subject
entry, serving, one might say, to re-establish the subjects original metric
position after the premature entry of the tenor in m. 21.
As regards the relationship between the episodes in this fugue, there is
a notable analogy between E1, E2, and E4. The remaining two episodes,
while both featuring the same segment from the motifs variation, fulfill
structurally different purposes and should therefore not be interpreted as
related. The motivically determined episodes bring about a significant color
contrast to the subject statements. Each begin with a slight diminuendo in
the descending sequence of M1b and end in a gentle crescendo in the
ascending sequences of the head. By contrast, the two shorter episodes
are integrated into the level of the primary materialone bridging between
two subject statements, the other providing the final relaxation, but both
without much dynamic development of their own.
The predominance of stepwise motion, coupled with the variety of note
values, determines this fugue as a piece in rather calm basic character. The
pace is serene but not hesitant. The tempo proportion between the prelude
and the fugue had best be complex, owing to the fact that both pieces are
in common time and based on the same note values. A simple proportion
would thus create an effect of dullness in the succession of the two pieces.
What works well is a rendition in which three quarter-notes in the prelude
correspond with two quarter-notes in the fugue. (Approximate metronome
settings: prelude beats = 108, fugue beats = 72.) The overall articulation is
legato. (Exceptions occur in the cadential-bass patterns of mm. 13, 17-18,
and 33-34, as well as in other consecutive quarter-note or eighth-note
leaps, e.g., m. 24: B, m. 31-32: T.)
The trill in the subject poses some problems. It is an integral part of the
thematic phraseso much so that playing the half-note without an
ornament would sound extremely dry.1 As the trill is approached in
stepwise motion, it begins on the main note. As the fastest note values in
this piece are 16th-notes, it shakes in 32nd-notes. And as its resolution
appearsso far as it appears at allon one of the two strong beats in the
reigning 4/4 time, it must end with a suffix. This long trill thus contains
altogether fifteen notes, to be played very regularly and, what is even more
1
This is certainly the impression we get when we play the voice alone. It is also what any
string or wind player, performing this fugue in a quartet, would feel. Only pianists, busy
with figures in other parts but still in their own ten fingers, occasionally ignore the singlevoice demands, with the excuse that there is already enough happening at this point.
308
WTC I/23
often neglected, with dynamic shading that reflects the decrease of tension
in the subject at this moment. The problems that arise with the trill in this
fugue are caused by several irregular endingsthose that were earlier
mentioned as statements with delayed resolution. In all of these cases,
the trill begins in the same manner but ends prematurely and without a
suffix, stopping short on the last main note before the bar line.2 To provide
performers who now decide against playing this fugue (just because of the
trills) with an incentive for trying, the following examples give the most
prominent occurrences spelled out, with some suggestions for appropriate
fingering in the trickier cases:
mm. 2-3 (similarly in mm. 4-5, 8-9)
mm. 30-31
mm. 12-13
mm. 32-33
mm. 6-7
mm. 19-20
The only statements that forgo the trill altogether are those with omitted resolutions or
varied endings, i.e., those in mm. 17, 21, and 25.
B major
309
When trying to determine how this fugue is structured, one cannot rely
on any of the data that normally guide such an analysis. Except for the
final measure, there are no explicit cadential formulas outside the confines
of the subject statements. The only obvious cadential-bass patterns mark
the endings of the fifth and sixth statements respectively. They thus follow
one another too closely to indicate section endings. The texture is unusual
insofar as only two among the twelve subject statements sound in full
ensemble. All other entries either feature one resting voice or maintain the
full four-part texture only for a short span. The harmonic design, too, is
atypical insofar as no noteworthy modulation takes. Except for one entry in
the subdominant, the statements alternate regularly between the tonic and
the dominant positions of the home key.
The only clues to determine the structural layout of the composition are
thus the analogies observed earlier in the design of the episodes. These
analogies can now be expanded:
4 consecutive subject
4 consecutive subject
statements (mm. 1-9)
.
statements (mm. 18-26)
(E3 regarded as an extension)
E1 (mm. 9-11)
.
E4 (mm. 26-29)
2 additional statements .
2 additional statements
interrupted by E2
.
rounded off by E5
As both the fourth statement in m. 9 and its counterpart in m. 26 close
with a perfect cadence while the ensuing episodes feature open endings
leading into the entry following next, we may assume that these episodes
open sections rather than close them, as in so many other fugues. The
result of these comparisons leaves us with four sections, of which
the first and third each consist of four consecutive subject statements and end in a perfect cadence (m. 9: F major, m. 26: C major),
the second and fourth each show an opening episode of analogous
material and design, two subject statements, and another episode
(incorporated into the second section but closing the fourth).
Concerning the dynamic design valid in the sections of this fugue, two
basically different patterns can be distinguished. The first and third sections
consist exclusively of subject statements. Within the first section, the
tension rises gradually along with the usual increase in the number of voices.
Owing to the fact, already mentioned above, that this section does not truly
establish the expected four-part texture, the growth should be restrained in
such a way as to avoid the sensation of a powerful climax. In the third
section, the two initial statements appear in inversion. Given the particular
shape of this subject with its two rising motions now converted to falling
310
WTC I/23
lines, the inversions sound much less cogent and self-confident than the
original. As a result, the four entries of this section also give the impression of curbed tension.
The second and fourth sections feature alternations of episodes and
subject statements. The decisive musical message in these sections is that
of color contrast. A very light and delicate touch in the episodes, conveying both a playful character and melodic openness, is set against a much
more assertive touch in the subject statements. In contrast to the episodes,
the statements are clearly directed toward their goal and assuredly closing.
Performers who wish to weigh the two entries in each section against each
other will find that in the fourth section, the final statement surpasses the
preceding one, both because of the rise in these consecutive entries from
the alto to the soprano and because of the temporary four-part texture. In
the second section, however, where the statements are separated by a
substantial episode, the question of any dynamic relation between the
statements seems beside the point.
WTC I/24 in B minor Prelude

This prelude is in three-part texture. The lower voice is conceived as a
thoroughbass in continuous eighth-note motion that is only given up at the
end of each of the two repeated sections. Meanwhile, the upper and middle
voices weave a polyphonic pattern with manifold imitations and occasional
freer contrapuntal passages. The principal motif with its rising fourth
followed by a syncopation on beat 2 and a stepwise descent is closely
related to the fugato motif in the E-major prelude:
compare Prelude in E major, mm. 10-11 tenor B-ED-C
with Prelude in B minor, mm. 1-2 middle voice F-BA-G.
The first harmonic progression concludes in m. 4. After a prominent
octave leap on the dominant note F, the lower voice ascends in a segment
of the melodic-minor scale to the keynote B at m. 41. At this moment, the
upper voice diverts its expected course with a leap to the F before
reaching the keynote on the second beat, which is also where the middle
voice belatedly resolves its suspension. This cadence marks the end of the
extended motifs initial statement in the upper voice. Particularly given its
overlap with the subsequent motivic statement, it cannot be regarded as a
structural caesura. The second harmonic close occurs at m. 73. The
cadential-bass pattern indicates a modulation to D major, the tonic relative,
supported by a typical closing formula in the soprano (D-C-D). With two
traditional formulas in the outer voices and the simultaneous resolution in
all three parts, this cadence qualifies as a structurally relevant close. The
16th-note figure in the upper voice, unique in this composition, makes it
possible to hear the third beat in this measure both as a resolution to the
preceding cadence and as the upbeat to the subsequent syncopation, i.e., as
the beginning of a new statement of the motif.
Further cadences in this prelude often make it difficult to distinguish
whether a harmonic close marks the end of a phrase or that of a structural
section. A progression of keys in the same order as the one just described
can be found in the second half of the prelude where a phrase-ending
cadence in D major at m. 211 is followed by a structurally more relevant
one in F minor at m. 273. After this, the sequence closes in E (m. 293)
the last perfect cadence for a long time.
311
312
WTC I/24
Strangely, the harmonic closes one would expect to be the most

straightforward, i.e., the ones before the repeat sign and at the end of the
composition, present themselves as evasive: On the downbeat of m. 16, the
traditional patternsthis time in the lower and middle voicesindicate the
return to the tonic B minor. Yet the next two measures with their growing
note values and eventual stop determine a definite structural break, depart
again from this tonic, and end in an imperfect cadence before the repeat
sign. Toward the end of the piece, the listeners anticipation of a return to
the tonic in the middle of m. 42 is deceived by the sudden twist to the
chord on VI. And even when at m. 463 the bass finally reaches the keynote
B, it takes quite a while for the higher voices to give in and settle.
This is a rather calm piece, with a complex rhythm including manifold
syncopations and distinct melodic tension between the notes. The tempo as
given by Bach, one of the few tempo indications in the Well-Tempered
Clavier, is Andante. The corresponding articulation is an overall legato.
Interruption of the continuous sound flow occurs only in the context of
phrasing. While the thoroughbass line in the lower voice is not conceived
with melodic qualities and should therefore be played without phrasing,
the upper and middle voices require careful structuring in order to fully
unfold their potential.
The score contains two ornaments, both located at the end of the first
repeated section (see L: m. 16, U: m. 17). The trill in the lower voice
resolves properly on the next downbeat and is therefore note-filling with a
suffix. As it is approached in stepwise motion, it begins on the main note.
Although 16th-notes do not play a major role in the rhythmic pattern of
this prelude, they are nevertheless the fastest note values to which the
speed of the trills shake has to relate. The appropriate note values in the
trill are therefore: an initial 16th-note, owing to the trills beginning on the
main note, followed by 32nd-notes. The ornament in the upper voice
designates a short mordent. It, too, sets in on the main note and contains
either three or, better, five fast notes (C-D-C or C-D-C-D-C).
The preludes first repeated section is built entirely from two motifs.
Both are introduced in the middle voice and imitated in stretto, with some
modifications, in the upper voice. M1 with its fourth interval ascending to
a syncopated half-note, followed by a quarter-note descending stepwise to
the subsequent downbeat, was already described. M2 begins with a
two-beat syncopation and descends in two eighth-notes to the next strong
beat. (That M2 does not, like M1, begin with an upbeat fourth may not
become quite clear in the first statement of m. 2; further appearances in the
prelude, however, confirm the beginning on the syncopation.)
B minor
313
The preludes first phrase comprises two group statements: the initial
middle-voice statements of M1 and M2 and their stretto imitations, starting
at half a measures distance in the upper voice, in which the final note of
M1 melts with the beginning of M2, and an M1 entry in the upper voice of
mm. 4-5 imitated now in the middle voice (ending rhythmically varied). A
further M1 statement, this time with a rhythmically varied beginning, can
be recognized in M: mm. 61-71. Its imitation in the upper voice bends its
final note back upward and thus creates a do-si-do formula. The second
phrase features a string of ascending M1 sequences in the upper voice that
follow one another in such a way that the final note of one simultaneously
serves as the beginning of the next. All are imitated after half a bar. The
same process can be observed, with the middle voice in the lead, in the
third phrase. After the perfect cadence in m. 16, the above-mentioned
extension to the imperfect cadence completes the section.
In the preludes second repeated section Bach introduces seemingly
new material. All of it is, however, related to the two earlier motifs. M3 is
a rhythmic variation and diminution of M1. Introduced in the middle voice
like the two earlier motifs, it consists of three eighth-notes leading to a
strong beat. The pitch pattern contains an ascending fourth followed by
descending steps. This motif reigns, extended through a few linking notes
in each voice, in the short fifth phrase of the prelude. M4 enters after the
D-major close in m. 21. It is rhythmically close to M3 with three eighthnotes leading to a strong beat, but its pitch pattern points in a single
direction, thus revealing its relationship to M2. The motif is introduced in
ascending direction in m. 21, but later recurs also in descent as, e.g., in
mm. 23 and 26. After a closely knit pattern with alternations of M4, M3,
and M2, the middle segment of the fifth phrase displays a pattern in
complementary rhythm with half-notes and syncopated half-notes (see
mm. 24-26) that closes with M2. The closing formula of this phrase, as
well as that of its modulating extension, presents yet another motif, made
up this time of the three ascending eighth-notes from M4 and a diminution
of the first motifs variation that sounds like the do-si-do formula. The new
motif, here called M5, also serves to indicate a cadential close (see M5 in
mm. 26-27 and 28-29, middle voice).
The sixth phrase, beginning after m. 293, sets out with motifs that are
by now familiar: in the upper voice, M2 gives way to a variation of M5
and partial sequences thereof, followed by a string of M4 sequences. The
middle voice eventually presents a sixth motif (M6 see mm. 324-333) that
is to play an important role in this phrase. Related to M2, it begins with a
syncopation followed by three eighth-notes leading to a strong beat. The
314
WTC I/24
direction, however, is changed to describe a curve, with the interval of a

third between syncopation and first eighth-note. M6 soon takes over all
active motion and from m. 36 onward appears in an almost continuous
stretto of the two upper voices. The phrase ends similarly to the preceding
one: a segment with complementary rhythm (half-notes and syncopated
half-notes) ending with M2 statements is followed by a closing formula
featuring M5 in the middle voice (compare mm. 41-42 with 26-27). After
the interrupted cadence in m. 42, the final phrase presents the last motif of
this prelude: M7 consists of a syncopated quarter-note followed by an
eighth-note that leaps up to the strong beat in alternations of the perfect
fourth, diminished fourth, and tritone. Introduced in the middle voice of
m. 42, this motif is then imitated and sequenced in ascending direction,
before giving way to the final closing formula.
As a conclusion one can state that Bach builds this prelude from seven
motifs, two of them initially opposed to one another and the five others
derived from the first pair. The piece thus conveys unity and development
in a unique blend. The preludes first section consists of three phrases plus
a cadential extension inroughlyan [a b b' c] pattern. The initial thematic
phrase introduces the first two motifs as well as the rhythmic pattern
(which, for the melodic voices, consists primarily of quarter-notes and
syncopated half-notes). The two phrases in the center of this section display structural analogies, particularly in their first halves. A short closing
formula leading to an imperfect cadence completes the section. The second
section comprises four phrases, laid out in a design that is roughly similar
to that described for the first section. The initial phrase introduces the third
motif with the diminution of the principal figure. At the same time it
announces a shift in the predominant rhythmic values, from quarter-notes
and half-notes to eighth-notes and quarter-notes. The two phrases in the
center of this section display structural analogies, particularly in their
second halves. A shorter phrase with a new motif and a different closing
formula completes the prelude. The second section should thus be regarded
as an intensified development of the first.
In terms of dynamics, the first section consists of a rounded initial
phrase with two gentle curves, two active phrases in which the level of
intensity is raised by means of ascending sequences (the climax falls on m.
14 in the middle of the third phrase), and an expansion that does not
provide the expected resolution owing to the interrupted cadence. In the
second section the first three phrases each begin with some emphasis
before they relax. Only the sixth phrase changes this dynamic direction.
After the anticipated cadence in m. 323 has been diverted, it surprises with
B minor
315
a conspicuous ascent not only in the two melodic voices but even in the
lower voice, which for once participates in the sequential patterns and even
describes a melodic contour in its metrically enhanced notes (see in mm.
32-38, U: G1 to B2, M: E1 to F2). This passage presents the climax of the
whole prelude. Its position in the structural layout corresponds with the
(lighter) first climax: one occurs in the middle of phrase III, the other in the
middle of phrase VI, the third phrase of the second section.
The final phrase is not restricted to concluding function but builds up
its own little climax at m. 453. In the metrically unusual cadential close, the
tonic is reached on the middle beat, while the two upper voices display a
distinct reluctance to give in to a final release of all tension. This generates
a close that can be interpreted either as hesitant or as fairly powerfuland
be played accordingly, with either a return to complete piano or a rich,
mezzoforte ending.
WTC I/24 in B minor Fugue

The subject of this fugue spans three measures. It begins after an
eighth-note rest and ends at m. 41. At first glance, this downbeat seems like
a melodic return to the note from which the statement set out. However,
the initial F is the fifth degree of the tonic B minor, whereas the same
pitch in m. 4 serves as the root of the minor-dominant harmony to which
the subject has modulated.
Pitch and rhythm within this phrase are quite unusual. With the exception of the ornamented penultimate note, the rhythm consists exclusively
of eighth-notes. Yet this rhythmic pattern is far from being characteristic of
the entire fugue, where 16th-notes, quarter-notes and a variety of tied notes
abound. The pitch pattern in the subject, particularly if considered for a
moment without Bachs slurs, includes a great number of large intervals
besides the two conspicuous broken chords, and an almost equal number of
semitone steps next to only two whole-tone steps (before and after the
trill). Again, this unusual combination of intervals is not shared by the
fugues other components. Outside the subject, stepwise motion in regular
diatonic progression prevails.
With regard to its phrase structure, this subject can be split into two
very unequal segments. Both begin with a falling broken chord, the second
to fourth eighth-notes in mm. 1 and 3. The first subphrase then winds its
way through six slurred note-pairs before coming to a halt on the next
316
WTC I/24
unpaired note, the B at m. 31. The second subphrase, much more concise,
comprises after the broken-chord upbeat only its target, the trill, and the
ensuing harmonic resolution.
When examining the subjects harmonic background it is crucial first
to determine the nature of the note-pairs. Slurred by Bach himself, each
pair acts out the relationship of appoggiatura-resolution. Having established
this, we know that only the resolutions are essential for the harmonic outline while the appoggiaturas create local harmonic relationships of secondary order. The example visualizes the two layers of harmonic events and
analyzes Bachs harmonization as found, e.g., in mm. 21-24 of the fugue.
B minor: i
F minor:
i V V/i7v
V/v7
V
7
V/ 7 V i iv 5 V i
V
In our search for the subjects dynamic outline, we want to look for
tension-enhancing features in each of the two subphrases before pondering
the relationship between the two climaxes. The highest degree of harmonic
tension within the initial two measures is reached in the chord marking the
modulation, i.e., in the C7 that determines the second half of m. 2. Within
this half measure, the D particularly captures attention. This note is exceptional in two respects. In terms of structure, it appears as the peak of the
ascending sequences (B-A, C-B, D-C). In terms of its scale degree, it
represents the sixth in F minor, thus serving as a secondary leading-note.
The climax of the second subphrase is the long G. This note is not only
much longer than all other values in the subject, but also incorporates the
two most essential steps of the target-key cadence: the subdominant (as a
six-five chord) and the dominant-seventh chord of F minor. Balancing the
two climaxes against each other, one discovers that the second represents
natural tension that is resolved immediately afterward, while the first
expresses artificial tension that, owing to the structural cut between B and
C, is released only indirectly. On a higher structural level one could thus
claim that the first subphrase creates a tension that is resolved in the
second subphrase.
This fugue contains thirteen full subject entries. Another seven, marked
with asterisks in the following table, are incomplete:
B minor
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
317
mm. 1-4
mm. 4-7
mm. 9-12
mm. 13-16
mm. 21-24
mm. 30-33
mm. 34-35
mm. 35-36
mm. 38-41
mm. 41-42
A
T
B
S
A
T
A*
S*
B
S*
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
mm. 42-43
mm. 43-44
mm. 44-47
mm. 47-50
mm. 53-56
mm. 57-60
mm. 60-63
mm. 69-70
mm. 70-73
mm. 74-75
A*
B*
T
B
T
B
T
T*
B
A*
The subjects most prominent modification is the abridged version,

particularly since the fragment Bach chooses is of exactly the same length
each time, breaking off after the third note-pair. Another expected change
occurs in the tonal answer, which adjusts the intervals in the initial broken
chord and in the first note-pair. Further alterations include the final resolution, which may be delayed (as in mm. 46-47) or omitted (as in mm. 1516, 55-56, and 62-63). True strettos are not used; all seeming combinations
feature an abridged entry in the lead and thus engender no significant
overlap (see statements nos. 7-8, 10-11, 12-13, and 18-19).
Bach has invented a single counter-subject for this fugue, introduced
against the second subject statement, in the alto part of mm. 4-7. Consisting of three segments that are often used separately, it is very versatile and
plays a vital role both as an accompaniment to the subject and, represented
by its components, in the episodes. CSa (m. 4: E to F) comprises two irregular ascents in 16th-notes, CSb (mm. 4-6: F to B) features a diatonic
descent in quarter-notes ending in a do-si-do figure, and CSc (mm. 6-7: C
to D) moves again in 16th-notes. While the counter-subjects initial statement is designed with CSa and CSb sharing one note (the F), Bach later
often separates the two segments and allocates them to different voices.1
All segments undergo modifications. CSa is shortened or lengthened,
inverted or even completely altered in pitch. CSb is frequently shortened at
the beginning and/or varied at the end, while CSc changes only its final
interval. (For a complete listing see the color graph on page 323.)
1
See, e.g., mm. 9-12 where an inversion of CSa appears in the currently highest voice
(which is the alto). The tenor follows with CSb + CSc. Similarly in mm. 13-15: an inversion of CSa is heard in the tenor, CSb + CSc in bass.
318
WTC I/24
Concerning dynamic shape, the first two segments (whether occurring

in the same voice or split into two) build one curve with a crescendo in
CSa and a gradual diminuendo in CSb. The third segment CSc creates its
own little build-up and relaxation within the few notes it comprises, due to
the very prominent ascent and subsequent fall of the pitch contour. The
counter-subject acts as a fairly regular companion to the subject; it accompanies (in a more or less complete version) the subject entries in mm. 4-7,
9-12, 13-16, 21-24, 30-33, 38-41, 44-47, 47-50, 53-56, 57-60 and 70-73.
Besides this counter-subject, there is another short motif that serves
several times to support the beginning of a subject entry. As this motif
materializes only in the context of a subject entry but never in an episode,
it must be regarded as a fragmentary second counter-subject and will
therefore be referred to as CSd. This motif is easily recognizable: with its
syncopation and ensuing 16th-note figures it constitutes a single relaxing
gesture. It is introduced in m. 21 (S) against the beginning of the fifth
subject entry, where its relationship to CSa becomes most obvious. It
recurs in mm. 34 (S), 35 (A), 38 (T), 41 (A) and 42 (S), almost exclusively
against incomplete subject entries.
This fugue contains twelve subject-free passages2:
E1 mm. 7-9
E5 mm. 33-34
E9 mm. 56-57
E2 mm. 12-13
E6 mm. 36-38
E10 mm. 63-69
E3 mm. 16-21
E7 m. 47
E11 mm. 73-74
E4 mm. 24-30
E8 mm. 50-53
E12 mm. 75-76
These episodes feature are only particles of the subject. Two are fake
entries: in m. 19, the alto (which is otherwise silent in mm. 173-211, i.e.
during almost the entire E3) presents the falling broken triad from the
subject beginning. It thus anticipates the true alto entry in m. 21. The same
is true for the fake tenor entry in m. 28, where the tenor, which had withdrawn in m. 21, resurfaces only in m. 30. In a third instance (see m. 16),
the tenor imitates the falling broken chord, the half-note and its resolution
from the subjects second subphrase, but with the pitch pattern of the
subject beginning (noticeable particularly in the sixth interval).
2
The incomplete subject entries have here been counted among the essential subject statements, both because of their material and because they appear accompanied by countersubject segments. This is, of course, a matter of interpretation. It is theoretically equally
possible to regard these incomplete statements as subject-related episode material. This
would lead to slightly different results in the counting of the episodes:
E4
mm. 24-30
E6
mm. 41-44
E11
mm. 73-74
E5
mm. 33-38
E10
mm. 63-70
E12
mm. 73-76
B minor
319
The counter-subject is all the more active in the episodes, contributing

its third segment (CSc) to each and every one of them. Even the extremely
short E7 features a partial sequence of CSc (S: m. 47), and the episodes E2,
E5, E6, E7, E8, E9, E11, and E12 are determined exclusively by this
16th-note figure.
In the first part of the fugue the CS segment is complemented by three
genuine episode motifs. M1 is introduced in E1, in imitation between alto
and tenor, and recurs in E4, in parallels (see S+A: mm. 24-26). It consists
of three eighth-notes leading to a longer note on the strong beat. M2 is first
heard in E3 (in the bass part of mm. 17-21) and recurs, likewise in the
bass, in E4 (mm. 26-30). It is very similar to M1: rhythmically it also
consists of three eighth-notes leading to the strong beat, and its pitch only
differs in the first note (an ascending instead of a descending step). M3 is
also presented in E3. From the first sequence onward a curved shape is
established (see S: mm. 17-18 E-D-C-D-E) that then remains consistent.
This motif consists primarily of 16th-notes. Its gentle climax falls on the
strong beat in the middle of the respective curves.
None of the episodes equals a disguised cadential close. In fact, none
of them contains any obvious cadential features. Close analysis is needed
to reveal the harmonic closes that do occur. When examining the episode
endings one finds that only E5 with its half-measure modulation and E12
end with a complete harmonic close. All other episodes feature one voice
with a suspension that is only resolved at the beginning of the subsequent
subject entry, thus preventing a feeling of closure. In addition to the two
closes coinciding with the end of an episode there are two others that
materialize at mid-point. In both cases, this occurs after a subject entry that
remains somewhat open. In m. 16, the subjects soprano entry ends
unresolved. The subject-derived fragment that follows in the tenor at the
beginning of E3 determines the harmonic background as F minor, and this
keythe minor dominant of the home keyis confirmed with a perfect
cadence in the first half of m. 17. Similarly, the tenor entry in mm. 60-63
ends without resolving its trill. Instead, the ensuing episode E10 contains
a cadential close in the home key on the third eighth-note of m. 65. In the
course of further examination it will thus be meaningful to distinguish the
following segments: E3a + E3b (mm. 16-173-21) and E10a + E10b (mm.
633-653-70). And seeing that two of the episodes consist of two structurally
detached portions, one should go ahead and have a closer look at the
others. There is in fact a third episode that can be thus divided: E4 also
consists of segments E4a and E4b (mm. 24-26-30).
320
WTC I/24
With regard to the relationship between the episodes, it is easy to

discover that E3b, E4b, and E10b are very much alike. All feature M2 in
the lowest voice and M3 in an imitation pattern of the two higher voices. A
more remote analogy can be found between E1 and E4a, both of which
contain M1 in connection with CSc (though the pattern does not quite
correspond).
The role each episode plays in the tension development is as follows:
E1
crescendo: ascending sequences
E2
diminuendo: descending sequences
E3a
closing
... E3b diminuendo after a new start: descending sequences
E4a
... E4b diminuendo: descending sequences
E5
closing
E6
diminuendo: descending peak notes
E7
almost like an extension: releasing
E8
self-contained, with crescendo / diminuendo
E9
resolving previous S, building up anew
E10a
closing
... E10b diminuendo: descending sequences
E11
E12
closing
The heading Largo indicates a calm character and a slow tempo. The
tempo proportion between prelude and fugue should be complex since both
pieces are in calm four-four time and simple proportion would therefore
give a monotonous result. There are two possible solutions: Performers
may equate three eighth-notes in the prelude with a quarter-note in the
fugue or an assumed triplet quarter-note in the prelude with an eighth-note
in the fugue. The approximate metronome settings could be the same for
the prelude beats (= 64), whereas the pulse in the fugue would beat at 42 in
the first case and at 48 in the second.
The articulation is legato in all notes with the exception of cadentialbass patterns and consecutive leaps. These exceptions, however, are many.
In the subject, the notes in the two broken chords should be played gently
detached, and the slurred note-pairs should also be very gently separated
from one another, not only because this allows the appoggiatura-resolution
structure in the pair to become clearer, but also because of the underlying
pattern of consecutive leaps (refer back to the melodically reduced version
given earlier). By contrast, the counter-subject segments and the three
episode motifs all require unbroken legato.
B minor
321
The only ornament in this fugue is the trill on the subjects penultimate
note. This trill always begins on the main note as it is approached stepwise.
After an initial note of 16th-note duration it moves in 32nd-notes (i.e.
twice as fast as the faster note values in the piece). Wherever the subject
ends with the resolution falling on the subsequent strong beat, the trill is a
note-filling one and ends with a suffix. The trill constitutes an integral part
of the subject. In this fugue as in many others, Bach does not add the
ornament symbol in later subject statements, just as he eventually stops
indicating the slurs. Contemporary performers could be trusted to retain all
the subjects characteristic features, particularly articulation and ornaments,
in all further statements without renewed reminders from the composer.
Regarding long trills, the decisive word is: where musically applicable.
It is necessary to study the melodic and harmonic surroundings of a
subject-ending (and not just the technical feasibility) to find out where a
trill is needed, where it may have to be modified, and where it should be
omitted. The three-part rule of the thumb is: (a) if the originally ornamented note is resolved on time, play the original trill; (b) if the originally
ornamented note is resolved early or belatedly, play an interrupted trill
without suffix, stopping short immediately before the bar line (in case of a
delayed resolution) or before the dot (in the case of an anticipated resolution); (c) if the originally ornamented note remains unresolved, no trill is
warranted. Moreover, this fugue features two instances where a complete
execution of the ornament is thwarted when the subjects contour is crossed
by a line from another voice. (In these cases, the trill may be pianistically
impossible but would be played in, say, a string-quartet rendition.) These
cases are marked with an asterisk in the table below. The conclusions of
the trill in subject statements are thus as follows:
m. 6 (a),
m. 11 (a), m. 15 (c), m. 23 (a)*,
m. 32 (a), m. 40 (a), m. 46 (b)*, m. 50 (a),
m. 56 (c), m. 60 (a), m. 63 (c), m. 72 (a).
The structural layout of this fugue does not reveal itself as easily as
that of others. For one thing, the texture is quite peculiar: although the
fugue is written in four voices, only two of the thirteen complete subject
entries appear in four-part setting: entry no. 4 in mm. 13-16 and the final
complete entry in mm. 70-73. As the latter only confirms the obvious, i.e.,
that the fugue is about to end, only the earlier four-part entry assumes a
crucial position. Moreover, as there are no episodes ending with earcatching cadential closes, structural understanding must rely on the three
cadential closes occurring within episodes and on the usage, so striking in
this piece, of incomplete subject statements.
322
WTC I/24
As has been shown, the first of the mid-episode perfect cadences

occurs in E3, in m. 172. It concludes the first section. Section I thus comprises four statements, A T B S, as well as the episodes E1, E2, and E3a.3
The second section begins with the first episode segment characterized by
episode-material: the interplay of M2 with the M3 imitations in E3b. The
corresponding episode segment with the same material in a similar setting
(E4a) precedes the next subject statement. The section ends with the fugues
only episode concluding in a fully resolved perfect cadence (see E5, mm.
33-34). The third section embarks on incomplete subject statements.
Including these abridged versions, it contains altogether eight entries. Its
confines are determined by the entering order of the complete statements:
when the bass that had provided the first unabridged entry (in mm. 38-41)
sets in again (in mm. 47-50), this signals that a round has been completed.
The actual closure occurs here on the final note of the bass entry, which
falls onto a fully resolved D-major chord (see m. 503).
Looking back on these three sections of the B-minor fugue one could
go so far as to suggest that they represent three expositionsin the sense
that each section exposes new material that will later be taken up: The
first section exposes the primary material (subject and counter-subject), the
episode-use of CSc (in E1, E2, and E3a), the first episode motif and its
ascending sequences, as well as the intended texture by presenting the
fourth entry in four-part setting. The second section exposes the episodetype that relies on the extended sequence pattern of M2 and M3 and is
conceived to precede, rather than follow, a subject statement. The third
section then introduces the pattern of several incomplete subject entries
preceding a complete one.
The fourth section is comparably easy to determine. Its third subject
statement appears, as did that in the third section, as a redundant entry
(here: T B T). Moreover, this third entry is followed by the episode that
was recognized as analogous to E3 in the first section: E10 also consists of
two segments the first of which concludes in a perfect cadence (m. 652).
The fact that the remainder of this episode then opens the ensuing final
section is consistent as it displays the M2 / M3 combination that in section
II also preceded the entries.
3
This in itself would not be remarkable at all. What does appear both unusual and ingenious
is the fact that E2 displays a clearly relaxing tendencyattentive listeners will get the
impression that the exposition of the ensemble is completed after the third entry!
Correlating this observation to the fact that four-part entries are in fact an exception in this
four-part fugue, it seems as if Bach consciously created the impression of a three-part
composition with only occasionally enhanced density in chosen moments.
B minor
323
Looked at from the same angle that earlier revealed three expositions,
one now finds that sections IV and V develop what has been presented
earlier. Section IV returns to the uncluttered structure with no incomplete
entries (thus leaning on sections I and II), but imitates the entry pattern of
section III (B T B becomes T B T). By contrast, section V takes up the
abridged entries from section III as well as the beginning with M2 / M3
from section II and the four-part statement from section I.
The fugues harmonic development confirms the layout described

above: Section I comprises only entries that modulate from the tonic to the
minor dominant and back. The two entries of section II are both conceived
in the original tonal setting (i.e., neither of them exhibits the features of the
324
WTC I/24
tonal answer). They are harmonically arranged in such a way that a return
to the tonic is nevertheless granted: the first modulates from i to v, the
second reciprocates with iv to i. The third section begins once again in the
i-to-v environment but then modulates for good, so that the remaining two
entries move from the tonic relative to the dominant relative and back
again, shading and concluding this section in the major mode. The fourth
section begins and ends on v, leaving the crucial return to the home key in
a iv-to-i modulation to the fifth section.
The dynamic outline is quite different in each of the five sections.
Section I is characterized by a build-up of tension that is gradual though
interrupted. The superimposed increase of tension occurs from the singlevoiced entry to the four-part statement. The first episode, after an initial
drop in intensity, contributes to the impression of mounting tension in its
ascending sequences. Only the second episode decreases and thus suggests
that we are dealing with a three-part fugue and a further entry would be
redundant. Yet the full four-part texture of the fourth subject statement
defies this. The section ends on a high level of intensity that, owing to the
fake entry and the fully maintained ensemble of four voices in the concluding episode segment, hardly abates before the cadence. In section II,
the protracted episodes with their descending motion precede two relatively soft statements, so that the entire section remains somewhat subdued. In
section III the apparent density of materialwith eight occurrences of the
subject headraises the overall tension level. At the same time, the
incompleteness of five of the entries and the fact that they appear accompanied by the relaxing CSd restrict the dynamic development, so that only
the final consecutive entries in T and B gain momentum.
Sections IV and V both begin with longer spans of diminishing tension.
The final entry of section IV is the only one to do without any segment of
the counter-subject. Instead it is accompanied by sequentially ascending
patterns in two voices that create intensity. Section V, having recalled
section II with its opening episode, then leaps directly to the final statement of section I. Prepared only by the incomplete tenor entry, the bass
entry in mm. 70-73 soon gains full four-part texture and thus provides the
fugue with a glorious ending. The full texture is maintained throughout the
ensuing E11 (which builds up tension in ascending sequences) as well as
the incomplete entry, and surpassed in the five-part setting of mm. 75-76.
When relating the climaxes to one another, it seems irrefutable that the
endings of sections I and V represent the overall highlights. In their midst,
section II is probably the softest, while the climax of section III is surpassed
by that of section IV.
WTC II/1 in C major Prelude

This C-major prelude presents itself as a tightly woven texture of four
complementary voices. There are no distinct motifs; recurring melodic
features appear more like formulas marking the beginning or end of a
structural unit. The preludes design is determined by its harmonic progressions, together with secondary features like peak-note lines.
The first harmonic progression concludes at m. 31. This is the only
phrase in the prelude in which all motion is provided by the soprano alone:
the alto has not set in, and tenor and bass are locked in a tonic pedal in
double octave. As the three other voices have not yet entered the musical
development, this cadential close cannot be regarded as structurally relevant.
The conclusion of a section occurs only after the reiterationextended and
with four active voicesof the C-major cadential progression. This second
cadence reaches the tonic in m. 5. When reading through the piece for the
first time, we might assume that this is a strong-beat ending in which the
close falls on the middle beat. As we compare this ending with similar
ones in mm. 8, 20, and 22, however, it becomes evident that all these
cadential closes end in weak-beat extensions of four 16th-notes. The first
section of the prelude thus concludes on the fourth beat of m. 5.
The pattern established in this first progression, i.e., a not-quite
conclusive cadence followed by a stronger confirmation, is repeated twice
in the course of the composition. In mm. 14 and 28, a modulation to a new
key area is harmonically complete but somewhat unconvincing owing to a
rest in the bass. The target key is subsequently corrected and the section
firmly closed a few measures later. There are altogether eight sections:
Section mm.
step
tonality
I
C major
Ia, b
1-31-54
I - ii
C major/D minor
II
54-82
ii confirmed
D minor
III
82-111
D minor/A minor/F major
IVa, b, c 11-142-163-202 ii - vi - IV
IV - V
F major/G major
V
202-224
V
G major
VI
224-253
V - ii - I
G major/D minor/C major
VIIa, b 253-284-302
I
C major
VIIIa, b 302-322-34
327
328
WTC II/1
These eight sections thus represent, on a large scale, the steps of the
simple progression: tonic (mm. 1-5), subdominant (mm. 6-20), dominant
(mm. 20-25) and return to the tonic.
The prelude comprises two extended passages that are built in strict
correspondence to one another: mm. 54-142 . 202-284. Another analogy
established through the use of pedal notes rather than melodic resemblance
exists between the opening measures and the closing measures: compare
mm. 1-31 with mm. 32-34. We can thus condense the structural overview
into the following schema:
opening section
mm. 1-3/3-5
main section
mm. 5-14
middle section
mm. 14-20
main-section recapitulation mm. 20-28
closing section
mm. 28-32/32-34
In this texture of voices complementing one another in a seemingly
endless stream of sound, no articulation or obvious phrasing is desirable.
Instead, the sound flow should meander from one voice to the other in an
unbroken legato. An appropriate tempo has to take into consideration both
the details of the surface pattern and the underlying metric pulse. It must
be calm enough to allow for clarity in the 16th-notes and occasional 32ndnotes in the weaving lines, but not so slow as to render the quarter-note
beats imperceptible.
When describing details of tension growth and decrease, we must keep
in mind that these appear only as shadings before the backdrop of the
predominant harmonic developments. The first phrase (mm. 1-31) displays
an ornamental line falling over an entire octave. The dynamic development
follows the shape, creating a single diminuendo.
The line reads like a hidden two-part structure in

double sixths. It anticipates what later frequently

materializes as a complementary pattern of two
adjacent voices.
The second phrase (mm. 3-53) appears as a more complex reiteration of
the same process. The octave descent occurs within the first half measure
in the form of a scale that falls through the alto and bass registers. Thereafter it is extended, first imitating the more elaborate ornamental pattern of
the first phrase (see mm. 33-4) and then continuing in a cadential bass
pattern (mm. 4-53). The falling line spans more than two octaves.
Only from m. 6 onward do listeners get the impression of true four-part
texture. Secondary features emerge with timid tendencies to counteract the
overwhelming impression of descent. The tenor emancipates itself in m. 4
C major
329
with a four-note ascent (a little crescendo), the soprano, which had been
motionless for more than a measure, follows with an ascending fourth (m.
4: D-G; a smaller crescendo), and the alto adds an even weaker ascending
step (m. 4: B-C). After this, all voices blend once again into the overall
relaxation. This is completed with the weak-beat ending in the bass. The
four ascending 16th-notes are passive here, and the C at m. 54 is best
played as a pianissimo note.
The first phrase of the main section introduces a figure that recurs
several times as an active gesture: the zigzag broken-chord descent in
16th-notes that, after reaching an artificial leading-note (F), resolves
indirectly onto G (B: mm. 54-62). This short figure will be referred to as
M1. Its little tension-curve with the dynamically active beginning inspires
two similar curves in the soprano (mm. 6-7) and bass (mm. 6-7). A last,
softer curve in the soprano (mm. 7-8) leads to the final relaxation of this
phrase with a weak-beat ending at m. 82. Here, too, the ascending notes in
the tenor are passive and end in pianissimo.
The second phrase of the main section also begins with M1 in the bass,
followed by a one-measure dynamic curve in the soprano (mm. 8-9). The
end of this phrase corresponds with that of the previous phrase (compare
S + A: m. 7 with B: m. 10, and the passive ascent in T: m. 8 with the passive
descent in S + T: m. 11). The performers shaping should be geared toward
underscoring such resemblances.
The next phrase consists of two consecutive larger curves. One begins
in m. 11 with three active gestures (S, B, A) followed by a protracted
relaxation through m. 132. The other is launched primarily by the prominent
bass line that builds up some tension in mm. 13-14. The phrase ends once
more with passive gestures (here in A + T).
In the middle section, overriding large-scale pitch motions determine
the tension design. The first peak line occurs in the bass. After two active
gestures describing ascending fourths (see mm. 14-15: C to F and F to B),
it descends gradually in an ornamented line that extends through the
remainder of this middle section, i.e., through six full measures. This bass
line is joined, in the second phrase within this section, by a similarly
embellished descent in the soprano. The most appropriate interpretation,
which admittedly asks for performers with a long breath, is one that
renders the entire descent as a single gradual diminuendo.
From m. 20 onward, the recapitulation of the main section follows,
developing along the same lines as the motions described earlier. It is
complemented in mm. 28-30 and 30-32 by two short phrases featuring
fragments of M1 in the bass (see mm. 28-29 and m. 30). These are
330
WTC II/1
integrated into another descending peak line, which ends in a dosido

formula. Again, rendering these measures in a continuous and almost
unbroken diminuendo would seem the best choice.
The final phrase, which begins after the passive gesture, i.e., at m. 322 ,
can be said to bring forth the real surprise of this prelude. It is launched by
yet another fragment of M1, this time in the alto, followed by an active
gesture in the soprano (mm. 32-33). The preludes penultimate measure
which, due to voice splitting, regains four-part texture despite the extended
bass pedalexpresses such obvious reluctance to calm down and resolve
that it is probably most appropriately interpreted as a final build-up to the
seven-part climactic chord that concludes the piece.
To sum up, the preludes main section contains fairly intricate patterns
of small-scale tension curves, as does its recapitulation. The middle section
and the two initial phrases of the closing section are characterized by largescale descents. The opening section also represents descending motions,
while the final phrase, structurally conceived as a coda, is the only one in
the entire prelude to be dominated by clearly increasing tension, which
ends this meditative prelude on a surprisingly assertive note.
WTC II/1 in C major Fugue

With its extension of four measures, divided by a rest exactly in its
middle, the subject of the C-major fugue reveals a very regular phrase
structure. The beginning on the second eighth-note converts the first
measure into an upbeat to m. 2. Similarly the third measure, with its
downbeat rest, also serves metrically as an upbeat to m. 4, after which the
phrase concludes on the first 16th-note of m. 5. The question whether this
phrase consists of two subphrases or one indivisible unit with a tensionsustaining rest in its middle allows for two answers. These have to be in
keeping with the interpretation of several other subject features. Let us
consider these first.
The pitch outline develops in a fairly restricted range, spanning a major
sixth. The beginning on the fifth gives the subject a hint of being already
in the middle of things, while the conclusion on the third has a gently
releasing quality. The pitch pattern features two accented inverted-mordent
figures (in mm. 1 and 3) along with an unaccented one (in m. 4). These
written-out ornaments suggest that the remaining 16th-notes, both within
the subject and in the fugue as a whole, also invite to be interpreted as
C major
331
ornamental rather than melodious. The two large intervals in mm. 1-2 form
consecutive leaps, thus corroborating the assumption of a lively character.
The target of these consecutive leaps, the A at m. 21, is supported in many
ways: melodically the highest pitch, rhythmically the first of two longer
notes, and harmonically the representative of the subdominant, it is further
emphasized by an ornament proper (see the inverted-mordent symbol)
before it relaxes slightly with the ensuing step downward.
If one regards mm. 3-5 as a varied sequence of mm. 1-2, seeing that
(GF)G-C-AG could become an ornamented (FE)F-(E)-DE, the subject
appears as consisting of two fairly balanced subphrases. The rest in m. 3
must then be played as an interruption before a new beginning, the
written-out inverted mordent following it (m. 3: F-E-F) being more active
than the fairly relaxed G in m. 2, and the D at m. 41 a second, softer
climax. Another view is equally possible and perhaps more conducive to
the transmission of overall unity. The entire string of 16th-notes in mm.
3-5 can be regarded as ornamental. One would then define the main
melodic steps in the subject as (GF)G-(C)-AGFE. In this case,
the gradual descent from A to E requires an uninterrupted line. The rest in
m. 3 is now perceived as tension-sustaining, after a G that sounds only
minimally softer than the preceding climax. The step A-G and the tensionsustaining rest will thus allow for a further release through the following F
and all its gradually retreating ornamental surroundings to the final E. This
interpretation thus renders the subject as an uninterrupted unit.
The subjects harmonic background does not reveal anything that might
decide the matter. The active step to the subdominant appears at m. 21
but then there never were any doubts about the overall climax anyway. The
subdominant relative, another harmonically active chord, is represented by
m. 4, followed by the dominant and the final tonic.
IV I

ii7
V7
The C-major fugue comprises eight subject entries:

1. mm. 1-5 M
4. mm. 21-25 M
6. mm. 39-43 L
2. mm. 5-9 U
5. mm. 25-29 U
7. mm. 47-51 M
3. mm. 9-13 L
8. mm. 51-55 U
332
WTC II/1
Apart from the interval modification in the answer (where the falling
fifth in the first measure becomes a fourth), the subject does not undergo
any changes. Neither does it appear in any stretto or parallel setting.
Against the answer, i.e., at a point where one expects a counter-subject,
the middle voice continues with a melodic line. Later in the piece, this line
recurs twice in its unabridged version (M: mm. 25-29 and 51-55); in two
other cases, only the initial measure reappears (U: mm. 9 and 39). While
independent in structure, this counter-subject is not entirely independent in
material. Its first two measures (see from D, the second 16th-note in m. 5)
are closely related to the subjects final two measures. Only its second half
contributes new components with an ascending scale, syncopation, and
closing formula. Another shortcoming of this counter-subject is the fact
that its unabridged version remains restricted to the middle voice, appearing exclusively in accompaniment to an upper-voice subject statement, and
thus lacks true polyphonic versatility.
The internal structure of this phrase poses a question that is crucial for
the fugues contrapuntal setting. Following the pitch pattern of descending
sequences in the first two measures, most performers will choose to
interpret these as expressing gradually lessening tension. After the lowest
note A, the ascending scale would then support an increase toward the
syncopation, after which the closing formula provides a relaxation.
With this dynamic outline in mind, let us return to the two options for
the interpretation of the subjects phrase structure. We shall find that the
concept described above for the counter-subject is ideally suited to balance
the concept of the subject as an indivisible phrase. In a reading of the subject assuming two dynamic curves, however,
this interpretation of the
counter-subject is unlikely as it would create
simultaneous phrasing
in both voices and thus
undercut the polyphonic
texture. Performers who
prefer a divided rendering of the subject should
thus play the countersubject as an unbroken
phrase, beginning with
an extended crescendo.
C major
333
The C-major fugue comprises only four subject-free passages.

E1
mm. 13-21
E3
mm. 43-47
E2
mm. 29-39
E4
mm. 55-83
The episodes make ample use of material from the subject. E1 features
an imitative pattern based on the subjects first half, complemented by an
eighth-note instead of the rest (see mm. 13-19, U and M; the final imitation
in M is inverted and varied). This pattern recurs faithfully in mm. 55-61. In
both instances, the lower voice adds a figure that is derived from the last
measure of the subject and/or from the first measure of the counter-subject.
The lower-voice 16th-notes then continue, with a little more liberty, up to
the end of the episode (m. 211) and up to the final cadence of the earlier
version (m. 671) respectively. The head of the subject further recurs in the
coda, where its imitations move through all three voices (L: mm. 68-72,
M: mm. 72-76, U: mm. 76-80). The first eight measures are once again
accompanied by the 16th-note figures from the subject ending and/or the
counter-subject beginning.
In E2, the first segment features a two-measure motif whose beginning
is rhythmically related to the subject head (U: mm. 29-31, sequenced in
mm. 31-33). The accompaniment, once again, is the string of thematic
16th-notes. The second segment of this episode brings an imitation built
exclusively on the 16th-note pattern (see U: mm. 33-371, M: mm. 34-391).
E3 displays a lower voice that is even more closely related to the end of
the subject as it sequences the final measure (see L: mm. 43-46).
Apart from the final measures of the coda, the only components of the
episodes that are not related to the primary material appear in the upper
and middle voices of mm. 43-47 and 61-67. They do, however, recall the
other episodes insofar as they also feature an imitative pattern in sequences.
The role each episode plays in the dynamic design of the fugue is easily
determined in accordance with the sequential patterns. E1 builds up
tension through mm. 13-19 but then brings a slight release in mm. 19-21.
E2 keeps a very low profile in its constantly falling lines. E3, on the
contrary, ascends continuously in all three voices and thus produces a
considerable increase. E4, like E1, begins with rising lines and a crescendo
(mm. 55-61) followed by a gradual descent (mm. 61-65) and a cadential
releasewhich now, in its revised version as an interrupted cadence,
expresses much more intensity than did the earlier perfect cadence. The
coda begins in a fairly leveled softer shade, followed only in mm. 80-83 by
a final strengthening (see the split voices).
The uniformity and simple relationship throughout both the primary
and secondary material adds to the fugues playful character. The impres-
334
WTC II/1
sion is further enhanced by rhythmic continuity: prior to the final cadence,

the 16th-note pulsation is almost constant, with only minimal interruptions
on four occasions (after the downbeats in mm. 11, 21, 22, and 41). The
simplicity of the rhythmic pattern, the ornamental structure of the pitch
line, and the leaps that occur both in the subject and in the episode material
join to express a rather lively character. The tempo, too, should be fairly
swift. The articulation requires non legato for the eighth-notes and quarternotes and legato for the 16th-notes.
Owing to the prevailing patterns of almost continuous 16th-notes in
both pieces, which would sound dull if the respective pulses were directly
related, the relative tempo had best be chosen in complex proportion. A
good and feasible solution is to translate each of the four quarter-notes in
the preludes final measure into triplet eighth-notes (instead of the previously felt four 16th-notes), and then turn these imagined triplet eighth-notes
into the eighth-notes of the fugue. In other words: an (assumed) triplet
eighth-note in the prelude corresponds with an eighth-note in the fugue
(Approximate metronome settings: prelude beats = 72, fugue beats = 108).
The prominent ornament notated with a symbol is the inverted mordent
in the subject. Its pitch does not pose a problem as it always uses the
natural (white key) for its lower neighbor note. As is the case with all
ornaments that form a characteristic feature of a subject, this inverted mordent must be played even where it is not marked (i.e., in mm. 22 and 26).
Conversely, the episode motifs deriving from the subject head need not be
ornamented as Bach does not indicate this in a single case. (The presence
or absence of the ornament has the further advantage to tell listeners
already after one measure whether what they are hearing is a subject entry
or a subject-related motifa distinction that, due to the uniformity of
material in this fugue, would otherwise not be so easy to make.)
Another ornament appears in two of the three counter-subject entries.
In m. 8 it is printed as an inverted mordent, while m. 28 features a mordent
symbol without the slash indicating inversion. The first reading is unusual
for a typical closing formula in which experienced performers might have
added a full trill even without any invitation from the composers hand.
The second reading appears more convincing and can safely be chosen in
both cases. As a leading-note trill approached stepwise it begins with a
16th-note on the main note, shakes in six 32nd-notes including the suffix,
and resolves smoothly into the subsequent downbeat. In m. 54 the closing
formula is varied and needs no ornament. Finally, the long trill with a
tie-suspension in mm. 37-38 begins regularly, i.e., from the upper note; it
then shakes in 32nd-notes until the very end of the measure where it ties
C major
335
the last F over to the next downbeat. No suffix is possible in this case since
the ornamented note lacks a resolution.
At this point it may be interesting to learn that Bachs first version of
this fugue (according to the Kellner manuscript) ended on m. 681. The
composer later rewrote the perfect cadence in mm. 67-68 as an interrupted
cadence and added 16 measures on a pedal note C. This extended coda
concludes the piece much more convincingly. At the same time, the fact
that it was not intended at the time of the earlier version is of great help for
a true understanding of the fugues architectonic design. This is determined by cadential formulas together with the key sequence of the subject
entries and the dynamic buildup in some of the episodes. Only the end of
the initial section may cause some doubt. In mm. 21-22, Bach presents the
first closing formula with typical features in both the upper and the lower
voices, thus creating a strong feeling of closure. The beginning of a subject
statement in the middle voice overlaps with this closing formula for an
entire measure. In this manner Bach strings the first and second sections
closely together.
Given that the ensemble is reduced to two voices after the cadence in
mm. 24-25, which would normally advocate the beginning of the second
section, one needs good grounds for a differing view. Four reasons support
the assumption that section I ends already in m. 22: As all three voices have
already presented the subject, the first section can only close here or after
an additional (redundant) entry. Yet the overlapping subject statement in
the middle voice ends with a cadential close that is not quite satisfactory
owing to a sudden break-off in the upper voice (m. 251), which creates a
strong link between the upper-voice line in m. 24 and the new beginning in
m. 25. The overlapping subject statement in mm. 21-25, while beginning
in the harmonic surroundings of G major, soon reveals its loyalty to the
key of D minor (see the Cs and Bs from m. 22 onward), thus forming a
pair with the following upper-voice in A minorthe minor dominant of
the preceding statement. A closer look at the beginning of both statements
further strengthens this reading as they display the interval structure of
subject + answer. (Compare mm. 21/25 with mm. 1/5.) If the subject entry
in mm. 25-29 were the first of a new section, then the episode that follows
would have to be regarded as linking two consecutive statements. The long
and definite tension decrease in E2, however, makes such an interpretation
very unlikely. Lastly, the dynamic design of E1 is distinctly related to that
of the original E4 (until m. 68): both describe a full curve, using similar
material.
336
WTC II/1
Having said all this, a conclusion for the remaining sections follows
without problems. The harmonically related entries in mm. 21-25 and
25-29 constitute, together with the decreasing E2, the second section. The
return to C major and to three-part texture in m. 39 marks the beginning of
the third section, which comprises three statements and a closing episode
(like the first section, the only difference being the inserted E3). The
exceptionally long coda must be regarded in the light of its relative in the
Classical period: it rounds off the entire piece, not just the final section.
In the first and the third sections, the tension rises from one entry to the
next. In the first section, this is bolstered by the gradual increase in texture;
in the third section, Bach uses for the same effect an ascent from the very
low keyboard register (mm. 39-43) and a mid-section episode with a
strongly increasing tendency. The tension diminishes in the second section
owing to the reduced number of voices in all but the first four measures,
the minor mode, and the strongly decreasing direction of the episode.
The coda begins softly but then builds up strength as the pedal moves
to the lower C and the lower-voice figures suggest hidden two-part structure.
When the other voices also split, the fugue ends in full resonance.
WTC II/2 in C minor Prelude

This prelude is composed in almost consistent two-part texture. There
are only three instancesmm. 12, 26, and 28where the simple pattern is
suspended in favor of an apparent third voice or voice-splitting into fuller
chords. Polyphonic independence of the two voices, however, is very
limited. Most measures feature open or slightly concealed parallel motion.
At the same time, each voice consists of various levels, often allowing a
distinction between melody and background. The background may be
a pedal note or an ornamental filler. On the melodic plane, several gestures
recur, both in imitation or sequence and in various modifications and
developments. Yet owing to the very uniform rhythmic structure consisting
exclusively of 16th-notes and eighth-notes and interrupted only in the final
measures of each section, these melodic gestures seem to bear less weight
than they might in different surroundings. The result is a light piece shaped
primarily by large motions of ascending and descending lines, reminiscent
of the B-major prelude from book I.
Harmonic data provide insight into the structural layout. Within the
first section, mm. 1-51 confirm the home key C minor, mm. 5-93 modulate
to the relative key E major, and mm. 93-12 confirm this secondary key. In
the second section, mm. 13-171 modulate to the subdominant F minor, mm.
17-223 confirm the subdominant, after which mm. 223-261 return to and
mm. 26-28 confirm the home key.
A simple rhythmic pattern and the presence of leaps and ornamental
figures in the pitch outline suggest a lively basic character. The articulation
requires non legato for the eighth-notes and a very crisp quasi legato for
the 16th-notes. The tempo is gently flowing. Performers will want to make
sure that listeners perceive a quarter-note pulse with ornamentation rather
than melodious 16th-notes while avoiding the impression of sheer finger
virtuosity. Not only did Bach choose 4/4 and not 2/2 for the meter, but he
also expects additional ornaments on some of the 16th-notes. The inverted
mordents in mm. 7 and 8 pose no problem: there is ample time for the three
notes and the pitch of the lower neighboring note is unquestionably G. The
trills in mm. 14 and 16 are more demanding. Even though they do not
encompass suffixes since the following note qualifies neither melodically
nor metrically as a resolution, four notes must be fit into a 16th-note, due
to the regular beginning on the upper neighbor note.
337
338
WTC II/2
With regard to color shading, careful gradation of intensity in the notes

belonging to the different melodic layers accounts for the beauty in this
prelude. To give an idea of how dynamic shading can convey this particular
textural design, the following example depicts these layers in mm. 1-2:
The initial motif (M1) begins in a metrically weak position. The bass
note C on the downbeat serves as an anchor: it takes part in neither the
melodic lines nor the regular background features. Its tone quality is
therefore sonorous but neutral. M1, as shown in the example above, consists of a melodic descent in quarter-notes doubled in compound thirds. On
the first level of background, a 16th-note-figure in the shape of an inverted
mordent serves both as a melodic link between the notes of M1 and as an
octave ornament of its parallel. Finally, the off-beat eighth-notes in the
accompanying voice represent a repeated C that, although it gives way to
the harmonically required B at the end of the measure, conveys a strong
sense of an indirect pedal note. All features recur in inverted voices in m. 2.
Moreover, they turn out to be the constituents of the entire piece.
M2, introduced in mm. 3-4, appears as if enlarged on various levels.
The steps of the melodic line, now ascending, move in strong-beat halfnotes so that we hear pairs of repeated notes in active/passive grouping:
D-D, E-E, F-F, G-(E). The parallel in the lower voice also begins, in m. 3,
with an ornamented half-note G. Only in m. 4 do we hear chromatic steps
in quarter-note pulse (A-A-B-B-C). The indirect pedal is abandoned in
this motif. M3 in mm. 5-73 drops an additional feature: the accompanying
eighth-notes. Instead, the parallel melodic steps, falling on beats 1, 2, and
3 of each of these measures, are surrounded by the ornamental figure that
is now intensified in parallels, and complemented by a scale that overruns
the obvious two-part texture by creating the illusion of a link between the
C minor
339
upper and the lower voices. In mm. 73-83, the lower part continues the
previous pattern while the upper part emancipates itself briefly with a
repeated ornamental figure. The harmonic progression is concluded in
mm. 83-93 with a closing formula in the bass and in the melodic layer of
the upper part (E-D-E). The first section ends with a transposition of M1
(compare mm. 93-113 with mm. 1-2) followed by another cadential close in
mm. 113-12.
The second section begins with a slightly more complex pattern. In
mm. 13-14, we hear a hidden-three part texture consisting of an upper
layer with melodic quarter-notes followed by a straightforward melodic
figure and its descending sequence, a lower layer with neutrally-colored
accompanying eighth-notes followed by quarter-notes, and an ornamental
background that, here, does not move together with the melodic lines but
presents an indirect pedal on B (m. 13, right hand), followed by an ornamented G and F (m. 14, left hand). The entire pattern is then repeated in
transposition (upper voice a fifth down, lower voice a fourth up), with only
minimal adjustments at the beginning of the pattern.
In the substantial center of this second section, Bach establishes no
motifs at all. Variations of earlier features can be made out, e.g., a modified version of M3 in mm. 19-20, a remote resemblance of M2 in m. 22,
and ornamented indirect pedal-notes in mm. 17 (r.h.: C) and 21 (l.h.: F).
Yet although a minimal sense of melodic recognition is encouraged by
sequences, the predominant features are not found in the details but rather
in the large-scale lines.
The following sketch, showing an excerpt of this prelude in its skeletal
version, concentrates on these features and thus gives guidelines for
intensity shading and dynamic shaping:
After the harmonic return to C minor, the preludes last three measures
recapitulate fragments of the various features. There is a variation of M2
beginning in the left hand with an upbeat of three 16th-notes to m. 26.
They are paralleled in the middle voice in m. 26 and picked up by the
340
WTC II/2
upper voice in m. 27 (here accompanied by the chromatic steps heard

several times before in the bass). The piece ends with a measure that
resembles the final measure in the first section, thus establishing a certain
sense of binary form that is otherwise easily lost in this piece.
Regarding the large-scale development of tension, the following
dynamic pattern emerges from the play of motifs and peak-note lines:
mm. 1 and 2 + 93-103-113 two self-contained curves
mm. 3-51
+ 113-121
crescendo
mm. 5-93
+ 12
diminuendo
mm. 13-14, 15-16
curves with climaxes at m. 141 and 161
mm. 17-20
diminuendo
mm. 21
l.h. diminuendo, r.h. crescendo
mm. 22-25
diminuendo
mm. 26-281
crescendo
m. 28
diminuendo
WTC II/2 in C minor Fugue

With only 28 measures, the C minor fugue seems relatively short. This
is also reflected in its subject, which spans only one measure, beginning
after an eighth-note rest and concluding on the next downbeat. As in the
preceding fugue, the subject sets out as if suspended on the fifth and closes
on the third. Two further features are also reminiscent of the C-major in
WTC II fugue: Regarding rhythm, the regular eighth-notes are complemented at the end of the phrase by two 16th-notes that, together with the
final note, form an inverted-mordent figure. Regarding pitch, the subject
features two consecutive leaps with large intervals (m. 1: G-C-F).
A considerable difference between the two thematic lines is that the
interval leaps in the C-major subject occur in a metrically strong position
(leading toward a downbeat) while the C-minor subject features the leaps
in a metrically weak position (relaxing after a strong beat). And while the
metric organization in the C-major subject leaves no room for doubt, the
three initial eighth-notes here give the impressioncertainly to all listeners
who are not following a scoreof an upbeat in a 2/4 measure, especially
since they describe a melodic curve that returns to its beginning: G-E-F-G.
(It is, in fact, almost impossibleand perhaps not even desirableto
attempt a rendition for the beginning of this fugue that does not leave the
metric organization unclear to the audience.)
C minor
341
The subjects simple rhythmic design is deceptive. From m. 5 onward,

16th-notes become a regular feature, occurring either in linear runs and
figures or in complementary patterns between two voices involving dotted
and tied notes. Finally, there are even instances where other note values
quarter-notes and 32nd-notesplay a role in the melodic material.
The harmonic background can be called rudimentary. It consists of a
minimal I-V-I, making do without even a proper representation of the
subdominant. Whatever the harmonization of the initial note, which Bach
keeps changing throughout the piece, the tonic is established weakly on
beat 2 and confirmed on the stronger beat 3. Where the connecting eighthnote is harmonized as a subdominant, this
chord is not only placed in a metrically
subordinate position but, more importantly,
heard as ornamentation between the two tonic
chords. The opposite is true for the fourth
beat of this measure.
Considering this unusually plain harmonic design in conjunction with
the metric particularities of the subject, there are two notes that might
contend for the privilege of serving as climax. The F on the fourth beat is
harmonically slightly enhanced and appears melodically prepared by the
two consecutive leaps. The G on the middle beat holds the metric focus
and could be regarded as the point of departure for an ornamented descent
in quarter-notes GFE D E. (In this interpretation, the C in the center
of the two consecutive leaps would be read as a kind of background-level
note not partaking in the main melodic line.)
This is a fugue with many subject statementstwenty-four altogether.
While identifying the entries is not difficult, the voice allocation is not as
obvious as it might seem at first glance. A straightforward reading would
distribute the statements among the four voices that Bachs title mentions.
Yet this obvious and apparently simple solution has several shortcomings:
The fugue is consistently in three-part texture until m. 19, i.e., for far more
than half of the piece. The exposition of a four-part fugue would have to
include the fourth statement. The fourth subject entry in mm. 7-8, however,
appears in a texture with only one other voice, and thereby gives the
impression of opening a new section. As if to support this reading, the
downbeat of m. 7 presents a cadential close that is melodically satisfactory,
while that at m. 81, after the fourth entry, is much less convincing.
L. Czaczkes in his analytical work on Bachs fugues makes a suggestion
for the structure of this fugue that may sound daring but has the advantage
that it works, i.e., that it makes sense without asking for compromises. He
342
WTC II/2
contends that what listeners perceiveand what Bach conceivedis a

three-part composition, with a fourth pedal voice entering only toward
the end and only for four repeated subject statements linked with cadentialbass notes. The table below lists the subject statement in a voice allocation
according to the concept described above and, in brackets, the designation
according to an open score (Thompson, Toronto) that represents the first
glance view. The descriptions in this chapter will refer to the former.
1 mm. 1-2 A
9 mm. 14-16 A
17 mm. 21-22 B
2 mm. 2-3 S
10 mm. 15-16 T
18 mm. 22-23 B
3 mm. 4-5 T
11 mm. 16-17 A
19 mm. 23-24 A
4 mm. 7-8 T (B) 12 mm. 16-17 S
20 mm. 23-24 S
5 mm. 8-9 S
13 mm. 17-18 T
21 mm. 24-25 S
6 mm. 10-11 A
14 mm. 17-18 S
22 mm. 25-26 A
7 mm. 11-12 T (B) 15 mm. 18-19 A
23 mm. 25-26 T
8 mm. 14-15 S
16 mm. 19-21 B
24 mm. 26-27 B
The subject undergoes a number of modifications as well as various

groupings. The answer differs from the original by a smaller initial interval: the major third is replaced by a major second. This allows the answer
to begin on the tonic but continue in the dominant. Throughout the fugue,
seven subject entries can be identified as answers. One of them adopts the
interval structure of S: mm. 2-31 (A: mm. 16-171) while the remaining five
reduce the initial interval to a minor second (B: mm. 7-81, A: mm. 10-111,
S: mm. 17-181, B: mm. 22-231, and S: mm. 23-241). Subject inversions
occur three times, all of them in the harmonic setting of an answer. Augmentation is used twice. Both enlarged entries quote the subjects original
shape without any modification.
Strettos occur frequently. They include combinations of the original,
the inversion and augmentation, combining two, three, or four statements.
In one case, the overlap is so dense that a repeated entry in one voice
unites altogether five statements into a group.
Modifications may appear at almost any place in the phrase; they can
serve as variations, harmonic changes, or the one resulting from the other.
Variations of the subjects head include an abbreviation of the first note
(mm. 8 and 17). Simple harmonic changes occur in the form of a change of
mode (mm. 7-81, 10-111, 22-231, and 25-261). The final note may be raised
in the manner of a Picardy third (m. 241) or diverted so as to transform the
inverted-mordent figure into a scale segment (m. 111). Where this
C minor
343
deviation of the final note occurs in the inversion of a minor-mode entry,

the subject ends in the melodic minor scale and thus contains shifting
accidentals (mm. 161 and 221). Similarly, a new harmonization of the
beginning converts the second and third subject notes into the sixth and
seventh degree of a melodic minor scalefollowed in the descent by notes
from the natural scale (mm. 16-19). Yet another change from major to
minor owed to a stretto position occurs in mm. 24-25. Finally, the last
statement surrounds its initial note with artificial leading-notes on both
sides and then breaks into a virtuoso flourish (see B: mm. 26-271).
There is no true counter-subject. The simple descent accompanying the
second and third subject entries (A: mm. 2-3, S: mm. 4-5) is an extension
and further relaxation of the subject and fails to recur later.
The fugue comprises seven subject-free passages, most of them short:
E1 mm. 3-41
E4
m. 11 (eighth-notes 2-5)
E2 mm. 5-71
E5
mm. 123-141
E3 mm. 9-101
E6
m. 23 (eighth-notes 2-5)
E7
mm. 27-28
The episodes are easily described in both their material and the role
they play in the overall dynamic outline. E1 with its descending lines links
two subject statements. E3 is related to the subject: it evolves from the
preceding entry in two descending sequences of all three voices. Both
episodes function as bridges, with slightly falling tension. E4, although
only half a measure long, has highly important effects: it shifts the subject
metrically (placing the climax on the downbeat) and harmonically (modulating with a chromatically enhanced bass line from C minor to F minor).
The tension rises here. E5 and E6 contain typical cadential formulas, as
does the final E7 that splits into five and (in the final chord) six voices.
Only the one longer episode, E2, introduces an independent motif. First
presented in the tenor, the motifs eight 16th-notes describe a fifth leap
answered in opposite direction by an almost complete scale. The motif is
sequenced once and then imitated in inversion in the upper voice, where it is
again sequenced. A further imitation, now in the middle voice, accompanies
the following subject entry. This episode is dynamically self-contained: the
tension rises through one measure and falls through the other. In mm. 12-13,
the ascending sequences in all three voices take up a mediating position
between primary and episode material. The tenor presents a free variation
of the subject (which includes a completely changed first eighth-note and
two octave displacements). This sequential measure complements E5,
extending it to two-measure length with a curved dynamic shape and thus
establishing a correspondence to E2.
344
WTC II/2
The basic character of this fugue requires careful pondering. Owing to

the subjects simple rhythm and consecutive leaps, the initial two measures
might easily lead one to assume a lively character, with detached eighthand quarter-notes. Yet this articulation fails already in m. 3 where the syncopated rhythm would be lost in detached style. From mm. 8-9 onward we
need not only legato playing but a calm mood to appreciate the texture.
Thus the complexity of the rhythmic pattern in the fugue as a whole
dominates in determining the basic character as rather calm. In terms of
articulation, the decision performers have made regarding the dynamic
shape of the subject influences its performance: If the F on beat 4 of the
subject is considered and played as the climax, then this climax is prepared
in a linear way by the fifth + fourth leaps. In this case, these leaps would
have to be played gently detached as they in fact interrupt the smooth
melodic line. If, by contrast, the middle-beat G epitomizes the peak of
tension, the situation is different. As was already briefly mentioned above,
this interpretation supports the subsurface line GFE D E in the second
half of the subject. In this concept, the C that causes the two leaps does not
belong to the principal melodic contour but sounds on a secondary level of
intensity. It thus acts as a momentary escape from a smooth line in
descending seconds, and not as an equal participant in a pattern of active
leaps. In this interpretation of the subject, the entire phrase would thus be
played legato.
or
In either case there are a few leaps outside the subject that require nonlegato rendering: the octave leap in m. 3 (G), the cadential leaps in the
lower voice of mm. 8-10 and 13-14, the alto on the middle beat of m. 19,
and the bass in m. 23 as well as in the three final notes in mm. 27-28.
The tempo of this fugue is moderately flowing: slow enough to allow
for full appreciation of the texture in all measures with a complementaryrhythm structure, but not so slow as to stretch the augmented statements
beyond recognition. These should still move in such a way that they could
be sung on one breath. The relative tempo of the prelude to the fugue may
be chosen in simple proportion since the two pieces are conceived in so
different a character: a half-note in the prelude corresponds with a quarternote in the fugue. (Approximate metronome settings: prelude beats = 108,
fugue beats = 54)
C minor
345
When describing the fugues layout, we must decide between Czaczkes

reading as a three-part fugue with late-coming pedal voice and the firstsight treatment as a four-part fugue with puzzling irregularities in the first
two-thirds (mm. 1-19). The analysis that follows is based on the former
assumption. The entering order of the voices in the three-part portion of
the piece, the texture in which they appear, the dynamic gestures of the
episodes, and the harmonic progression all support this plan.
The first section comprises three statementsone in each of the three
voicesas well as a linking episode (E1) and a self-contained one (E2)
concluding the section at m. 71. The second section begins with the ensemble reduced to two voices and the subject statements representing the
answer now in the major mode. There are four entries in this section, the
last of which is redundant and appears in F minor. It is followed by a
strongly varied sequence that shifts to G minor, the key in which the
section ends with a cadential formula at m. 141. Both episodes included in
this section are bridges between consecutive entries. E3 corresponds with
E1 in the first section both with regard to its descending sequences and
with regard to its position between the second and the third statements.
The third section begins again in reduced ensemble and is characterized by
the introduction of an augmented subject in stretto with an original entry
and an inverted entry. In this three-part stretto, the augmented alto entry
serves as leader, although it enters second. In the next stretto, the salient
feature is a repeated soprano entry (also entering second). This repeated
entry takes up the same time as the augmented entry before. Secondary
entries in this stretto appear in the alto, the tenor and again in the alto. The
third component of this section is provided by the newly entering bass.
Matching the preceding strettos with three entries, the bass covers consecutively the augmented, the inverted, and the original versions of the subject.
The cadential close of E6 completes this section. The fourth and last section
begins once more with a drastic reduction in the number of voices, but
embarks on two more strettos. The first presents an interplay of alto and
soprano, with sketchy support of the tenor and only an eighth-note C as a
pedal note in the bass. The second stretto begins similarly (entries in
soprano and alto, similar tenor figure, same bass eighth-note) but then
includes a tenor entry and a strongly varied bass entry, together with
fragments of the subject in the upper voices. E7 concludes this section.
The first and second sections are almost equally long and similarly
designed. Moreover, sections I + II (mm. 1-14) and III + IV (mm. 14-28)
are also of corresponding length.
346
WTC II/2
The first two sections remain understated in their emotional content.

The tension rises little in the course of the first section where the increase
in texture is counteracted by descending lines in the accompanying voices
and in the episode. The same holds true for the first three entries in the
second section. Only the redundant entry presents a heightened level of
intensity. The third section, by contrast, conveys impressions of density (in
the five-entry stretto in mm. 16-19) and of grandeur (in the augmented alto
entry, the reiterated soprano entry, and the threefold bass statement). While
in the fourth section the bass pedal evokes a foreboding of the impending
end, the impression of density is repeated in the final four-part stretto and
the voice splitting of the closing episode.
WTC II/3 in C major Prelude

The prelude in C major is the only one in the second book of Bachs
Well-Tempered Clavier to appear as two in one: It comprises two clearly
designated sections contrasting in every respect: in meter, rhythmic
features, melodic relevance, and texture. The tempo indication in m. 25
recalls the C-minor prelude of Book I, while the sequence prelude / fugato
/ fugue is reminiscent of the prelude in E also from the first volumewith
the difference that there, the threefold form took place within the confines
of the prelude itself whereas here, it involves the fugue proper.
The regularity of the rhythmic features within the first twenty-four
measures defines the prelude-portion as harmonically determined. The
apparent four-part texture hardly disguises an underlying chordal structure.
In terms of measure numbers, the change to a faster tempo and triple meter
occurs almost exactly in the middle of the piece. In terms of actual playing
time, however, the proportion between the two structural halves is quite
different, with the Allegro taking up little more than one fourth of the
duration of the prelude within the prelude. Yet the fugato achieves
musical balance through the intensity of its polyphonic design.
The preludes first harmonic progression, extending over a tonic pedal,
closes at the end of m. 3 where the C-major harmony is regained. The following three measures provide a modulation to G major, the dominant,
which is reached at the end of m. 6. The third harmonic progression is
longer and more complex, passing through several secondary keys before
resolving finally into the subdominant F major at the end of m. 13. The
fourth harmonic section is the most vivid and deserves closer inspection
later. Suffice it here to mention that it consists of a series of dominantseventh chords all deceived in their resolution. These four measures lead
into a transposition of the second progression (mm. 18-20 . 4-6), which
brings the return to the tonic. While this may give a sense of closure, there
is a fifth progression that leaves C major once again to turn to the
dominant, from where the fugato is then launched.
The fugato follows with three further harmonic progressions that do
not, however, make any attempt to leave the area of tonic and dominant.
The very pronounced closing formula in mm. 33-34 marks the return to the
tonic. A weaker cadential formula turns once more to the dominant in
347
348
WTC II/3
m. 411, and the final ten measures confirm the tonic with voice-splitting to
a four-part chord in m. 50. Here is an overview of the layout:
Prelude
I
mm. 1-3
I
(C major)
II
mm. 4-6
I-V
(modulation to G major)
III
mm. 7-13
V-ii-vi-I-IV (modulation to F major)
IV
mm. 14-20
IV - I
(return to C major)
V
mm. 21-251 I - V
Fugato
VI
mm. 25-341 V - I
(return to C major)
VII mm. 34-411 I - V
VIII mm. 41-50
V-I
(final return to C major)
The challenge posed by this piece is to create the greatest possible
difference between the two halveswhile at the same time conveying the
message that they belong integrally together.
Difference in character is achieved mainly by variations in touch and
intensity: In the prelude, no note, whether of 16th-, eighth-, or quarternote duration, should stand out from the chordal texture of which it is but
a part. Dynamically this means that any small-scale increase, any accented
or agogically delayed stroke automatically sounds like a pretense of melodic
independence, which is not what is wanted here. (In order to create this
non-melodic effect, performers might even strive to counteract natural
tendencies and play the soprano-ascent at the beginning of each half
measure with an imperceptible decrease.) In addition to such dynamic
treatment of note-groups, the intensity in each single note should be kept
very low. This is obtained by a combination of neutrally colored touch and
very even articulation. While the connection between the 16th-notes is
definitely legato and should not pose a problem, an even articulation in the
eighth-notes is less easy to achieve and needs more attention than pianists
are often willing to pay in the case of such secondary features. The choice
is between, on the one hand, legato (which means finger-legato in note
repetitions) and, on the other hand, a very gentle non legato (in which
skips are no more separated than steps).
In the fugato, the opposite holds true: the texture consists of nothing
but melodic features. Every note is alive in color and touch, and every line
is dynamically shaped. Articulation, too, is different: the legato in the
16th-notes is lighter than that in the prelude; unmarked eighth-notes are
well detached, and the wedge-carrying eighth-notes in the fugato-motif
(as well as in all further entries where this is not specifically indicated)
sound staccato. Unity between the two unequal halves is achieved by
C major
349
means of tempo proportion: one beat (a quarter-note) in the prelude

corresponds with a whole measure (a dotted quarter-note) in the fugato.
Ornaments appear only in the fugato. One is the inverted mordent in
m. 28. It is played with the regular F as the lower neighboring note. In the
imitation, the inverted mordent also uses the whole tone D-C-D. Despite
the rule that thematic ornaments are to be transferred to further entries, and
the fact that the fugato motif recurs several times, such transferals do not
take place in this piece. They are hampered either by the simultaneous use
of the originally ornamented note in a cadential bass pattern (as in mm. 33
and 40) or by a varied ending of the motif (as in mm. 44 and 45). Another
ornament in this fugato is the cadential trill in m. 33. As its resolution,
C, is anticipated on the 16th-note before the strong beat, this ornament is
a point darrt trill: it ends without a suffix, stopping short on (or
preferably immediately before) the beat preceding the anticipation. In this
case a possible realization consists of four 32nd-notes of which the last is
tied over to the third eighth-note beat. This ornament, too, remains
singular. Finally, there is the grace-note preceding the first downbeat in the
fugato motif. As an appoggiatura it enhances the motifs climax. It is
indicated in U: mm. 26 and M: m 27 but can be played equally effectively
in L: m. 31. Its note-value allows for two solutions. According to the rule
of rhythmic splitting in dotted values, an eighth-note appoggiatura
followed by a quarter-note resolution is perfectly acceptable. Thinking,
however, of the usage to let sustained note-values in imitative texture
disappear as early as possible so as to direct the listeners ears toward the
newly entering voice, a realization with an appoggiatura of only 16th-note
duration also finds support. In this second case, transferal of the ornament
beyond the first round of entries is not possible since all further statements
of the motif are accompanied by 16th-note lines. An eighth-note appoggiatura does not cause harmonic clashes in later entries but may sound
somewhat odd; it is definitely excluded in m. 38 where it would lead to
hidden octave parallels.
The first two harmonic progressions within the prelude describe
simple and very smooth dynamic curves, with the climax on the respective
subdominant chords (i.e., the downbeats of mm. 2 and 5 respectively). As
the second progression contains an active modulation, its dynamic increase
may be slighter more pronounced than that in the first progression.
The final measure of the second progression features in mm. 6-7 and 8-9
a tentative melodic emancipation of the tenor. This melodic sequence,
apart from highlighting the relationship between the measures, remains a
transitory secondary feature. Harmonically, the progression moves swiftly
350
WTC II/3
through various modulations, reaching on its way steps ii (D minor in m.

73) and vi (A minor in m. 83). The dynamic equivalent demands an initial
increase that is slightly stronger than any previous one (toward m. 71)
followed by an incomplete release from this V7 toward the inverted D
minor chord (m. 73), and a weaker increase toward the next V7 (m. 81)
followed by a more definite release to the target chord A minor (m. 83).
The melodic sequence in the tenor, beginning on the fourth beat of this
measure, engineers another increase that now launches a very gradual
protracted relaxation, up to the resolution into F major in m. 133.
The fourth progression consists of two curves. The first is launched by
the most powerful dynamic increase in the prelude, triggered by the
unexpected harmonic turn in m. 14. It is complemented by an almost
reluctant and very gradual decrease through a series of wrong resolutions
and ends unresolved (both in terms of harmony and of dynamic tension) in
m. 17.1 The second curve, once again a regular and complete cadential
progression, begins with a slight increase toward the subdominant (here:
its representative ii7, see m. 191) from where the tension falls to a complete
release in the return to the tonic (m. 203).
The final progression within the prelude could be played in various
ways. Most convincing, in view of the unity between the two halves of this
composition, is to render this renewed modulation to the dominant as a
long crescendo, so as to arrive at the fugato in a tone quality distinctly
different from (i.e., more intense than) that in the prelude but not jarring
with the earlier color as would be the case with an abrupt contrast. The
diagram below aims to provide a visual suggestion of the relationship
between the harmonic developments and the dynamic shading in the
prelude within the prelude.
1
Apropos wrong resolutions: The harmonic progression here is most bizarre: m. 14

begins with a G-major seventh chord that creates an expectation for C major, but is
followed instead by an A-major seventh chord! The latter generates a longing for D
which is reached only in the bass, while the other voices prolong the dominant-seventh.
The delayed resolution is once again not satisfactory: in m. 153, D is reached but as before
in the form of a dominant-seventh chord. In the same way, the resolution of this chord
appears as a G-major seventh (m. 161), and its resolution is another dominant-seventh
chord (on C, see m. 163). The consequent F major chord in m. 171 comes with a major
sevenththus not as a dominant but as a tonic with a leading-note that one expects to hear
resolved in the second half of the measure. Yet, instead of granting this melodic release,
which now seems so close at hand, Bach leads the soprano downward, thus creating what
turns out to be the subdominant-six-five of C major. At this point of what could be called
harmonic frustration, the listener thus comes to understand that no repetition of the F
major cadence was intended, but just a delay of the return to the home key.
C major
351
mm. 1
14
18
21
In the fugato, the motivic beginning with a 4/16 upbeat suggests an

increase toward m. 261, a note additionally enhanced by an appoggiatura.
A complementing decrease is thwarted by the absence of the expected
leading-note F (are we in G major or still in C?), the energy of the two
staccato eighth-notes, and the rhythmic accent of the final syncopation,
which is intensified by an ornament. There is thus no dynamic relaxation.
The motifs initial entry is answered in fugal style by imitations on the
fifth (M) and the octave (L), before a half-measure closing formula concludes the section. Section II begins with two inverted and varied entries
(L: mm. 34-371, U: mm. 35-381). A statement in the original shape (L: mm.
37-40), an incomplete inverted entry (U: mm. 38-39), and another brief
cadential formula follow. The final section features two entries of the motif
(M: mm. 41- 451, U: mm. 42-45) above ornamental runs (L). The cadential
close is extended to allow for a dynamic relaxation. The piece ends in a
color almost as soft, though not quite as pale, as it had begun.
WTC II/3 in C major Fugue

The subject of this fugue combines so many unusual features that it is
worth listing them before addressing the details. With just six notes spanning
the length of one measure (from the second eighth-note of m. 1 to m. 21),
this subject is one of the shortest in the Well-Tempered Clavier.2 If brevity
2
The subject of the C-minor fugue in vol. I has only five notes, but these expand over more
than three measures in fairly calm tempo. The subject of the E-major fugue from vol. I is
slightly shorter than this one and also fast-paced but rhythmically far more exciting.
352
WTC II/3
and simplicity seem noteworthy, even more surprising is that this brief unit
appears in stretto from its very first appearance. Listeners will necessarily
assume an even shorter subject, one comprising only the four notes that
sound unaccompanied before the second entry. Although the harmonic
layout proves that this segment is insufficient for a subject, Bach seems to
take the foreseeable misunderstanding into account when he uses this
four-note fragment several times later in the fugue to replace the entire
subject (see particularly in mm. 25-35). If the early entry of the second
subject statement takes listeners by surprise, they must be all the more
astonished to hear that the third entry (M: mm. 2-3) already uses the
inversion. Moreover, almost every subject statement is followed by a motif
that, serving as a regular extension to the entry in the particular section of
the fugue, is imitated in all voices. These motifs, which often sound against
a subsequent subject entry but are actually designed as strung behind the
preceding one, confound listeners expectations for a counter-subject.
Let us now look into details. The pitch pattern comprises a broken
chord followed by a short stepwise descent (C E C G F E). The steps of
this descent are interrupted by rests; hence the overall effect of a melodic
line in separated notes. The rhythm is simple, comprising only eighth-note
values. Rhythmic variety is provided in the course of the fugue by the
regular motifs, which introduce 16th-notes, 32nd-notes, and tied notes. The
subjects harmonic background is as simple as its melodic line. Owing to
the stretto overlap, only the minimal i-V-i materializes: the three initial
eighth-notes represent the tonic; the remaining three notes stand for the
dominant, the dominant-seventh, and the final return to the tonic. The
dynamic follows this simple design. The climax falls on G as it represents
at the same time the middle of the subject, the onset of the only non-tonic
harmony, and the target of the largest leap.
When listing the subject entries of this fugue, one can adopt different
views: Counting only the twelve entries that quote the subject in its full
lengththese will be listed below in bold faceis not good enough for a
structural explanation as it leaves the fugues second half devoid of its
subject. Statements that, although reduced to their incomplete four-note
scope, retain the original rhythm (or its augmentation or diminution) must
therefore be included; these are given below in Roman characters printed
in normal, widened, or narrowed typeface. Inversions are indicated by an
asterisk. Not included among subject statements are figures that resemble
the abridged subject but are falsified in meter or interval pattern, either
because their climax falls on a weak beat or because their contour is not
launched with the characteristic third. This still leaves 34 statements:
C major
1
2
3
4
5
6
7
8
9
10
353
mm.
mm.
mm.
m.
mm.
mm.
mm.
m.
mm.
m.
1-2
1-2
2-3
3
4-5
4-5
5-6
5
5-6
6
L
U
M*
L
U
M
L
U
M
L*
11
12
13
14
15
16
17
18
19
20
mm. 7-8
mm. 7-8
mm. 8-9
m.
9
mm. 9-10
mm. 9-10
mm. 10-11
m. 11
m. 11
mm. 11-12
L
M
U
L
M
U
L*
U*
M
L
21
22
23
24
25
26
27
28
29
30
31
32
33
34
mm.
mm.
mm.
m.
mm.
m.
mm.
m.
m.
mm.
mm.
mm.
mm.
mm.
14-15
15-16
15-16
17
18-19
19
19-20
24
25
25-26
25-26
27-28
28-29
31-32
L
M
U*
L
M*
L*
L
L
U*
M
L
L
U
L
This table demonstrates a number of facts:

There are four blocks of full-length entries, each comprising three
statements (mm. 1-3: L U M, mm. 4-6: U M L, mm. 7-9: L M U,
and 14-16: L M U).
The second half of the fugue does not feature a single full-length
entry. These only occur in the first sixteen of altogether thirty-five
measures.
Incomplete entries in original rhythm also form four regular blocks
(mm. 9-10: L M U, mm. 10-12: L U M L, mm. 25-26: U M L, and
mm. 27-29: L U).
Further regular groups are built by the diminished entries (mm. 5-6:
U M L and mm. 18-20: M L L).
Four shortened subject statements seem outside the groups; interestingly, all of them are in the lower voice (mm. 3, 17, 24, and 31-32).
As was already mentioned earlier, the initial subject entry in the lower
voice is followed by a melodic unit that recurs several times in the course
of the fugue (see L: m. 2 D-CD-E-C-F). It sounds against another
subject statement and could thus be heard as a counter-subject. Yet this
seems not to be the chief purpose of this motif, as it is conceived as an
extension to an entry. This is obvious in the many cases where it occurs
only after but not against a subject statement (e.g., U: mm. 2-3, M: m. 16,
and U: m. 16-17). The same is true for the extension introduced in m. 8
(see L: B to F). It, too, appears always strung after a subject entry, and
354
WTC II/3
recurs as suchwith slight variationsquite frequently in the course of

the fugue. We had therefore better abandon the concept of counter-subjects
for this fugue and speak of motifs. As these motifs partake consistently in
the imitation process, it would also be counter-productive to consider them
purely as secondary material. This becomes all the more obvious as we
discover that both of them are elaborate variations of the incomplete subject itself: compare
the interval structure in the motifs with that in
the four-note answer. Moreover, since the line is
blurred in this fugue between measures determined by subject statements and subject-free
passages, it seems advisable to forego any
attempt to distinguish episodes, and to describe
the process of material development chronologically rather than systematically.
The tempo in this fugue may not be too fast, so that it can convincingly
accommodate the tied 16th-notes as well as the 32nd-note runs at the end
of the piece. The relative tempo of the prelude to the fugue should be
chosen in logical progression from one pace to the next: a quarter-note in
the prelude corresponds with a whole measure in the Allegro, while an
eighth-note in the Allegro corresponds with an eighth-note in the fugue.
(Approximate metronome settings: prelude beats = 64, Allegro whole
measure = 64, fugue beats = 96.)
As ambiguous as the fugues material is its basic character, since the
complex rhythmic pattern and the ubiquitous broken-chord figures seem to
point in opposite directions. Yet the conclusion drawn from the rhythmic
featuresthat the basic tempo is rather calmis supported by a detail
inherent in the subject itself. The rests between the descending notes reveal
this clearly. Had Bach conceived the character of this piece as lively, then
the rests would have been unnecessary, for the three notes, if written as
quarter-notes, would have been played non legato anyway. It is only in
calm basic character that they would be rendered legato and need rests to
indicate the desired separation. For the articulation of most of the notes in
this fugue, however, the distinction between lively and calm character has
no relevance. We are dealing here almost exclusively with two features:
eighth-notes in patterns of leaps (which would be played non legato in
either character), and shorter notes in stepwise patterns (which would be
played legato in either character). The distinction between the two characters only has implications for the longer note values in stepwise motion
above all, the descending syncopations in mm. 30-31: U.
C major
355
The following table describes the events of this fugue in chronological

order. Roman numbers (designating the beginning of a new section in a
traditional analysis) indicate that a harmonic conclusion coincides with
features suggesting a new beginning, like a reduced ensemble, a new stretto,
etc. Owing to the overwhelming use of strettos, traditional criteria for
counting subject entries in a section must be disregarded here.
I
mm. 1-41: Stretto of three complete subject statements (the third of
them inverted), followed by four statements of M1, whereby the
M1 imitation in the middle voice is met by an incomplete inverted
subject statement in the lower voice. Cadential close in the tonic
with dosi(omitted)do formula.
II mm. 4-71: Stretto of three complete subject statements (the third of
them slightly shortened), with extra density achieved by a false
entry (Ms) in the lower voice. Followed by M1 in the leading voice
and three incomplete subject statements in diminution (the third of
them inverted). Cadential close onto the dominant, once again with
incomplete resolution as the target note in the bass is omitted.
III mm. 7-143: Stretto of three complete subject statements, each of
them followed first by M2 and then by an incomplete inverted
entry. Complemented by two further statements of M2 (L original,
M inverted) followed by four incomplete subject entries (L + U
inverted, M + L original). Descending sequences in all three voices
leading to an imperfect cadence (E major = V/vi).
IV mm. 143-25: Stretto of three complete subject statements (the third
of them inverted), condensed with a false entry (Ms) as in section II,
followed first by four statements of M1 (L + M original, U inverted
+ sequenced) and then by two statements of M2 (the first of them
inverted), met by an incomplete subject statement in the lower
voice plus a variant of the false entry (Ms). Complemented by three
diminished subject entries as well as a variant of M2 with two
sequences. Descending sequences. Cadential pattern toward the
subdominant with denied resolution in the upper voice. Ascending
sequences with three entries of Ms in the lower voice and incomplete
entry in the bass met by threefold M2 quotation.
V mm. 25-35: Densest stretto in this fugue, all statements incomplete,
the first inverted, the second augmented, surrounded by M2 and Ms.
Followed by nine M2 imitations and sequences with a new augmented entry. Emergence of a much varied dominant pedal that
defines the music up to the onset of the tonic pedal in the final
measures.
356
WTC II/3
The rise and fall of tension in this fugue describes a regular pattern.
Each section begins with fairly high intensity. The intensity is slightly
higher in the second and fourth sections than in the third because of the
additional false entry with Ms, and higher even in section V because of the
particular density and the augmented statement. In each section, the tension
then decreases. In section I tension declines sharply after the stretto, in
section II it subsides gradually through the diminished entries, and in
section III it abates even more gently through two strettos with incomplete
entries. In sections IV and V the decrease in tension seems to follow that in
section I, but is then prolonged by an extended stretch in moderate
intensity. Sections II and III end in fairly soft tones, while sections I and
IV present material up to the very end and thus do not create so much
relaxation. Only the extended cadential close in section V provides a
complete release.
WTC II/4 in C minor Prelude

The texture of this prelude consists of three polyphonically independent voices. The original thematic idea and its contrapuntal companion,
which characterize the beginning of this piece, recur repeatedly in the
course of the composition, interspersed with independent motifs. What
distinguishes this prelude from others that one would describe as written
in the style of an invention is, above all, the phrase structure within the
thematic material. The main theme, spanning four measures (U: mm. 16-53),
encompasses three separate gestures. Their melodic meanderings are
reminiscent of the expressive lines in a Baroque aria. The high degree of
ornamentation adds to this impression, as does the slow tempo and the
gently swinging 9/8 time. This piece can thus best be characterized as an
instrumental aria in polyphonic three-part setting.
The main theme occurs four times, in two entry pairs connected with
brief links (U: mm. 16-53 + M: mm. 76-116 and M: mm. 336-373 + U:
mm. 396-433). The analogy of the leading entries in each pair extends to the
contrapuntal lines (L: mm. 11-51 . 331-371). There is a strong impression
of recapitulation from m. 33 onward.1 Other structural correspondences
can be discovered, yet they do not continue straightaway from the abovementioned passages: mm. 5-7 recur in mm. 43-45, mm. 11-17 in mm.
56-62, and mm. 27-31 in mm. 50-54.
1
The harmonic layout shows a small but significant difference. In mm. 33-37 the material
sounds on the subdominant. This is a function traditionally used at the beginning of Baroque
recapitulations as it guarantees closure in the tonic without the need of adjustments in the
modulation: I-V becomes IV-I. It comes therefore as a surprise that the imperfect cadence
at the end of the main theme (mm. 4-5: G-C-G) is changed (mm. 36-37: C-F-G7). The
transposition of the melodic lines continues (U: mm. 56-74 . M: mm. 376-394), but the
changed harmonic circumstances allow the second statement of the main theme to enter in
the original key (M: mm. 76-116 . U: mm. 396-433). The end of the phrase is once again
harmonized differently: m. 111 represents V/V while m. 431 is a simple imperfect cadence.
Adding the observations regarding the differently harmonized phrase endings to the fact
that the voices carrying the main theme in exposition and recapitulation are exchanged,
we discern a cross relationship in the symmetrical portions: mm. 1-5 corresponds with mm.
39-43 insofar as in both, the main theme is sounded in the upper voice and harmonized as
V-I-V. Conversely, mm. 7-11 correspond with mm. 33-37 insofar as in both, the main
theme appears in the middle voice and the phrase ending is harmonized as V-I-V/V7.
357
358
WTC II/4
Structurally relevant harmonic closes with perfect cadences can be

found in mm. 16-17 (on G minor, the minor dominant), mm. 26-27 (on
E major, the tonic relative), mm. 32-33 (on F minor, the subdominant), in
mm. 38-39 (on C minor, the tonic), m. 56 (on C minor, albeit in
weak-beat position), and in the final measure.
The preludes aria-like character can best be expressed in a calm mood
combined with expressive intensity. The meter must be read as indicating
compound time, so that the main pulse determining the tempo is the dotted
quarter-note. The appropriate articulation in such a piece is overall legato.
This will include the slow broken-chord figures in the main theme, in its
contrapuntal companion and in any further motivic material. Only genuine
cadential-bass patterns must be taken non legato. This applies above all to
the left-hands latter two-thirds in mm. 4, 16, 26, 32, and 42. Cuts in the
melodic flow occur only in the context of phrasing. These cuts, while
crucial for a subtle interpretation, must be very gentle so as not to disturb
the mood by an exaggerated focus on eloquence.
The many ornaments require in-depth examination. Ornamentation in
compositions of arioso style was often executed on the spur of the moment.
This practice derived from pieces in which the melodic lines appear
enveloped in homophonic or at least less intricate texture, so that there was
little risk of clashes even if the lines were given to different players in an
ensemble. At any given moment, there was usually only one line that
invited embellishments. In a polyphonic composition, the need arose for a
certain agreementbetween the players of an ensemble at any rate. In the
case of a keyboard composition with a single performer, particularly in the
case of a work that the composer expected to perform or teach, the notation
of these conventional ornaments was treated with a certain negligence.
This is why the copies of various pupils show a different distribution and
density of ornaments in the same piece, proving that more than one solution
seemed acceptable in many instances and that Bach took into consideration
the particular disposition (emotional as well as, probably, technical) of
each player. Because of these considerations the composer, who would
expect performers in strictly imitative works like fugues to transfer thematic
ornaments regardless of awkward fingerings, was more lenient in works of
this gentle nature. He thus often avoided the more difficult middle-voice
ornaments and substituted them with an embellishment in the upper voice.
Todays performers, whose technique has been sharpened in tasks of a
much more demanding nature, should therefore be aware that they have the
choice between following the convention of such allowances and observing
a more logical imitative use of embellishments.
C minor
359
Lastly, it is vital to understand that there are a few different ornaments

expressing a similar melodic attitude: an indirect approach to the target
note in the sense of appoggiatura-resolution. These comprise the writtenout grace note (which creates a strongly accented, longer appoggiatura)
and the mordent (which equally features the on-beat appoggiatura, albeit in
a shorter value and repeated pattern). To be aware of their relationship is
important as Bach uses them interchangeably in similar circumstances,
depending on whether a longer or shorter appoggiatura fits better into the
surrounding harmonic design and texture.
Let us begin by comparing the four statements of the main theme and
consider the possibility of retaining or interchanging ornaments depending
on their changed surroundings. The inverted mordent on the second note
does not depend on harmonic preconditions. As retaining it in each entry is
definitely conducive to the listeners recognition of the theme, this inverted
mordent should be transferred to both middle-voice entries (m. 7: C and
m. 33: F). In m. 7, this ornament had best replace the preceding one in the
upper voice to avoid congestion. The same applies to two subsequent
inverted mordents (mm. 2 and 3): if both are played in the initial entry of
the main theme, there is no musical reason why they should be omitted
later. This implies: if the second inverted mordent is added in M: m. 9 on
F, it replaces the ornament indicated for the upper-voice note.
Another case needs more courage. In the fourth entry of the main
theme, an additional appoggiatura (see U: m. 411) seems to disturb rather
than beautify the theme. On second thought, however, it is easy to appreciate why Bach added a grace-note here: owing to the longer note-values in
all three voices, the unembellished version causes a sudden interruption of
the rhythmic flow. In each of the previous statements, this continuous flow
has been taken care of by one of the other voices: in m. 3 by the appoggiatura to the lower voice, in m. 9 by the eighth-note (representing a writtenout appoggiatura) in the lower voice, and in m. 35 by the appoggiatura in
the accompanying upper voice. At m. 411, too, an appoggiatura E in the
lower (rather than the upper) voice would benefit the flowing motion while
leaving the thematic line intact.
The fourth note of the main theme, which in m. 21 is preceded by an
appoggiatura, meets different circumstances in each of the four statements.
While in m. 2, the eighth-note of the ornament resolves in a movement of
parallel tenths with the lower-voice line, a similar ornament in m. 8 would
result in impossible parallel sevenths. In this case the note can either be left
unornamented or embellished instead with a mordent. In m. 34, however,
a middle-voice appoggiatura highlighting the main theme is preferable to
360
WTC II/4
the printed upper-voice appoggiatura. The appoggiatura toward the end of

the theme (m. 4: E) is replaced by a mordent in m. 10. Given the eighth-note
motion in the lower voice this is the only possible solution and should not
be changed. Other corresponding motivic details must be treated in the
same way; in this context they can be mentioned only very briefly.2
Besides the many grace-notes that, given their stepwise resolution and
harmonic independence, represent appoggiaturas, there are some that must
be read as acciaccaturas. These are notes of very short duration falling on
the beat and immediately giving way to the main note they highlight in a
virtuoso (and not in a harmonic) manner. Such notes can be recognized
either as pitches belonging to the same harmony as the main note or as
chromatic sparkles without harmonic value. In this prelude we find four
acciaccaturas (U: mm. 67, 74, 447, and M: m. 387).
2
C The eighth-note appoggiaturas in U: mm. 5-6 cannot be retained in M: m. 37 but should

be substituted with a mordent or, better still because of the approach from below, an
inverted mordent (as indicated by Bach).
The eighth-note appoggiatura of m. 71 can be transferred to m. 391.
In the bass line that accompanies this motif, ornaments may or may not be adjusted to
highlight the sequence. Thus the inverted mordent in m. 6 may be moved from beat 2 to
beat 3, as suggested also in m. 44.
In U: m. 11, either of the two ornaments should be selected. Since an eighth-note
appoggiatura causes the B to sound in octave with the lower voice, a mordent seems
preferable, particularly since a mordent is also possible in the analogous position (M: m.
56), while the eighth-note delay is here excluded.
Later in the same line, both grace-note and inverted mordent can be played in U: m. 13
and M: m. 58, while the simultaneous grace-note in U: m. 58 may be omitted.
The inverted mordent in U: m. 141 should be transferred to M: m. 59 where it sounds
much better than the upper-voice grace-note found in the score.
The cadential mordent in U: m. 16 should be transferred to U: m. 61. This kind of
ornament needed no mention in Bachs times.
In mm. 17-23, ornamentation in the three-part imitation is consistent apart from the
additional inverted mordent on the ending-note in L: m. 23. The brackets suggest that
there were second thoughts, and omitting it is certainly conducive to structural clarity.
In m. 18, the accompanying lower voice features a mordent instead of the inverted
mordent conventionally used in such patterns. This may be a miswriting on Bachs part.
In these measures, the secondary voices may also benefit from less ornamentation.
Consider omitting the grace-notes in M: m. 21, L: m. 25, and U: m. 26.
In the subsequent three-part imitation (mm. 27-30, recurring in mm. 50-53), bracketed
ornaments can safely be ignored (see the second beats in U: m. 27 and L: mm. 29, 30),
whereas the characteristic inverted mordent on the longer note (beat 3) should be
retained and transferred to the middle voice entry (M: m. 28) as well as to the
corresponding passage (U: m. 50, M: m. 51, L: m. 52).
The ornaments in L: mm. 31 and 53, although in parenthesis, are an asset since they
embellish partial sequences.
C minor
361
Finally, there are several long trills. Three of them occur in a chain
and, uncomfortably, in the left hand (mm. 14-15, 19-20, and 59-60). In all
three instances, the second trill is approached stepwise and thus begins on
the (prolonged) main note. The same holds true for the first trill in m. 19.
The other two trills embellish chromatic progressions, thus also suggesting
emphasis on the main notes. As a result, all trills may begin on the main
(16th-) note, then shake four times in 32nd-notes with the upper auxiliary,
and end in a suffix as indicated. The three remaining trills, occurring in U:
m. 31, U: m. 54, and L: m. 50, have exactly the same shapeexcept that
the latter, very softly, shakes twice as long.
The complex question of ornamentation in this aria-style prelude duly
addressed, let us turn to the works layout. The first section (mm. 1-171)
comprises three phrases and a free continuation (codetta). The leading
features are the main theme with its imitation, the contrapuntal accompaniment with its later free variation, and an interlocked motif (M1) with
its own contrapuntal line. The main theme includes three subphrases.
The first is one measure long, with a climax on the appoggiatura on m. 21.
The second subphrase consists of a 3/16-note ascent preparing the climax,
and a subsequent double-note repetition providing some relaxation. The
third subphrase begins like a varied sequence: it reaches its climax in an
immediate leap, its double-note repetition is rhythmically extended, and a
one-measure tail complements the phrase. Of the three climaxes, either the
first or the second can be played as prevalent in the theme. Their metric
placement is intriguing as it creates the impression of hemiolas: they fall
consecutively on beats 1, 3, and 2. After the main climax, the impression
of a gradual decrease should prevail, and the tail is best kept free of any
further accent. The contrapuntal accompaniment to the main theme
contains two subphrases: one ends on the long B in m. 2 after climaxing on
the tied C, while the second is characterized by a protracted decrease after
a climax on the appoggiatura (m. 31). A third voice fills in the texture in
mm. 1-31 but displays more independence thereafter. It creates a tensionincrease toward the appoggiatura (m. 41) before it resolves together with
the other two voices.
On the occasion of the second main theme statement, the contrapuntal accompaniment is strongly varied. With only its second climax
retained (m. 91) it describes a simple curve. The third voice keeps a low
profile during the statement but then adds a shortened imitation of the third
subphrase from the main theme (see U: m. 10-12). The first independent
motif (M1) is introduced in mm. 5-7. Its two subphrases both peak on the
downbeat-appoggiaturas (U: mm. 61 and 71). The accompaniment in the
362
WTC II/4
lower voice (M1a) retains many features of the main theme, particularly
the division into two subphrases and the similarities at the beginning of
each subphrase. The climaxes thus fall on mm. 52 and 62 respectively.
To distinguish the free development in the codetta from the preceding
tight-knit phrases, the dynamic lines should be kept as simple as possible
there. A convincing solution is a long and gradual build-up (from L: m. 11
and U: m. 12 respectively) that peaks in m. 15 and is followed by a
relaxation up to the close of the section at m. 171.
The second section begins, over an accompaniment launched once
more from an ascending broken chord, with a second motif (M2). This
motif, again, contains two subphrases. The first is closely related to that of
M1 (U: mm. 5-6 .17-18), and even in the second we can discover, with
some imagination, traces of the second half of M1 (mm. 6-7 .18-19). Not
surprisingly, climaxes fall on the respective downbeats. As this section is
characterized by the immediate imitation of M2, the performers attention
will be captured by one voice at a time. From m. 23 onward, the lower
voice dominates with partial sequences of the last-heard motif statement.
Climaxes of gradually lessening intensity fall on mm. 233 and 243. The
section is rounded off with a two-measure cadential close (mm. 255-271).
The gradual decline that begins with the third imitation of M2 and
continues through to the end of the section is underpinned by a descending
peak-note line in the upper voice (see U, mm. 21: F, 22: E, 23: D, 24: C,
25: B, 26: A-G-F, 27: E).
The third section begins similarly with a motif in three-part imitation
(M3, see U: mm. 27-28). While M2 recalls M1, M3 appears as a condensation of M1a. Like the second subphrase in M1a, M3 reaches its climax
after the broken-chord ascent on beat 2 and continues in a single unbroken
decrease. As in the previous section, the third motif statement is taken up
in two partial sequences (L: mm. 30-321) and rounded off by a short
cadential close.
It has been mentioned earlier that a large portion of section IV, mm.
33-451, is conceived as a (somewhat irregular) recapitulation of the
thematic passage of section I, and that another passage (mm. 50-55) very
faithfully takes up section III. Between the two, the lower voice retrieves
the habit achieved in the two preceding sections and continues with partial
sequences (see L: mm. 45-481). The other two voices recall fragments of
earlier motifs in variation (e.g., M: mm. 45-46 from M1, partly sequenced
but differently varied in mm. 47-48). A descending peak-note line in the
upper voice creates once more the effect of gradual decline (U mm. 44: A,
45: G, 46: F-F, 47: E-E, 48: D-C-B-B, 49: B-A-G-F-E, 50: D).
C minor
363
After the weak-beat cadential close in m. 56, the seven final measures
represent the coda. As was mentioned earlier, Bach had already used this
coda in the conclusion of the first section.
The internal design of this prelude is thus far from simple. It contains
enough repetitions to serve the intended mood of calm expressive power,
and enough variation to provide for constant surprises.
section
I
II
III
IV
Coda
mm.
1-17
17-27
27-33
33-56
56-62
material
m.th.
M2
M3
m.th.
M3 codetta
M1
M2
M3
M1
M3
m.th.
M2
M3
m.th.
M3
codetta sequences sequences M1
sequences
close
close
sequences
WTC II/4 in C minor Fugue

The subject ends on the middle beat of m. 2 where the dominant-seventh
chord (represented by D, G, and F) resolves onto the tonics third. The
rhythmic pattern is consists exclusively of 16th-notes. These note values and
their relationship to the pulse deserve careful consideration. The fugue is
notated in the compound time signature of 12/16. This should be translated
into a pulse in 4/4 time with triplets in each beat. The pitch pattern must be
interpreted as primarily ornamental. If one assumes a simplified tune
underneath the ornamented surface, it would read C-D-GFE. This
(skeletal) tune reveals more about melodic tension than the busy and
elaborate 16th-note motion.
The subjects harmonic background is ambiguous as Bach harmonizes
each entry differently. The first measure may come as a I-V-I curve or,
particularly toward the end of the fugue, begin on the subdominant and
resolve into the major tonic. In the second measure, one usually finds a
complete cadence with the subdominant or its representative on the
downbeat, followed by the dominant-seventh and the return to the tonic on
beats 2 and 3 respectively. In other entries, the dominant is already reached
at the beginning of the subjects
second measure. This example
indicates only two of the various
harmonizations Bach employs.
364
WTC II/4
The subjects dynamic outline follows from the above observations.

The tension grows from the initial keynote through D, onward through the
low G that builds the basis for the large interval, and up to F. This F at
m. 21, reached thus in a one-measure crescendo, captures all the tensionenhancing features: harmonically, it represents the active step to the
subdominant; melodically, it is the target of the high-tension seventh
interval in the simplified tune or the summit of a powerful thrust upward
expressed in the ascending scale. This climax is followed by a gradual
relaxation through the 4-3 step downward in the simplified tune. (Great
care should be taken not to play Romantic wavesdynamic curves that
follow the ups and downs of the pitches instead of expressing a sense of
purpose. Thus G, although it is the lowest sound, is dynamically powerful
as it propels the run up to the climax.)
The subject appears 16 times in the course of the fugue.
1 mm. 1-2
L
9
mm. 28-29
Linv
2 mm. 2-4
U
10
mm. 30-31
M
3 mm. 5-6
M
11
mm. 48-49
U
4 mm. 16-17
U
12
mm. 53-54
Minv
5 mm. 17-19
M
13
mm. 55-56
L
6 mm. 20-21
L
14
mm. 61-62
M
7 mm. 24-25
Uinv
15
mm. 66-67
M
8 mm. 26-27
Minv
16
mm. 67-69
L
Four of the subject statements are inverted, but none appears in stretto
or parallel. The subject receives a real answer without interval adjustments.
Other modifications, too, occur only rarely. They never affect shape but
only tonality and materialize only toward the end of the fugue. In m. 53,
the thematic fifth interval is diminished to a fourth, and the two notes after
m. 541 are shifted one tone up. In m. 55, Bach introduces an additional
accidental on the note before the climax, and in m. 62 he raises the final
note to the major third. The most dramatic effect among these small alterations occurs in mm. 67-68 where the beginning of the subjects answer
sounds for once not in G major but on the fifth degree of C minor and the
perfect-fifth leap is significantly altered to a diminished fifth. As this
subject entry also picks up alterations heard separately before (the raising
of the note before the climax and of the final note; see B m. 68-69), the
final entry of the fugue sounds harmonically somewhat eccentric.
C minor
365
The only counter-subject Bach invents for this fugue comes with two
surprises. It develops as it goes, passing through several entries and
various attempts before reaching what can later be recognized as its final
shape. Moreover, this component does not feel confined to the passages
reserved for the primary material. Instead, three counter-subject statements
invade the long subject-free passage in the middle of the fugue. They behave like entries in a section and have thus led some analysts to assume a
fugue with two subjects. While we can appreciate that the three countersubject statements in mm. 35-39 create this effect, we must acknowledge
that this thematic phrase was heard before. It is not presented in an independent exposition here (or anywhere else, for that matter).
In its final shape (L: mm. 30-31, U: mm. 35-36, L: mm. 48-49, U:
mm. 55-56, L: mm. 61-62, U: mm. 66-67, and M: mm. 68-69), the countersubject describes a continuous diminuendo. The first four notes are melodic,
whereas the final leaps give the impression of a cadential-bass pattern.
Comparing the other statements with this version, we find the following
variations. The first statement of CS (L: mm. 2-4) enters late, shortening
the initial note. This note is tied and thus launches a diatonic descent
instead of the chromatic one characteristic of the final version. The high
pitch Fneeded here to complement the harmonyreplaces the expected
D. The second and third statements of CS (M: mm. 17-19, U: mm. 20-21)
come close to the final version. The former still omits the chromaticism by
suspending the initial note, while the latter introduces the chromatic descent
but features a variation (a three-note group instead of a leap) toward the
end. The sixth and seventh statements (M: mm. 36-37, L: mm. 37-39)
begin a quarter-note late; in addition, the former entry omits the final note.
Owing to the harmonic modifications in the final subject entry, the partnering CS (L: mm. 68-69) begins and ends half a measure late.
The contrapuntal dynamics of subject and counter-subject are simple:
The fugue comprises thirteen subject-free passages. Some of them are

very short, only filling up the half-measure between the end of one subject
366
WTC II/4
entry and the beginning of the ensuing statement on the next downbeat.
These filler-episodes, linking what should be conceived as consecutive
entries, will be indicated below with an asterisk.
E1 m. 4
E5 m. 25*
E9 mm. 498-52
E2 mm. 68-15
E6 m. 27*
E10 m. 54*
E3 m. 19
E7 m. 29*
E11 mm. 568-60
E4 mm. 218-23
E8 mm. 318-47
E12 mm. 628-65
E13 mm. 69-71
Within these episodes, sequences of the subjects ending play a major
role; they will be referred to as Ms (motif derived from the subject).
Several other components recur frequently and therefore deserve to be
pointed out. One is the conventional melodic closing formula consisting of
a syncopated keynote followed by the leading-note and a downbeat keynote (see, e.g., L: mm. 4-5). In this fugue, this do-si-do group is frequently
used outside cadential conclusions. For convenience, it will be referred to
below as the close. Furthermore, Bach enriches the episodes with four
independent motifs. M1 is first heard in U: mm. 107-126, M2 in U: mm.
117-1212, M3 in U: mm. 13-14 (from B to B, a convex curve ending in a
syncopation), and M4 in U: mm. 448-451. The very short episodes and segments of the two very long ones are built in symmetry: E1 . E3, E5 . E6,
and E7 . E10. Moreover, within E2 and E8, mm. 6-8 . 31-33, mm. 8-13
. 39-44, and mm. 13-15 . 44-47.
The role each of the short episodes plays in the dynamic design of the
fugue follows from the direction of their sequences. Thus E1 and E3 are
decreasing, but E5, E6, E7, and E10 are increasing. All of them link consecutive subject statements. The dynamic layout of the two long episodes
is more complex. Each features a conspicuous peak-note descent marked at
its end with a close and a cadence (E2 mm. 6-9: A-G-F-E-D in U, close
in M, cadence in C minor; mm. 10-13: F-E-D-C-B-A-G in U, close in
M, cadence in E major). In these segments, the tension is diminishing. In
the final segments, however, the direction of the sequences is ascending
and, with rising tension, prepares the ensuing subject entry.
Of the remaining episodes, E4 is divided: mm. 217-227 diminish with
descending sequences in all voices, while mm. 227-23 increase again. E9 is
simple in structure, with only decreasing tendency. E11 sets off with a
cadential close in C minor (mm. 56-57), overlapping with the beginning
of an Ms stretto and a dynamic curve formed by ascending M2 sequences
followed by a cadential resolution. All this occurs over a dominant pedal.
E12 is once again divided in its dynamic shape, featuring a decrease (up to
m. 647) followed by an increase that prepares the next pair of entries.
C minor
367
The subjects ornamental nature suggests a rather lively character. This

is supported both by the consecutive leaps marking the counter-subjects
final shape and by the gigue-like rhythm accompanying the inverted
subject entries in mm. 24-29. The tempo may be fairly swift, but performers should still be able to shape every motif. The tempo proportion
between the prelude and the fugue may be simple since the change from
9/8 to 12/16 time (i.e., from compound triple to quadruple meter) and the
shift from the preludes calm character to the fugues boisterous runs provides ample contrast. Thus a dotted quarter-note in the prelude corresponds
with a dotted quarter-note in the fugue. (Approximate metronome settings:
60 for the dotted quarter-note in both pieces.) The appropriate articulation
entails non legato in the dotted eighth-notes and gigue-patterns (eighthnote + 16th-note) and legato in the 16th-note figures. It is possiblebut
rather demandingto distinguish a tighter legato in the inverted-mordent
figures from a crisper quasi legato in all other 16th-notes. The chromatic
segments in the counter-subject can be rendered in either legato or gentle
non legato. Exceptions from the non-legato touch in the longer note values
occur above all in the do-si-do closes.3
The score indicates four ornaments. All are note-filling trills shaking in
values twice as fast as the shortest notes in the score (i.e., in 32nd-notes)
and resolving after a suffix onto the target note. The first trill (m. 26) is
approached stepwise. It thus begins on the main note and contains one
longer and four shorter notes. The second trill, which appears in the context of an imperfect cadence (see mm. 32-33), is launched from the upper
auxiliary B and encompasses twelve notes including the suffix. The other
two ornaments appear in consecutive measures (mm. 60 and 61). Both are
weak-beat trills on F followed by a middle-beat resolution onto G. This
resemblance makes them appear as an imitation, although the melodic
context in the two voices is otherwise different. A possible interpretation
is that the ornament in m. 60 (which, as the brackets indicate, was added
later) helps, through its relationship with the next trill, to tie the subject
statement in mm. 61-62 to the previous section. Both ornaments should be
rendered as note-filling trills, beginning on the upper neighbor and resolving
with a suffix taken from the melodic minor scale (G-F-G-F-E-F-G).
A last comment concerns the final chord. It is surprisingly short: only
a dotted eighth-note instead of the expected (and, out of negligence, often
3
Some of these melodic closing formulas are easily overlooked, so here is a list: L: mm. 4-5
and 5-6; M: 8-9, 12-13, 39-40, 43-44, and 48-49; U: mm. 19-20, 32-33, 53-54, 56-57, and
70-71.
368
WTC II/4
played) dotted half-note of full-measure length. This rather unceremonious

ending does not invite a pronounced ritardando. A mere hint of a relaxation is most adequate.
Determining the design of this fugue is not too easy. In the absence of
prominent cadential closes and renewed build-ups of the ensemble, interpreters and analysts have to rely on symmetries and on the logic suggested
by the order of the entries.
The striking analogy of episodes E2 and E8 establishes a correspondence between mm. 1-15 and mm. 24-47. The latter appears as
an enlarged modification of the former: Both sections begin with
three consecutive subject statements, and both end with ascending
motifs and scale portions (mm. 13-15: M2 + ascents in M, U; mm.
44-47: M2 + ascents in M, L + M3).
Between these two sections lies a shorter second section that, for
once, begins with one voice resting. Even more convincing is the
correspondence between the beginnings of the first and second
sections: both string together a subject entry on the tonic (mm. 1-2
. 16-17), an answer on the dominant (mm. 2-3 . 17-18), and a
one-measure episode with identical notes (m. 4 U . 19 M and m. 4
L . 19 U). The second section features three entries, the third of
them in the relative major key, and a concluding episode that, at its
very end after an initial descent, musters new energy to prepare the
advent of the next section (see the ascending lines in mm. 22-23).
The fourth section begins in m. 48 with the return to the tonic. Its
three initial subject statements all seem, at first glance, to be rooted
in the tonic. Yet their individual surroundings define their harmonic
background quite differently and reveal that the return to the tonic
occurs only very gradually (see mm. 48-49: F minor to C minor;
mm. 53-54: A major beginning, unresolved ending; mm. 55-56: C
minor). The longer episode E11 with its ascending sequences of M2
announces the conclusion of this section, but the above-mentioned
trill imitation invites listeners to include a further entry, the answer
in mm. 61-62, into this round. Once again an episode ends with
rising tension (mm. 64-65) to prepare the forthcoming section.
The brief final section is launched once more in F minor, thus
picking up the beginning of the fourth section. Its two subject statements appear in the keys and metric distance of the initial entries in
the fugues first and second sections, and even the ensuing episode
measure is very similar (mm. 66-69 . 1-4 . 16-19). The fifth
section concludes with a final cadential ending in E13.
C minor
369
The structural balance Bach achieves in this fugue is worth noticing.

The five sections group into three blocks: sections I + II, section III,
sections IV + V.
Both sections I and II and sections IV and V feature very closely
related beginnings.
In both cases the shorter follow-up sections fulfill the harmonic
purpose not achieved by the larger sectionsin section II the
modulation away from the tonic, in section V the firm re-establishment of the tonic.
The third section provides some contrast with its newly introduced
subject inversions, gigue-rhythm, and the independent round of
entries of the counter-subject.
The three-part layout is further supported by the length of the three
portions:
sections I + II = 23 measures
section III
= 24 measures
sections IV + V = 23 1/4 measures
This fugue is very playful in character, thus dynamic increases and
decreases are gentle rather than dramatic. Most influential in the overall
design are the extensive episodes, whose dynamic impact is based on
descending or ascending sequences or peak-note lines.
370
WTC II/4
What remains to be decided is the relationship between consecutive

subject statements. Within the three entries that open the first and second
sections, the tension rises gradually, supported both by the increase in the
number of voices and, in the third entry of section II, by the modulation to
the major key. A considerable build-up of tension is prevented, however,
by the relaxing attitude of the linking episodes (E1, E3). Section III, by
contrast, sets out with a much lighter character. It is softened even further
in the two-part statement of mm. 26-27 but recovers a little in the ensuing
lower-voice entry. The fourth statement returns to the uninverted shape of
the subject. But since it appears in the weak middle-voice position, it
reaches only moderate intensity.
The fourth section returns to the original touch and color but fails to
create similarly persuasive dynamic groupings. Only the short final section
with its compact design reaches the intensity of the two initial sections.
WTC II/5 in D major Prelude

On the surface, this prelude appears like a polyphonic composition, yet
there are several passages that, like the opening, are distinctly homophonic
in texture. Further observations do not immediately clear these doubts. On
the one hand, we find consistent writing in three parts, abandoned only for
rare instances of voice splitting. Moreover, most compositional techniques
expected in polyphonic worksimitation and instances of contrapuntal
setting, sequencing and inversion of motifs, as well as motivic development
are employed. On the other hand, close inspection reveals that many of
the measures that at first glance seem to feature independent voices
actually represent homophony in disguise: only one voice is melodically
active while the others describe either ornamented parallels or rhythmically
embellished accompaniment figures.
This ambivalence of polyphonic and homophonic elements continues
in the choice of the thematic material and in the structural design. The
ubiquitous presence of a single motif recalls the design of inventions,
whereas the distinctive coloring of the segments within this motif and their
use in separate sections of the prelude, together with the ternary layout of
the work and its harmonic development, suggest a classical form.
The preludes main motif ends with the essential steps of a cadence.
Yet since regular participation of all voices is postponed for four measures,
the harmonic closures on the downbeats of mm. 3 or 5 cannot be counted
as indicative of the structural layout. The first relevant cadential formula
occurs in mm. 12-13, after a modulation to the dominant launched in m. 5.
Thereafter, the dominant key is confirmed in another cadence. The subsequent section appears similarly built. The four opening measures remain
in A major but fail to convince as a structural section in its own right. They
are followed by an extended modulatory process that eventually closes
with a cadential formula in B minor, the tonic relative in D major. A
shorter section brings about the return to the home key. The passage from
m. 41 onward presents a recapitulation. Notwithstanding some small
modifications and harmonically caused adjustments, it is a faithful review
of the preludes first 16 measures. The relevant harmonic processes encompass a modulation to the subdominant in the initial four measures, the
return to the tonic, and the confirmation of the home key in the final four
measures:
371
372
WTC II/5
mm. 1-51
4 mm. opening
D majorD major
mm. 5-131 8 mm. modulation (1 fifth up) D majorA major
mm. 13-16 4 mm. confirmation
A majorA major
mm. 17-211 4 mm. opening
A majorA major
mm. 21-331 12 mm. modulation
A majorB minor
mm. 33-411 8 mm. return to the tonic
B minorD major
mm. 41-451 4 mm. modul. (1 fifth down) D majorG major
mm. 45-531 8 mm. modulation (1 fifth up) G majorD major
mm. 53-56 4 mm. confirmation
D majorD major
The character of this prelude is rather lively. This is indicated in the
broken chords and ornamental 16th-notes as well as in the way Bach
expresses his idea of meter. The dual time signature is revealing: while the
12/8 time reflects typical Baroque reluctance to accept triplets as basic note
values, the alla breve not only hints at the occasional suspension of the
compound metric pattern in favor of a simpler order, but also determines
that the main beat is to be the half-measure pulse. The dual time signature
also serves to eliminate doubt about the value of the eighth-notes: they take
up a third of a quarter-note in all patterns like that of m. 1, but half of a
quarter-note in all patterns like that in the second half of m. 2. Bach seems
to conceive all notes outside the groups of three in the alla breve pulse.
Thus wherever he uses accompanying rhythms in gigue-style, he writes
these in dotted rhythm (see, e.g., mm. 5-7, 12, 16, etc., and mm. 23-26).
These dotted-note groups have the rhythmic value that matches the triplets,
i.e., they stand for a quarter-note-plus-eighth-note combination. (Bach may
have avoided writing this way because he is using the quarter-note as a
larger unithalf of the half-note beatelsewhere and would thus have
created more confusion with two kinds of quarter-notes.)
The preludes tempo should combine two requirements: swiftness (in
the 16th-notes) and stateliness (in the quarter-notes and non-triplet eighthnotes). The triplet eighth-notes act as a link between the two. A good way
to approach them is to make it a point not to think in groups of three but in
groups of six, in accordance with the alla breve indication.
Ornaments occur in various forms. The cadential mordents in mm. 12
and 32 begin on the main note in an interpretation that stresses melodic
steps, but may begin on the upper note if a performer chooses to focus on
the virtuoso element. More melodically defined mordents like the one in
m. 23, however, must begin on the main note. The inverted mordents in
mm. 13 and 33 must fit into the local tonality. They thus touch use the
leading-notes G and A respectively. The inverted mordents in mm. 13,
14, and 15 should be transferred to the corresponding mm. 53, 54, and 55.
D major
373
The long trills need to be carefully timed as they appear in combination

with triplet eighth-notes (m. 19), duplet eighth-notes (m. 20) and tripletbased 16th-notes (m. 40). The tempo of the shakes must be the same in all
cases: triplet-based 32nd-notes. This posits a different number of ornamental notes against each note of the accompanying voice: 4:1 in m. 19,
6:1 in m. 20 and 2:1 in m. 40. The two former ornaments are unresolved
trills ending without a suffix in a tie. The first begins on the main (16th-)
note while the second sets out from the upper neighboring note shaking
right away. Both stop short on the final 32nd-note before the bar line. The
ornament in m. 40, by contrast, is resolved on the following strong beat
and thus ends regularly, with a suffix.1
The prelude begins in almost orchestral colors. The opening downbeatoctavea feature not usually associated with Bachs musical languageis
followed by a figure that, in its combination of run + zigzagging broken
chord, is reminiscent of a fanfare. The effect is completed in the one-bar
cadential pattern in straightforward homophony and with strong emphasis
on metric order that evokes an orchestral tutti picking up after a solo
gesture. Fanfare and tutti cadence are then repeated (see mm. 3-51:
U/M inverted, L an octave lower). From m. 5 onward, the character is
different. The absence of any homophonic doubling and the appearance of
a kind of dialogue between the upper and middle voices seem to suggest a
change of instrumentation from larger ensemble to chamber-music setting.
In this new color, the fanfare is used in a much shorter and therefore less
clamorous version, alternating inversion (M) and original (U). The dynamic
outburst traditionally inherent in fanfares appears softened to a gentle
crescendo that becomes even more subdued in the lower sequence (m. 6).
When the figure recurs in its original length (mm. 7-81: L, 10-111: M,
11-12: U), its character is considerably softened first by the middle-voice
position, then by the inverted shape.
In mm. 5-7 and 10-11, the accompaniment consists of a four-note
rhythmic figure that retains some of the stateliness of the opening measures.
Yet even these drum-beats are interspersed with softening runs (see L:
mm. 5-6, 6-7, and 10-11). Inside the bracket built by the drum-beats, the
1
To further enhance the preludes stately character it is possible (though by no means

necessary) to add ornaments in certain typical patterns, where they are not specifically indicated. Notes that could be thus enhanced are: the first dotted note in a group (see, e.g., L:
mm. 5 and 6, second quarter-note, etc.) and the final treble note in otherwise unornamented
cadential patterns (see, e.g., U: mm. 16, 56). Such additional ornamentation emphasizes the
ceremonious aspect of the piece; an interpretation that aims at highlighting the polyphonic
features would therefore not make use of them.
374
WTC II/5
two original components of the fanfare, run and zigzagging broken chord,
are split between the upper and lower voices in m. 8, while only runs are
left in m. 9. A cadential formula closes the development in mm. 12-13.
The four measures preceding the repeat sign present a descending
sequence (compare the second to ninth eighth-note in m. 13 with the same
segment in mm. 14 and 15). The material consists of varied fragments of
the fanfare (descending runs and a three-note broken chord) along with
the drum-beathere with an ornament replacing the dotted-note group.
The fourth measure adds a renewed cadential close in the dominant.
With regard to this preludes material and form, two observations are
crucial:
All components of the thematic material, even when expressing
different colors, are derived from the initial two measures. The
fanfare may appear split into fragments of various lengths, the
homophonic cadential formula may assume different shapes, and
the assertive opening octave beat may turn into different drumbeat patterns. Yet all can be traced back to a single two-bar cell.
The overall design reveals a ternary form, with some specific attributes: The initial section is not only repeated but also recurs in
corresponding design and material. The significant difference lies
not in the inversion of voices (U/M: mm. 1-9 . M/U: mm. 41-49)
or the few melodic variations (compare mm. 13-16 with 53-56), but
in the harmonic development. The combination of a repeated
exposition that modulates to the dominant, is then followed by a
middle section, and is finally complemented by a recapitulation that
begins and ends on the tonic, suggests sonata form.
We are thus dealing here with a prelude in sonata form based on a single
thematic cellin other words: a monothematic sonata movement.
The middle section fits well into the notion of development section.
Its beginning in mm. 17-20 is based on the main motif, i.e., the
combination of octave-beat, fanfare, and cadential close. For the first
time in the prelude, the entire two-bar unit appears in inversion and at the
same time contrasted with a counterpoint (U: mm. 18-20). In mm. 21-331,
Bach develops the chamber-music version of the thematic components
as it appeared in mm. 5-141. Familiar passages with only little modification
(see e.g. U/M: mm. 21-231 and L: mm. 23-251) are set against new material
(see L: mm. 21-231 and U/M mm. 223-251). The combination is then
further developed (mm. 25-26), gives way momentarily to disguised
parallels and virtuoso figures (mm. 27-27), and ends by returning to
familiar grounds (m. 30 . m. 11, m. 31 . m. 10, m. 32 . m. 12).
D major
375
The final portion within the development section begins in m. 33 in the

same way as did, in m. 13, the third portion within the exposition.
Thereafter, however, mm. 34-40 display new material that prepares the
transition to the recapitulation in simple pitch lines (mm. 34-35: descending sequences; mm. 36-37: ascending sequences; mm. 38-39: descending
sequences).
WTC II/5 in D major Fugue

Beginning on the second eighth-note of an alla breve measure, the
subject of this fugue concludes after three half-note beats, in the middle of
m. 2. Its line falls from the octave to the third of the D-major scale. (The
harmonic resolution, the logic of metrical organization, and also Bachs
further use of the subject prove that the D on the last quarter-note of m. 2
does not belong to the subject.) In terms of phrase structure, the subject
consists of a simple, indivisible unit. This is particularly noteworthy in this
piece as the second half of the subject, used as a motif, permeates the entire
composition so densely and frequently that one might be tempted to regard
it as separable from the initial measure. Yet the indivisibility is corroborated by the harmonic layout. The subject is designed as a simple I-ii-V-I
cadence. The repeated D represents I, followed by ii in the second half of
m. 1. With the beginning of the new measure,
i.e., during the course of the tied syncopation,
the harmony changes to a V9 chord and resolves to I on the middle beat of this measure.
The rhythmic pattern in the subject comprises eighth-notes, quarternotes, and a syncopated note of 3/8 duration. In the further course of the
fugue, Bach adds merely a negligible number of 16th-notes. The pitch
pattern in the subject consists of a tripartite note repetition, a broken chord,
two consecutive larger leaps (a perfect fifth and a perfect fourth), and only
at the very end two small intervals.
The dynamic development expresses both the harmonic design and the
particular metric/rhythmic organization since the two coincide with regard
to their tension-enhancing features. The tied note B is the obvious choice
for a climax as it is not only a syncopation (and as such the most striking
feature in this short unit) but also represents the shift from one harmonic
step to another, a fact conferring additional tension to this note.
The fugue comprises twenty-four subject statements.
376
WTC II/5
1
2
3
4
5
6
7
8
mm. 1-2
mm. 2-4
mm. 5-6
mm. 6-7
mm. 10-11
mm. 11-13
mm. 14-15
mm. 14-16
T
A
S
B
A
S
A
S
9
10
11
12
13
14
15
16
mm. 21-22
mm. 22-23
mm. 22-24
mm. 25-26
mm. 27-29
mm. 27-29
mm. 28-29
mm. 33-34
T
S
A
B
B
S
A
T
17
18
19
20
21
22
23
24
mm. 33-34
mm. 33-35
mm. 40-41
mm. 43-44
mm. 44-46
mm. 44-46
mm. 45-46
mm. 45-46
A
S
T
B
S
A
T
B
Only one of these entries is varied: the alto statement in mm. 28-29
substitutes the syncopation with a rest and transposes the four final notes
a fourth up. Modifications of the beginning or the end do not occur. There
are no interval adjustments in the answer, and the subject is never inverted.
Strettos, however, are frequent and involve from two entries (introduced
already in the first section; see mm. 5-7) to all four parts (see mm. 44-46).
Bach does not invent a counter-subject. The fugues secondary material
derives entirely from the subject itself, thus giving this piece a singularly
dense and unified atmosphere. The subjects four-note ending (Ms for
simpler reference) appears both as a counterpart to the subject and as an
episode motif. When accompanying the subject it can appear in three
positions: as a rhythmic complement, beginning after the fourth subject
note (see mm. 3, 5, 6, 10, 21, 25, and 40), as a rhythmic parallel to the
beginning eighth-note motion (see mm. 5, 11, 14, 21, 25, and 33), or as a
stretto imitation to the subjects ending (see mm. 7, 29, 34, and 41).
D major
377
The fugue encompasses ten subject-free passages.

E1 mm. 4-51
E4 mm. 16-211
E7 mm. 292-331
E2 mm. 72-101 E5 mm. 24-251
E8 mm. 35-401
E3 mm. 13-141 E6 mm. 262-272
E9 mm. 412-43
E10 mm. 47-50
One of the episodes is subdivided: in E4 we distinguish a principal
segment that closes with a cadential formula in the middle of m. 20 from a
half-measure appendix that prepares the next subject statement. In the
absence of independent material, no episode presents a color contrast to the
subject-determined portions. On the contrary, all sound almost like extensions to a preceding entry. This effect is achieved by Ms, the four-note
group from the end of the subject.
Among the episodes we can distinguish four groups. A first group
begins as a partial sequence, with Ms in the voice that just presented the
subject.2 The three episodes in this group are identical in length and
analogous in structure. All serve as bridges between consecutive statements
within a section. Their dynamic tendency is slightly rising. A second group
begins in stretto imitation, often involving many entries of Ms.3 These
episodes are longer. The striking density of their strettos is counteracted by
descending peak-note lines (see E2: A-C, E7: E-G, E9: D-F). While the
first of these lines features a closing formula (see mm. 9-10), the conclusion
is transitory. (The chord at m. 101 comes with a sudden thinning of the
texture, brought about by the unexpected rest in the tenor.) The other two
episodes of this group have linking function. The dynamic tendency of all
three episodes is that of a very gradual lessening of tension. A third group
follows a stretto of subject statements with Ms strettos.4 The three episodes
in this group are long and complex. Each concludes a section. This is
obvious in the case of E10 at the end of the fugue. In the case of E4 it is
supported by the closing formula (see S and B in m. 20). E8 does not show
2
E1: subject ending in A, Ms sequence in A, imitated in T. E3: subject ending in S, two Ms

sequences in S. E5: subject ending in A, Ms sequence in A, imitated in T.
3
E2: subject ending in B, imitated in uninterrupted stretto with A/T/S/B/A/T, non-stretto

imitation in B. E7: subject ending in A, imitated in uninterrupted stretto with T/S/A/B/S/T/A,
second stretto with A+T/B, non-stretto imitation in B. E9: subject ending in T, imitated in
stretto with A/Svar, non-stretto imitations in T+B and in S.
4
E4: ending of subject stretto (A/S) imitated by stretto 1 with A/S/A, stretto 2 with T/S/T,
stretto 3 with B/S/B/A/T, non-stretto imitation in B. E8: ending of subject stretto (T/A/S)
imitated by stretto 1 with T/A/S, stretto 2 with T/A/S/T/B/A/B/S/A+B/S/A, non-stretto
imitations in A and in S. E10: ending of subject stretto (S/A/T/B) imitated by stretto 1 with
T/S/B/A, stretto 2 with T/B/A, stretto 3 with T/A/S, non-stretto imitation in T.
378
WTC II/5
any obvious closing features, but it distinctly marks the return to the tonic
(see particularly the two-octave descending scale in the bass of mm. 38-40).
Structured in themselves, these three episodes are also the most selfcontained in their dynamic shape: E4 describes a curve, beginning with a
buildup (see the special treatment of texture in mm. 16-18) and ends with
a definite relaxation; E8, despite its descending D-major scale at the end,
strikes listeners with steady ascents (see the chromatic ascent in the bass of
mm. 35-36, followed by ascending sequences, and the rising peak-notes in
the soprano). The conclusion of this section is thus prepared as a triumph.
Conversely, E10 features only descending lines and thus completes the
piece on a soft note. A fourth group serves only to prepare or close a subject
statement.5
The subjects pitch pattern with note repetition, broken chord, and consecutive fifth and fourth intervals indicates a rather lively character, as
does the rhythm that, as mentioned earlier, is simple throughout the piece.
The alla breve time signature confirms this reading. The tempo, however,
finds its upper limit in the desired clarity of the material: the density of the
Ms strettos can be neither properly expressed nor fully perceived if the
pace is too fast. The two beats in each alla breve measure can therefore
best be imagined as generously swinging. In view of the existing contrast
in prevailing note values, the tempo proportion between the prelude and
the fugue may be simple: half a measure in the prelude corresponds with
half a measure in the fugue. (Approximate metronome settings: 92 for the
dotted quarter-notes in the prelude, 46 for the half-notes in the fugue.) The
articulation includes non legato, quasi legato, and legato. Within the non
legato we should distinguish between more definite cuts after the quarternotes (fourth subject note) and smoother interruptions after the syncopations
smooth enough to guarantee that the syncopation can truly be heard as
swallowing up the strong beat. The quasi legato, too, may sound denser
in the ending of the subject (Ms) than in the note repetitions. True legato
applies to the few 16th-notes but also, more importantly, to the closing
formulas (see U: mm. 20, 27, and 44; M: m. 50). The score does not
feature any ornaments.
The fugue consists of five sections. The development of tension inside
each section as well as in the entire fugue is determined by the occurrence
of strettos and by the tendencies expressed in the episodes.
5
E4a: a single statement of Ms serves, together with the two bass notes, to modulate from
A major to B major. E6: cadential formula in F minor, for once no quotation of Ms.
D major
379
Section I comprises the initial round of statements, the linking E1, and
E2, which concludes the section in the dominant key. The tension rises
through the subject entries but declines in E2. Section II presents two single
entries followed after the bridging E3 by a stretto of the same voices. E4
closes this section, once more in the dominant key. By unexpectedly
weakening the cadential resolution in m. 10 with a resting tenor, Bach
combines the two sections. Their union is also fortified harmonically: the
weakened A-major cadence of m. 10 is taken up in a fully-voiced close at
the end of section II. Beginning in four-part texture, this section picks up
from the raised level of the previous development. The linking episode
with its ascending sequences of Ms aids the rising motion. The growth is
furthered by the advent of the first stretto and is doubled shortly in the first
portion of the ensuing episode. Only then does the tension abate. The first
and second sections together thus build a pair not only on the harmonic
level but also with regard to the development of tension.
Section III consists of the half-bar modulation in E4b, four partly overlapping entries, the linking episode E5, and the cadential close in E6. It
relates harmonically to B minor/major and F minor/major. The beginning
launches a new tension development, enhanced by the change of harmony
and the thinning of texture. Although the sequence of the three opening
entries dramatically raises the tension level, the mood is more restrained
here than in the preceding section. The subsequent single statement and the
simple closing episode underpin the lessening of drive in this short section.
Section IV encompasses two tightly interwoven three-part strettos and
two longer episodes with multiple Ms strettos. It begins harmonically in
the dominant (stretto at the octave), progresses through the subdominant
(m. 331, also m. 351), and ends with a return to the tonic. Section V begins
with two single subject statements linked by E9. It is then crowned by the
triumphant quadruple stretto, after which its tension subsides gradually
throughout the descending lines of the final episode E10.
Sections IV + V together form a large pair, corresponding to that built
by sections I + II: Just as the A-major cadence in m. 20 completes the transitory A-major cadence in m. 10, so does the D-major cadence in the final
measures complete the return to the tonic that at the end of section IV lacks
the support of any cadential formula. Just as the four single entries of
section I were surpassed not by the two single entries at the beginning of
section II but by its stretto, so the two three-part strettos in section IV are
outdone not by the two single entries at the beginning of section V but by
its four-part stretto. The consistent decline of the pitch line in the final
episode is understandable after such a powerful buildup.
380
WTC II/5
WTC II/6 in D minor Prelude

This prelude is in two-part texture. The material includes components
usually indicative of homophonic style (the broken-chord accompaniment
patterns in mm. 2-4) as well as a number of motifs that are developed
evenly in both parts. Counterpoint and imitation are amply used.
The first harmonic progression closes at m. 51. The first structurally
relevant harmonic close is not reached until m. 261, where it affirms the
modulation to the dominant. The prelude then continues uninterrupted as
motifs not only follow each other but also develop from one another in
such a way that distinguishing sections seems futile. The subdominant
field within the large-scale harmonic development is established in m. 42,
but it immediately gives way to a dominant pedal. This launches a liquidation process that proceeds through various steps up to a final tonic pedal.
The preludes layout can thus be presented as follows:
Ia
mm. 1- 91
tonic confirmed
Ib
mm. 9-261
modulation to the dominant
II
mm. 26-421
modulation to the subdominant
III
mm. 42-61
return to the tonic
The character is determined above all by the uniquely homogeneous
rhythmic pattern. The only interruptions of the continuous 16th-note motion
occur not because of longer note values that would allow breathing space
but in favor of brief ornamental runs in 32nd-notes (see mm. 22, 24). The
rhythmic simplicity is matched by a pitch pattern consisting predominantly
of broken-chord figures, scale passages, and ornamental figures. The
preludes basic character is thus rather lively. The tempo may be swift, not
hurried but in a pace that discourages listeners from focusing on single
16th-notes but that conveys patterns like those in mm. 2 and 3 as units.
Articulation is crisp. The non legato in the eighth-notes should rebound
without appearing weighty. Apart from the few groups that Bach marks
legato (U: mm. 22 and 24; L: mm. 43-45), all faster note values should be
played quasi legato, resembling the tone quality that in Viennese Classical
style was to be called leggiero.
The only ornaments indicated are inverted mordents. They use for their
lower neighbor the pitch belonging to the tonality of the measure: in mm.
1-3, 44, and 50 where the key is D minor, the ornaments touch the leadingnote C, whereas in mm. 27-28 where the secondary key A minor is
381
382
WTC II/6
reached, G and not G is the proper auxiliary. In mm. 43 and 45, however,
the inverted mordents use the whole-tone step.
The falling octave in the lower voice at the beginning of the prelude
never recurs and should therefore be regarded as apart from the thematic
material. Moreover, on two strategic occasions (see mm. 5 and 26) the
descending scale is conceived as an upbeat feature beginning in a metrically weak position on the second 16th-note of a measure. One might thus
contend that the prelude begins with a kind of assertive signalthe
triple Dand only launches its motivic material from the next weak beat.
This material encompasses four distinct motifs. One of them appears with
a specific accompaniment that it retains, a second is escorted by varying
non-motivic note-groups, and the other two are introduced in stretto
imitation, thus not leaving room for any contrapuntal figure.
M1 consists of a melodic strand in eighth-notes (L: mm. 2-51) and an
accompaniment (M1a) in which a descending scale triggers a broken-chord
motion. While M1a is harmonically determined and reaches its climax on
the subdominant chord (m. 2) followed by a diminuendo until the end of
the motif, the leading part of M1 follows melodic features. The rests
together with the sequential layout at the beginning indicate sub-phrasing.
The climax of the entire motif falls on the B in m. 4, a note that is
preceded and followed by high-tension intervals (a minor sixth and a
diminished fifth respectively). In this motif m. 3, a kind of harmonic
variation of m. 2, takes part in the gradual buildup to the climax and should
therefore increase in intensity.
Both voices come to a relaxed ending on the downbeat of m. 5. The

repetition in inverted voices begins with a renewed secondary buildup of
M1a, followed by a reiteration of the processes described above. Later in
the prelude, M1a recurs in L: mm. 26-301 and U: mm. 30-341, while M1 in
its original melodic shape is heard again only once, in U: mm. 27-301.
Fragments of the motifs melodic material can also be recognized in L: m.
42, U: m. 43, and L: m. 50.
D minor
383
M2 is also introduced in the lower voice. Beginning on the second

16th-note of m. 9, it describes a curve that culminates on D (m. 93), later
backtracks, and is rounded off by a broken chord also in the shape of a
curve (mm. 103-111). In an interpretation emphasizing pitch patterns, the
tension peaks on this D. Conversely, an interpretation that supports metric
features will place the climax on the B at m. 101. In either case, the climax
is followed by a gradual decrease of tension that approaches the motifs
final note motif without passing through any further accent. M2 is set in
stretto imitation: it recurs in U: mm. 10-121 (ending varied, with tied note)
and in L: mm. 11-131. Furthermoreunusual for an imitative processwe
can recognize two halves that separately fill up the spaces left vacant at the
beginning and end of this stretto (see the second half in U: mm. 9-101 and
the first half in U: mm. 12-131).
M3 presents itself under structural conditions very similar to those
found in M2. Introduced in the lower voice, it is imitated in stretto. Its first
complete statement is heard in the lower voice of mm. 13-151. Its initial
measure consists of a 2/16 upbeat to a broken chord in zigzag (see mm.
13-14: D-F-B-D-G) complemented by a continuous 16th-note motion
continuing the zigzag line. An interpretation aiming at revealing actual and
hidden pitch shapes should place the dynamic climax on the note at the
outset of the descending broken-chord pattern, the D at m. 132, whereas
performers who have chosen to focus on rhythmic and metric features will
increase tension until m. 141 and decrease thereafter. M3 recurs in U: mm.
14-161, in L: mm. 15-171, and in U: mm. 16-181. As in the case of M2,
there are also half statements (see the second half in U: mm. 13-141 and the
first half in L: mm. 17-181).
M4, the last thematic component in this prelude, is introduced much
later and, to make it stand out even more, in the treble (U: mm. 35-371). In
its first measure, the continuous 16th-notes are conceived in hidden twopart structure, displaying a peak-note line ascending before the backdrop
of a repeated E-D-C. The harmonic resolution of the thus described
diminished seventh chord F-A-C-E materializes on the subsequent
downbeat (with B-G for G minor), followed by a descending G-minor
scale that, after a bend, ends on C. For the dynamic representation of this
process there are again two options. One can emphasize the harmonic play
of tension and relaxation by taking m. 361 softly, as the beginning of a
resolution. Or one can choose the metrically oriented version in which the
climax falls on the strongest beatwhich here, moreover, contains a
melodic surprise as the central B deceives the listeners expectations for
a continuation of the ascending peak-note line. M4 differs from the
384
WTC II/6
preludes other motifs in that it does not come accompanied by a fixed

figure in the second part. The imitation sets in after the original figure has
ended, and the eighth-notes of mm. 35-36 never return in the same shape.
M4 recurs, with slight variations at beginning and end, in L: mm. 37-391.
Thereafter, only its first half appears in U: mm. 39-401. (The descending
16th-notes in L: mm. 40-41 tempt us to hear them as a continuation of the
motif in octave displacementwhich, strictly speaking, they are not.)
The structure of the D-minor prelude is determined by the appearance
of the four motifs and their respective inherent impetus. Section I displays
a lineup of the motivic material followed by an extended cadential figure.
The main motif in its original setting and in inverted voices respectively is
presented in mm. 1-51 and mm. 5-91. This repeated statement of M1 is
balanced by the two strettos of M2 (mm. 9-131) and M3 (mm. 13-181) that
effect the modulation to the dominant key. In the following pattern, which
exposes a chordal progression in hidden two-part structure, four measures
with an indirect pedal on A (see the peak notes in the upper voice) are
matched by four measures with an indirect pedal on D. In m. 25, the treble
D has turned into the seventh of an E-major chord and thus leads to the
completion of the modulation to A minor at m. 261. With regard to the
development of tension, M1 is clearly the strongest thematic component in
this section, both because of its greater length (four measures versus only
two each in M2 and M3) and because of the extended dynamic buildup
within its melodic part. The two smaller motifs both feature descending
sequences. While the tension may momentarily flare up at the outset, it
soon subsides, reaching a level of quiet suspension at the beginning of the
pedal-note passage. Throughout the eight cadential measures, the tension
then rises, reinforced by the sudden 32nd-note flushes in mm. 22 and 24,
until the dynamic level of the preludes beginning is regained.
Section II begins in an intensity similar to that of section I. Yet while
the immediate recurrence of M1 sounds as triumphant as before, its
repetition in inverted voices is so strongly varied that the same character
cannot be maintained. An inserted extra measure, reminiscent of the initial
run of M1a, seems to launch a new attempt but leads instead into M4. The
section concludes after two measures serving both as a complement to the
aborted M4 sequence and as a transition, at m. 421. The decline of dynamic
intensity is much more drastic in this section than in the previous one. It
begins with the variation of M1 in which the lack of the characteristic
melody creates a thematic void. The impression of reduced polyphonic
density continues throughout the measures dominated by M4 where only
one voice seems relevantas supported by the switch to parallel texture in
D minor
385
mm. 38-39. In the transitional measures, the lower voice continues the
decreasing tendency while the upper voice opposes this with a stringent
increase in (mostly chromatically) rising eighth-notes.
Section III differs from the two preceding sections in almost every
respect. No motif is quoted in its entirety. Instead, the initial measure with
fragments of M1 is immediately followed by a development that combines
faint remembrances of M1 with the extended pedal concluding section I
(see A in L: mm. 43-46 and in U: mm. 47-49). Two measures of hidden
two-part structure with parallels recall the end of section II (mm. 51-52 .
38-39) and lead to four measures in toccata-style (mm. 53-56). After the
home key is reestablished in its major mode (Picardy third from m. 57
onward), the prelude ends with four cadential measures on a bass pedal D
and a final chord that, with its four voices, confirms the four-part structure
indirectly evoked in all the preludes non-polyphonic passages. Dynamically, the third section describes a prolonged decrease of tension. Caused
primarily by the absence of thematic substance, this decrease is further
supported by a descending peak-note line (see treble E: m. 43, D: m. 44,
C: m. 45, B: m. 46, A: mm. 47-49, G: m. 50, F-E-D: m. 51, E-D-C: m.
52, B: m. 53). While the toccata-like measures describe superficial waves
rather than a consequential increase in tension, the final development over
the tonic pedal musters the force to counteract the decreasing tendency and
conclude the prelude on an assertive note.
WTC II/6 in D minor Fugue

The subject begins on the downbeat and ends at m. 31, thus extending
over a little more than two measures. The two extra eighth-notes following
the final note before the rest have neither harmonic function nor do they
ever recur. They should therefore not be regarded as part of the subject.
The F on the downbeat, on the other hand, is indispensable to complete the
D minor cadence.
A description of the pitch and rhythmic patterns in this fugue depends
very much on the interpreters personal preference. There are two possible
interpretations that will influence how the basic character is determined.
Both can be supported by sound reasons: One can describe the rhythmic
pattern as consisting of a variety of note values (eighth-notes, 16th-notes,
triplet 16th-notes, and syncopations), and the pitch pattern as containing
predominantly stepwise motion. Or one can argue that while there are three
386
WTC II/6
regularly occurring note values, the rhythmic pattern can nevertheless be

regarded as simple in its overall impression, particularly since the triplets
conceived in stepwise motion display ornamental character.
The subjects phrase structure also allows for alternative interpretations
(plus some variation in one of them). As each of these interpretations is
linked to a particular view of the preludes character, it is vital not only to
recognize the various possibilities but to be aware of the ground on which
each of them stands, and the consequences each of them demands.
The subject may be regarded as an indivisible unit. This view can
be convincing in a very lively character.
The subject may also be interpreted as consisting of two subphrases,
with the first subphrase ending on the first of the eighth-notes, the
A at m. 13, which is the natural target note of the winding ascent in
the 16th-notes. The second subphrase then comprises the descent
from the octave D to the final F. In a rendering in lively character
this concept is possible, in rather calm character it is probably the
only meaningful interpretation.
On basically the same grounds, the subject can be described as
encompassing three segments: a first subphrase from the beginning
to the A at m. 13 (as described above), a second subphrase whose
main body ends on the F after the tied G, and a cadential tail. The
argument runs as follows. The syncopated G acts as a harmonic
embodiment of both the subdominant and the dominant chord, after
which the subsequent eighth-note F brings the return to the tonic.
So we have a point for regarding the subjects main body as ending
here. Yet metrically this is unsatisfactory. Moreover, it would seem
somewhat arbitrary as the remaining three notes with their cadential
skips, which present a reassuring orientation after so much chromatic mystery, provide a convincing conclusion. To regard these
notes as a tail might therefore not be such a bad idea: it shows
that they are not essential although they belong, that they are
somehow self-contained (owing to their harmonic background) but
not weighty enough to constitute a subphrase of their own, and that
they can occasionally be omitted without damage to the main body
(as happens, e.g., in mm. 10-12, 14-15, and 17-18).
The subjects harmonic layout is very complex, particularly in the
second half where the high degree of chromaticism creates altered chords
and chords of secondary order. Bachs own realization displays harmonic
shifts on every eighth-notewith only one exception, to which we shall
return). While this constant change makes the search for what could be
D minor
387
determined as the main harmonic functions difficult, it contains a message

of its own that is worth taking into consideration. On the one hand, such
dense harmonic action requires time to be appreciatedand thus rules out
the fast tempo that the 16th-note runs might otherwise tempt us to adopt.
On the other hand, the only exception from the constant harmonic changes
occurs on the third beat of m. 1 where both the A and the D stand for the
tonic. This supports the view of the subject consisting of two subphrases,
both of which display a roughly similar harmonic design:
As to the dynamic outline, any suggestion would obviously have to be

in keeping with decisions made in other respects. The following options
seem open:
For performers who, despite the dangers for an adequate appreciation of the rich harmonic fabric, opt for a fast tempo with no subphrasing, a consistent dynamic design places the climax on the
highest note D, with a crescendo in the ascending runs and a
diminuendo in the gradual descent. This interpretation is based on
the view of the subjectand, to a degree, of the entire fugueas a
virtuoso display in which pitch curves dominate. (Harmonic details,
however, are overrun in this interpretation.)
For interpreters who opt for a moderately fast tempo and a divided
subject, the first subphrase also exhibits a crescendo in the ascent.
The second subphrase can then give credit to the syncopation and
its harmonic background, by having the tension increase in the
chromatic descent up to G and decrease gradually thereafter.1
For those who pledge that in addition to two subphrases there is a
cadential tail, the final decrease after the syncopated G must be
fairly abrupt, so as to allow for the subsequent F to sound fully
relaxed. The three final eighth-notes E-A-F then sound in a neutral,
dispassionate tone color.
1
One occasionally hears the very subtle version that increases tension in the first subphrase
only up to the B before resolving into the tonic on A. This is beautifulas long as it does
not lead to a similar up-down on the first beat, which would with one stroke ruin the
subphrase by making it reminiscent of the emotional waves of 19th-century music.
388
WTC II/6
The D minor fugue comprises nine complete subject statements.
1
mm. 1-3 M
5
mm. 14-15
M
2
mm. 3-5 U
6
mm. 14-17
U
3
mm. 6-8 L
7
mm. 17-18
M
4
mm. 10-13 U
8
mm. 17-18
L
9
mm. 25-27
U
None of the further statements sounds exactly like the initial one. The
second and third entries set out without major modifications, but the final
note concludes the melodic cadence even more definitely than at the
beginning of the fugue by falling back to the keynote. The fourth subject
statement breaks off immediately before the expected syncopation and,
after an octave displacement, sequences the eighth-note descent but alters
the ending and breaks off unresolved. Statements 5 and 6, as well as 7 and
8, are positioned in stretto, with the latter pair using the subject inversion.
None of these entries is complete. The fifth entry anticipates the syncopation (M, m. 15: on A instead of G) and then breaks off suddenly. Its
partner corrects the position of the syncopation (U: m. 15) but, instead of
resolving afterward and thus bringing the subject to a close, continues and
expands the chromatic descent in sequences of the rhythmic figure. In the
following stretto with inverted entries, the leader behaves like the leader of
the preceding stretto: it anticipates the syncopation, lengthens it and then
breaks off (M: m. 18). Its partner is even less complete, as it does not reach
any syncopation but gives way to another motif halfway through the
second segment. Only the statement that concludes the fugue recaptures
the original version. Yet while the syncopation is in place, the cadential tail
is substituted by a melodic closing formula.
This fugue features only one proper counter-subject. It is introduced
from m. 3 onward where it begins immediately after the middle beat. Its
end is unclear. One feels that the counter-subject was meant to end with a
closing formula AG-A. But Bach obviously decided to deviate from this
expected ending, to write G instead of G and delay the final note, which
now resolves late and into another key. (In M: m. 5, the resolution comes
after a suspension on the second 16th-note and suggests not a chord on the
dominant A but one on the tonic D. The ending on the 16th-note F is thus
the logical conclusion, although it is quite impossible to convey this in
performance as suspension and resolution overlap with the beginning of
the next motif.)
D minor
389
In terms of phrasing, this counter-subject can be interpreted either as

indivisible or as consisting of two subphrases, with the second beginning
in the form of a varied sequence after the downbeat A. The dynamic layout
allows for a host of possibilitiesseveral within each of the different
options for the phrase structure. If the counter-subject is interpreted as an
indivisible phrase, the climax can be chosen on the basis of melodic,
rhythmic, or harmonic criteria and thus fall on the high-tension interval
(the tritone C-F in m. 4, F-B in m. 7), on the syncopation, or on the
subdominant harmony, each time preceded by an unbroken crescendo and
followed by a diminuendo. Conversely, if the counter-subject is interpreted
as consisting of two subphrases and interrupted by a gentle caesura after
the first downbeat, a first climax falls on the downbeat and its subdominant
harmony can be combined with a second climax on the syncopation.
The following example gives the counter-subject as it sounds against
the tonic version of the subject in mm. 6-8. Listed here are only two of the
manifold contrapuntal patterns that result from the different options for
phrase-structure and dynamic outline in subject and counter-subject.
There are five subject-free passages in this fugue.

E1 m. 5
E2 mm. 8-103 (M overlaps, beginning already in m. 73)
E3 mm. 13-141 (M overlaps, beginning already in m. 123)
E4 m. 16
E5 m. 183-253
All the material that appears in these episodes either derives from the
subject and the counter-subject or is introduced against it (as a fake
counter-subject). E1 quotes (M) and imitates (U) the inversion of the
subjects first phrase segment, with the second half of the subjects initial
measure in retrograde serving as counterpoint. E4 takes up an excerpt of
the counter-subject, juxtaposing it to descending lines reminiscent of the
subjects second segment. E2 uses an independent motif introduced just
before in the context of the last half measure of the subject/counter-subject
juxtaposition (see M: mm. 73-81). The motif consists of a 7/16 upbeat
leading to a target note that is complemented in various ways in the course
390
WTC II/6
of the fugue. The second half of E2 recalls the subjects head. E5 is a

complex variant of E2: before Bach quotes M1 toward the end of the
episode he builds from the subject head a motif in which six triplet 16thnotes lead toward an ascending third, then places this motif into a chain of
strettos and finally develops strettos of the entire inverted subject head.
Only E3, the shortest of the episodes, is without analogy, consisting solely
of M1 statements.
The role played by each of the four episodes in the dynamic development of the fugue follows directly from the analysis of the material. E1 and
E4 bridge two consecutive subject statements. E3, despite its short range,
effects a relaxation with independent material. The same is true for the first
half of E2 and for the M1-determined segment of E5. In both cases, the
episodes final half measure turns around and prepares for what appears
like (and is intended to appear like) an unexpected additional entry.
As expounded above, contour and rhythm in this composition allow for
two conclusions regarding the basic character. Given the specifics of the
material in this fugue, the ultimate choice will have only limited influence
on the articulation. In a lively character, the triplet 16th-notes are (quasi)
legato including the link to the A on beat 3, while the eighth-notes are non
legato. In rather calm character, all melodic notes should sound legato.
Different options are, however, open for the three final notes depending on
whether or not they are taken as cadential, in which case they had best be
played non legato. (In this option, the subject would appear as follows:
m. 11-3: legato, phrasing, from upper D to F: legato, after that: non legato).
The articulation in the counter-subject and in M1 is overall legato in
whichever character a performer chooses. Similarly, the articulation in the
broken-chord extension of mm. 19-20 is non legato in any case.
The tempo is limited on both sides by certain features: it must be fast
enough to ensure that quarter-notes, and not eighth-notes, are perceived as
the beats, and it must be slow enough to allow listeners to fully appreciate
the chromaticism and its harmonization as well as the syncopations and
complementary rhythmic patterns. The most convincing proportion between
prelude and fugue is one that translates the main pulse: a quarter-note in
the prelude corresponds with an eighth-note in the fugue. If, however, a
considerably faster tempo is desired for the fugue, a more complex proportion is preferable: one measure in the prelude then corresponds with half a
measure in the fugue. Approximate metronome settings: (a) prelude beats
= 108, fugue beats = 54 / eighth-notes = 108; (b) prelude beats = 108 /
measures = 36, fugue beats = 72 / half-notes = 36.
D minor
391
The fugue features two ornaments. The trill in m. 16 represents a long,

note-filling ornament since it resolves dulyin harmonic, melodic, and
metric respects. As it is approached stepwise, it begins on the main note,
which is prolonged to triplet-16th-note duration. The following four notes
are the upper/main/lower/main notes and lead smoothly into the E at m. 171.
The cadential trill in the fugues final measure, by contrast, is a point
darrt trill rather than a note-filling ornament since its resolution is
anticipated before the strong beat. After a regular beginning on the upper
neighbor note, the shake includes four or, preferably, six notes and then
stops short before the next eighth-note in the middle and lower voices.
The layout of the D-minor fugue is conveyed mainly by three features:
the introduction of the subject stretto in the second section, correspondences between subject entries (and, particularly, between episodes), and
the dynamic function of the episodes. The most prominent correspondence
occurs between mm. 10 and 25. In both cases, the head of the subject
appears in stretto with its inversion, at a distance of one quarter-note, thus
preparing the next subject entry. While the voices are inverted and the
subject statement begins earlier in m. 25 than in m. 10, the basic analogy is
unmistakable. Furthermore, mm. 10 and 25 both appear at the end of an
episode in which, after an overall relaxation of tension, they build up new
tension in a last-minute preparation for another subject statement. (This is
even more obvious in m. 10, which is reached in a cadential formula, than
in m. 25.) These two measures thus determine that the following entries are
perceived as part of the ongoing section. Another structural correspondence exists between the main body of E2 and the latter half of E5:
compare the four statements of M1 in mm. 7-9 with those in mm. 21-23.
E1 and E4 fulfill an analogous function as both consist exclusively of
primary material and both bridge consecutive subject entries. Finally, the
two subject statements that form the stretto in mm. 14-15 are identical,
both in notes and with regard to the voices presenting them, with the two
initial entries in mm. 1-3 and 3-5 respectively. The counter-subject of mm.
3-4 also recurs here in octave transposition.
As far as the thinning of the ensemble is concerned, no subject statement returns to two-part texture. The prominent lower-voice rests in m. 14,
however, convey this impression although the three-part texture is regained
halfway through the stretto with the entry of the counter-subject. The
fugue thus presents itself as consisting of two structurally corresponding
and equally long halves. The abridgment achieved in the second section by
means of the stretto is made up by the additional measures in the first half
of E5.
392
WTC II/6
While the harmonization is complex in its details, the overall design is

simple: all statements remain in the key of D minor. The dominant is not
reached in cadential confirmation and the subdominant appears only
transitorily at the end of the second stretto.
In its overall design, the fugue is built in one large dynamic curve: the
first half represents an overall increase and the second half an overall
decrease of tension. Within the first section, the tension rises, both owing
to the growing ensemble and the increase in polyphonic density. While E2
brings forth a certain decline, its final half measure with the intense
preparation of the subject entry makes up for the loss by propelling the
dynamic level up again. The second section is governed predominantly by
the opposite tendency. It begins with the subject in stretto, accompanied by
the counter-subjecta combination that represents the highest degree of
polyphonic complexity achieved in this fugue. The extension produces a
protracted diminuendo, followed by the more subdued character of the
inverted-entry stretto. E5 continues this tendency with a prevalence of
falling lines and descending sequences. The dynamic decrease ends at m.
251 where the preparation for the final subject statement achieves a last
short rise of tension.
WTC II/7 in E major Prelude

The most striking feature in this prelude is its rhythmic pattern. The
eighth-note is not only the predominant note value, it constitutes a constant
pulse suspended only briefly in mm. 67-68. In terms of texture, patterns in
complementary rhythm (e.g., mm. 1-2) alternate with homophonically
accompanied contours in strongly metric patterns (e.g., mm. 3 and 6-8),
parallel lines (e.g., mm. 13-16), and occasional pedal notes (e.g., mm. 5-8:
E, mm. 13-16: B ). Although a number of recurring melodic figures can
be identified, the prelude does not come across as dominated by thematic
material. The overwhelming impression is of the continuously flowing
pulsation in eighth-notes. The piece might therefore be described as
metrically determined, and belongs to the compositions expressing a
meditative mood.
The first harmonic progression is completed in m. 4, with a metrically
weak tail up to the seventh eighth-note in the lower voice. The subsequent modulation to the dominant concludes in m. 12 where it features the
same closing gesture. There are altogether seven sections:
I
mm. 1-47 tonic confirmed
II mm. 48-127 modulation to the dominant
III mm. 128-244 modulation to the tonic relative (C minor)
IV mm. 245-324 return to the dominant
V mm. 325-504 modulation to the dominant relative (G minor)
VI mm. 505-611 return to the tonic
VII mm. 611-71 tonic confirmed
Several structural correspondences meet the eye. Section I recurs at the
beginning of section VII, albeit with small variants (mm. 1-4 . 618-641).
Section II is sequenced in section III, again with variations and an enhanced
texture (mm. 4-81 . 128-161). Sections III+IV are partially quoted in sections
V+VI, with many small melodic modifications (mm. 21-287 . 47-547).
Section V contains a three-measure sequence involving both hands without
adjustments (mm. 325-354 . 355-384). Furthermore, there are two partial
sequences (mm. 385-404 and 405-424).
Questions of performance practice in this prelude are closely linked to
the meditative character. The pulse of continuous eighth-notes represents
timing in triple meter. From a metric viewpoint, the triplets within each
393
394
WTC II/07
of the beats appear as gentle embellishments that lack not only rhythmic
variety but also melodic intensity.
As a result, articulation is not an issue here. At any given moment, the
sound flows without interruption in at least one of the voices. As Bach
takes the trouble to write many of the triple-meter beats in the left-hand
part with rests (see mm. 3, 11, 20-23 etc.), we may assume he took it for
granted that all notes, whatever their value, would be sustained for the
entire duration. The performers concern is therefore merely to distinguish
between passages of true complementary motion, in which neither interruption nor overlapping may occur (see e.g. in mm. 245-324, 565-609), notes
to be sustained explicitly in order to transform broken chords into vertical
chords (mm. 18, 34, 37, 9-10, and 26-28), and measures in which a
subordinated rhythmic pattern needs to be observed (particularly those
with a pedal note: mm. 5-8, 13-16, etc.).
In the interest of the atmosphere the composer tries to convey, the
dynamic level throughout the prelude should remain even. Fine shades in
intensity create light and shadow or foreground and background, but strong
increases and decreases in tension are not called for. Details that may guide
performers in this delicate task can be discovered in answer to questions
like these: Does the melodic flow proceed from one eighth-note to another?
(In this case, each note carries active melodic weight.) Are subsequent
eighth-notes conceived as spread-out chords with additional rhythmic
fills or tail? (In this case, a larger number of notes may represent a
single musical unit, and may therefore have to appear under the impression
of a single gesture, with not more than one weighted note.) Do subsequent
notes represent two different melodic layers? (In such cases of hidden
two-part structure, different coloring of the parts is required.)
Several ornaments appear in this piece. The grace-notes all represent
eighth-note appoggiaturas resolving into quarter-note resolutions.1 The
inverted mordent in the final measure is straightforward while the one in
m. 50 raises second thoughts regarding the pitch of the lower neighbor
note. Despite the A in the key signature, this ornament requires an A as it
occurs in the harmonic context of G minor. The cadential mordent in m. 68
begins on the main note and contains three or, better still, five notes.
1
The question whether the dotted note should be split into one third appoggiatura and two
thirds resolution (as applies most often in J.S. Bach) or into two thirds appoggiatura and one
third resolution (as demanded by his son C.P.E. Bach in the famous treatise On the True Art
of Playing the Keyboard) does not arise here, for in all cases has the lower voice abandoned
the harmony on the third beat and a resolution at that moment would thus come too late.
E major
395
Phrase I is conceived in an ornamented chordal structure that unfolds

as a simple cadence. The only melodic feature recurring repeatedly is
the figure in the lower voice that appears like a female extension to
the downbeat chords. It will be referred to as [x] (L: mm. 1, 2, 4).
Phrase II is characterized by a motif whose main beats are ornamented,
either by a pre-beat inverted mordent (mm. 5, 6, 7 etc.: beats 1 and 2)
or by an escape note (see the third beats in the same measures). M1 is
accompanied by a rhythmic pedal note (mm. 5-81: tonic pedal). Three
non-motivic measures then focus on an F-major seventh chord that
triggers the modulation to the dominant. This harmonically significant
moment is underpinned by [x] in the lower voice. The phrase ends with
a cadential close analogous to that at the end of the first phrase (mm.
10-12 . 2-4).
Phrase III begins with M1 transposed to the dominant. Thereafter,
three measures suggest a return to the tonic (see the twofold [x] stressing the harmonic significance of mm. 17-18). The recurrence in mm.
19-20 of M1 in inverted voices confirms the V7 chord, but the fourmeasure cadential close turns instead to the relative minor key.
Phrase IV begins with almost perfect complementary motion. Figure
[x] appears in the right hand in the form of a V7 chord. Its resolution is
conceived as another figure in which tied notes form a chord. This
two-measure motif (M2) is then sequenced (mm. 268-287 . 248-267).
The phrase ends with a modulation to B major.
Phrase V begins with a new motif (M3). Its first half (mm. 325-334)
consists of the same seventh chord in both voices, preceded and followed by inverted-mordent figures. Its second half (mm. 335-354) complements this, after a repetition of the seventh chord, with a descent
featuring sequencing sevenths in both voices. After one complete and
two partial sequences, all in ascending order, the climax (m. 431)
spawns a four-measure extension. The phrase ends with a four-measure
cadential close similar to the one that concludes the third phrase.
Phrase VI continues the analogy by beginning with M2. It ends with
ascending sequences, in a pattern that is complementary except for the
dominant pedal on each downbeat (see mm. 56-60).
The final phrase begins similarly to the initial one. The four-measure
cadence is here extended and, after a short general pause (m. 67), the
return to the tonic is substantiated with a renewed perfect cadence in
E major (mm. 68-69) and an additional three-measure key confirmation.
396
WTC II/07
The following simplified version of the seven phrases aims at visualizing the main features:
cadence a
motif 1
cadence a
II
rhythmic pedal
motif 1
rh. pedal
- - - - - cadence b - - - - -
III
rhythmic pedal
motif 2
IV
- - - - cadence b - - - -
motif 3
4-bar
insert
motif 2
sequences
VI
dominant pedal
VII
opening cadence, expanded
E major
397
Dynamic shading in this prelude is very restrained so as not to jeopardize the meditative calm of the composition. At the outset of phrases II-VI,
the three motifs describe melodic curves of one, two, and three measures
respectively, while the endings of these phrases and the framing cadences
are defined by very slight harmonically induced tension developments.
WTC II/7 in E major Fugue

The E-major subject appears firmly rooted at the outset but ambiguous
at its end. Beginning with the keynote on the downbeat, it spans a little
more than five or six measures. As the return to the tonic harmony occurs
already on the G at m. 61, the question whether or not mm. 6-71 belong to
the subject has led to many heated discussions. On the one hand, both the
structural design of the phrase and Bachs very consistent use of the entire
length support the view that what is harmonically but a tail should be
regarded as an integral component of the subject. On the other hand,
Bachs harmonization of this tail does not always remain in the tonic but
often leads away from it (see, e.g., mm. 12-13, 19-20, 26-27, etc.). This is
a strong argument for the shorter version. In the following analysis, we
will try to keep both aspects in mind: the melodically consistent longer
version [a] (which comprises a complete sequence) as well as version [b],
one measure shorter but harmonically more reliable.
With regard to its pitch pattern, version [a] describes a curve that sets
out from and returns to the E-B interval, enclosing several ups and downs
in-between. Skips and steps seem fairly equally distributed in this phrase.
Apart from the initial fifth, which swings back in steps, there is a perfect
fourth (m. 3) followed by seconds that lead to a broken chord (mm. 4-5).
The combination of fourth + seconds + broken chord is then sequenced. The
rhythmic pattern displays great variety, with four different note values
featured in the subject itself: a whole-note, several half-notes (two of them
as syncopations), quarter-notes, and eighth-notes. Later in the fugue Bach
adds dotted quarter-notes, 16th-notes, and various tied values.
The subjects harmonic background is interesting as it comprises two
complete cadential progressions. Bach harmonizes the initial whole-note
with an active step from the tonic to the subdominant (see, e.g., m. 14) or
to the secondary dominant (see, e.g., m. 60), followed in m. 2 by the
dominant (or dominant-seventh) and at m. 31 by the return to the tonic. From
there onward, the syncopated C initiates another subdominant harmony.
398
WTC II/07
This gives way, on or before the syncopated B (see mm. 64 and 18

respectively), to the dominant, and then returns to the tonic on the
downbeat of the statements sixth measure.
(
I IV V7
I IV- - ii7
IV V V7 I
The phrase structure behind all these details encompasses three subphrases. The first coincides with the first cadential progression, thus
closing at m. 31. The second subphrase begins with the syncopated C and
closes before the sequencing syncopated B, i.e., at m. 51, and the third
subphrase consists of the complete or shortened sequence of the second. In
the light of this phrase structure, our earlier assessment of the interval
structure calls for a small revision. The perfect fourths are not actually
conceived as intervals (in the sense of steps within a melodic context) but
mark cuts between adjacent subphrases. The pitch pattern thus balances the
initial fifth and the broken chord at the end of the second and third
subphrases with stepwise motion in-between.
Determining the overall climax is not difficult. The C in m. 3 not only
stands out rhythmically and melodically, it also represents the definite turn
to the subdominant harmony. Furthermore, the gradual relaxation in the
second subphrase is echoed in the third subphrase, with a smaller climax
on the syncopated B and a more complete relaxation to the tonic. In the
first subphrase, the initial note should be felt as an implied increase of
tension (which, obviously, the keyboard is unable to produce but any
string or wind player would provide very convincingly). The subphrase
then peaks fairly gently on the B in m. 2 before relaxing toward its end.
The fugue comprises twelve subject statements. Among them, one
(marked with an asterisk in the table below) is significantly changed at
both the beginning and the end. All other subject statements are given here
in their full melodic length, including the contested tail.
1 mm. 1-7
B
7
mm. 37-43
A
2 mm. 7-13
T
8
mm. 38-44
S
3 mm. 14-20
A
9
mm. 53-59
T
4 mm. 21-27
S
10
mm. 55-59
A*
5 mm. 30-36
T
11
mm. 59-65
S
6 mm. 31-37
B
12
mm. 60-66
B
E major
399
Apart from the adjustment of the initial interval in the answer and the
overall modification in the tenth entry, the subject remains basically
unchanged throughout the fugue. Slight changes in pitch appear in mm. 31,
38, and 60 where the second interval in the tenor, alto, and soprano entries
respectively is altered to a note repetition instead of the original descending
second. The only rhythmic variation is inconsequential. It occurs in the
statements beginning in mm. 30, 31, and 53 where the initial whole-note is
shortened to a half-note.
Inversion and augmentation are not used, but strettos appear frequently.
Their structural position and the kinds of combination chosen are interesting. The four stretto settings in this fugue include pairing of the two
lower voices (T+B, mm. 30-37), the two upper voices (A+S, mm. 37-44),
the two inner voices (T+A, mm. 53-59) and the two outer voices (S+B,
mm. 59-66). In each of these pairs one entry is modified. An entry with an
interval adjustment or any other significant variation carries less weight
than one that sounds exactly like the original. The sequence of the leading
voices of the four sets is thus B S T B (from TB, AS, TA, SB).
Proceeding in the traditional way, one can expect a counter-subject
surfacing against the second subject statement (i.e., against the answer).
Within the melodic progression introduced in the bass line in mm. 7-13,
two segments can be distinguished. Their different recurrence in the course
of the fugue will tell whether or not any of the segments, each one of them,
or only the whole qualifies as a self-assured companion to the subject.
The first segment begins with a descending sequence of the subjects
tail (mm. 7-81), which it extends into a cadential close (up to m. 91). This
gesture recurs, with small variations at the end, three timesalbeit never
against the subject. Instead, it appears as an extension of subject statements
(see S: mm. 27-30, A: mm. 43-45, S: mm. 44-46, B: mm. 66-68; shortened
also in T: mm. 13-14). We must thus conclude that this figure is a horizontal extension of the subjects tail rather than a vertical counterpart in the
manner of a counter-subject. The second segment introduced against the
answer (mm. 92-121) begins and ends with the keynote (here, on the
dominant, this is B ). After an initial syncopation, the curve between the
two Bs resembles a familiar closing formula, particularly from the dotted
note in m. 11 onward. This melodic unit recurs twice, against the third and
the fourth subject statements (see T: mm. 16-19 and A: mm. 23-26; in the
latter, see also the parallel in the tenor). We can safely thus speak of a
counter-subject. The fact that it does not reappear later may be excused by
the grouping of all further subject statements in strettos.
400
WTC II/07
S
CS
The number of subject-free passages in this fugue depends on the interpretation of the subjects length, as the assessment of the subjects tail
makes all the difference. The table below is based on the following
distinction: Whenever a subject statement reaches its harmonic conclusion
only one measure before the beginning of the next statementif, in other
words, the subjects tail ends on the initial note of the following entrythe
tail and its vertical surroundings are not considered a subject-free measure,
although it must be recognized that it is harmonically defensible to hold
this view. Wherever the tail is accompanied by a motif that plays a role
in the further development of the episode, the measure is considered part of
the episode. This leaves five genuine episodes:
E1 mm. 12-13
E4 mm. 43-53
E2 mm. 19-20
E5 mm. 65-70
E3 mm. 26-30
Four of these five episodes grow out of the end of a preceding subject
statement by extending the tail in sequences: In E1, the tail and its
sequence are accompanied by a four-note motif in the bass (B-B-A-G)
that is also sequenced a third lower. In E3, the same bass motif appears,
albeit with a smaller initial interval, against the soprano entrys tail and its
sequence (see F-B-A-G in mm. 26-27, sequenced again a third lower).
After this variation of the first episode, E3 concludes with a cadential
formula in B major. In E2, the subjects tail and its descending sequence
are accompanied by partial parallels in the tenor and (descending) cadential steps in the bass. In E5, the descending sequence of the subjects tail is
matched again by descending sequences in all other voices (mm. 65-66,
66-67). The conclusion follows here with an extended cadence.
The only longer episode is E4. Its bass motif, though not immediately
related to the earlier ones, is one measure long and sequenced in descending
thirds, once in identical and once in varied shape (mm. 43-46). Thereafter
the two voices of the preceding stretto introduce a counterpoint of two
substantial motifs (M1: alto m. 46-48 G to B, M2: soprano m. 46-48 E to
E). Exchanged in cross imitation these form a threefold chain. The overall
motion is falling, accompanied by a two-measure bass figure.
E major
401
The role played by these episodes in the development of the fugue can
easily be deduced from the observations made above. E1 and E2 link
consecutive entries with a slight decline in tension each time. E3 and E5
conclude harmonic as well as structural progressions with full cadences.
E4 also creates a gradual decrease of tension but does not spell out an
explicit close, thus maintaining some connection with what follows.
The fugues basic character is determined by the variety in its rhythmic
pattern that, particularly with its manifold syncopations, needs much calm
to unfold. Meanwhile, the alla breve time signature prevents too slow a
motion. The resulting calm but flowing style is also reflected in the interval pattern, which demands an alternation of tight legato in all steps and
single skips and non legato for the characteristic broken chords in the tail
of the second and third subphrases of the subject.
As the contemplative mood of the prelude with its triplet motion differs
considerably from the duple meter in the fugue, the tempo proportion
between the two pieces can be simple: a beat (= a dotted quarter-note) in
the prelude becomes a beat (= a half-note) in the fugue. Approximate
metronome settings: 72 for the dotted quarter-notes in the prelude and the
half-notes in the fugue.
The subject with its distinct material pervades much of the fugue. Its
articulation should therefore be carefully studied. In the first subphrase,
E-B and A-G are legato. The interruption by the tension-sustaining rest
should be very smooth, giving the impression that the fully sustained
half-note reaches silently across the rest where the A picks up in almost
equal intensity. (A drop in tension after the resta mistake that occurs all
too easilycompletely destroys the expressive significance of this rest.)
Thereafter, GC and FB two measures later are separated by phrasing.
The descents C-B-A and B-A-G are legato, followed by the gentlest of
breaks for the note repetitions. The inverted-mordent figures resume the
legato, before a softly detaching non legato distinguishes the broken
chords A-C-F and G-E-E.
The obvious grouping of subject statements into three blocks makes it
seemingly easy to determine the scope of the sections, but only the first is
simple. It contains the initial statements in each of the four voices, connected by the short episodes E1 and E2 and concluded with a perfect cadence
on the dominant at m. 301. Thereafter, the introduction of strettos marks a
new section. The first stretto leads without interruption into the second,
which is followed by E4, the episode that stands out both for the significance of its motivic material and for its length. Two more strettos, equally
tightly linked, are rounded off by the final episode that closes the fugue.
402
WTC II/07
The sequence of two strettos + episode, two strettos + episode may lead
one to assume a division of these measures into two sections. The fact that
each stretto-pair features a complete round of voices would support this
view were it not for the lack of a harmonic close between the two blocks.
In m. 53 where the tenor sets in, the other voices are in the midst of their
motifs. Moreover, the B-major harmony is inflected by the D on the middle
beat and grows into an E7 chord on the next downbeat. The expected
resolution into the subdominant A major materializes only in a metrically
weak position after the second quarter-note in m. 54.
Another approach to this fugues structure is therefore to assume a long
second section, which develops from the first through overall expansion:
I = entries (B, T), E1 (link), entry (A)/E2 (link)/entry (S), E3 (close)
. II = strettos (TB, AS), E4 (link), strettos (TA, SB),
E5 (close).
In the first section, the tension rises very gradually with the increasing
number and rising order of entering voices. As both the subjects tail and
the links of E1 and E2 with their descending sequences and retroactive
harmonic steps have an appeasing quality, the buildup does not reach a
high level, and the cadential close in mm. 29-30 concludes almost as softly
as the piece began. The dynamic processes in the fugues second section
repeat those of the first section on a heightened level of intensity. The rise
of tension between the first and the second strettos is stronger because of
the increase in voices and the more drastic rise in pitch level. E4 with its
descending sequences causes a depletion of tension, but it is counteracted
by tight polyphonic interplay. The third stretto enters with slightly reduced
intensity, a fact that is supported by the texture (resting soprano) and the
choice of subject-carrying voices (T+A form the least exposed combination).
The final stretto brings the climax, with a brilliant combination of S+B
subject statements and a parallel of the two inner voices.
WTC II/8 in D minor Prelude

This prelude is written in consistent two-part texture, with imitation
occurring regularly. Techniques of counterpoint are also widely used, and
there is not one moment where the feeling of polyphonic independence of
the voices is lost. At a casual glance one might consider having to decide
between a two-part fugue and an invention. The fact that the first measure
features bass notes that serve as a harmonic support but do not take part in
the polyphonic play is shared by the various invention-style preludes. But
even the other prelude of the Well-Tempered Clavier that is composed as
a two-part fugue, the prelude in F minor vol. I/14, also introduces its
subject against a neutral accompaniment in the lower voice. Nonetheless,
an interpretation as a fugue encounters some more serious irregularities:
The imitation of the main idea occurs in tonic position and not, as usual in
a fugue, on the dominant. The main melodic idea only covers thirteen of
the thirty-six measures, but its various forms of development account for
much more. Moreover, there are repeat signs for the first 16 as well as the
following 20 measures. It may therefore be less controversial to describe
this prelude as a two-part invention.
The sections in this invention are defined, in addition to the usual
harmonic processes, by the recurrence of the thematic material. This first
overview gives only the harmonic processes.
I
mm. 1- 91 tonic to tonic relative (D minor to F major)
II
mm. 9-16 on to the dominant (F major to A minor)
III
mm. 17-281 back to the tonic (A minor to D minor)
IV
mm. 28-36 tonic confirmed (D minor)
The number of structurally analogous measures is striking:
mm. 1-2 . 28-29 (same key, counterpoint varied),
mm. U: 3-4 / L: 33-53 . L: 21-22 / U: 223-233 (transposed),
mm. 6-8 . 25-27 (transposed),
mm. 154-16 . 354-36 (transposed),
mm. 17-19 . 323-343 (transposed, voices inverted)
To sum up the findings with regard to harmonic design and structural
analogies, we can recognize that this invention conceals a typical Baroque
binary form, found, e.g., in many of Bachs suites, most particularly in
allemandes.
403
404
WTC II/08
The rhythmic pattern is generally simple. Most measures feature a

combination of eighth- and 16th-notes. The pitch pattern encompasses
frequent leaps both in the eighth-notes (see, e.g., L: mm. 6-8) and the 16thnotes (see, e.g., U: mm. 6-8). This simplicity is, however, counteracted not
only by a few 32nd-notes but also by a very large number of ornaments.
The basic character of this allemande-style invention is thus rather lively,
but the tempo should be moderate rather than fast. The articulation must
contrast a fairly gentle non legato in the eighth- and quarter-notes with
legato style in the 16th-notes. Among the eighth-notes one can distinguish
melodically intensive touch (as in U: m. 2) from a more neutral touch used
in accompanying groups (as in L: mm. 6-8). The only occasions where
eighth-notes are legato because they are paired as appoggiatura/resolution
are indicated by the composer (see mm. 16 and 36). In the tied notes (mm.
4-5, 5-6, etc.), the 16th-note extension is treated as a 16th-note. Wherever
abbreviation of this note due to phrasing is desired, it must be guaranteed
that the syncopated effect is not destroyed. Expressing phrasing by
dynamic means alone may be a better choice here.
Among the numerous ornaments, the turn in m. 2 with four 32nd-notes
and the inverted mordents pose no problem as none of them needs an
accidental for an artificial leading note. (Their lower neighbor notes are
regular: F in m. 3, B and E in m. 4, A and G in m. 5; G in m. 21, C
and F in m. 22, B in m. 23.) Trills and mordents come with and without
suffixes and also begin differently. In mm. 2 and 9, the mordents in the
upper voice begin on the main note since they are approached stepwise and
comprise a single three-note shake. The trill in m. 14, by contrast, begins
on the upper neighbor note A, shakes in three pairs, and ends with the
indicated suffix.
The main melodic idea upon which this invention rests, M1, is a little
over one measure long. It begins and ends on a downbeat. Built exclusively of 16th-notes, it describes a curved line in which a zigzagging ascent to
the fifth in the first half is answered by a descending scale in the second
half. Surprising in this line are the pitches at the outset of the descent: they
represent the melodic minor scale, which is more common in ascents. The
dynamic shaping follows the straightforward curve. Thus the climax falls
either on the octave D (if the tempo is taken slow enough to feel an
offbeat 16th-note) or on the fifth A (the rhythmically and metrically more
convincing choice, particularly at a swifter pace).
There are altogether thirteen statements of M1. Most appear in pairs at
the octave. The following table gives the keys only of the statements thus
paired, with capital letters for major, lower-case letters for minor keys.
D minor
1.
2.
3.
4.
5.
6.
405
mm. 1-2
mm. 2-3
mm. 9-10
mm. 10-11
mm. 11-12
mm. 12-13
U (d)
L (d)
L (F)
U (F)
U (a)
L (a)
7. mm. 17-18
L
8. mm. 19-20
U
9. mm. 28-29
L (d)
10. mm. 29-30
U (d)
11. mm. 323-333
U
12. mm. 333-343
L
13. mm. 343-353
U
M1 undergoes a number of small changes in the course of the piece. In
mm. 181, 201, 333, and 343, the descent in its second half begins not from
the octave but from the seventh, with the effect that it reaches the keynote
on the final 16th-note of the measure. Consequently, the interval to the
ensuing downbeat is modified in all cases. In mm. 18 and 20 it is raised to
a fourth, whereas in mm. 33 and 34 the transition to the counter-motif is
such that the strong-beat note is not perceived as belonging to the motif.
Bach creates two companions for M1. In keeping with the terminology
used in polyphonic compositions they will be referred to as countermotifs. CM1, introduced in mm. 2-3, plays around the keynote and fifth.
It begins with three eighth-notes (the leading-note followed by the keynote
and its octave), centers in an ornamented descent, and concludes, after a
rest, with another leading-note/keynote pair. The single climax of this
counter-motif falls on the middle beat and thus coincides (almost or
exactly) with that of the main motif. CM1 recurs several times, albeit
heavily disguised. In m. 9, the 16th-note figure begins with F-E (instead
of E-F) and ends on the octave. The ornamented descent is the same, but
the resolving final pair is omitted. In m. 12, the initial eighth-notes octave
leap is diminished to a sixth. The descent, twice as fast as before, occurs an
octave lower than expected and is followed by another leap, and the ending
(G-A) appears faster and a little earlier than it should, with unexpected
notes following. Finally in mm. 28 and 29, only the descending figure,
now even on a different degree of the scale, can be identified.
The second counter-motif (CM2) is introduced in the second half of the
prelude (U: mm. 17-18). It is characterized by a symmetrical structure in
which the second half acts as an inversion of the first. There are two broken
seventh chords in the same rhythmic pattern on beats 1 and 3 respectively
and two turn-figures on beats 2-3 and 4-1. Regarding its dynamic design,
the rhythmic details, surprising in a composition with an otherwise so
simple rhythmic pattern, demand that the climaxes fall on the two syncopations. This counter-motif thus exhibits a greater polyphonic independence
than the first. It recurs four times, in L: mm. 19-20, L: mm. 323-333, U:
mm. 333-343 and, considerably modified, in L: mm. 343-353.
406
WTC II/08
The invention features two other motifs besides M1. Naming them
creates a little problem as the first obviously derives from the main motif
and the second, though not related to the main motif, derives very
definitely from the previous one. To avoid confusion, they will appear here
with consecutive numbering regardless of their relationships. The initial
two statements of M1 are followed by M2 (see U: mm. 3-4). It begins on
the second eighth-note of the measure with a fragment of the zigzag known
from M1 and continues with a short descent also reminiscent of the main
motif. The two segments are then complemented by an eighth-note leaping
up a fourth to an ornamented peak note that is tied over. This pattern distinguishes this motif from the preceding main material. M2 is imitated in
stretto (L: mm. 33-43). The sequence of both the original and its imitation
(U: mm. 4-5, L: mm. 43-53) features a strongly varied first half and thus
severs all ties with M1. A partial sequence of the fourth leap follows (L:
second half of m. 5) and can be recognized, without its ornament, at the
beginning of the accompaniment figures in the three following measures.
With these broken chords in the lower voice Bach creates a gradual
descent that contributes to the shaping of this portion of the prelude:
E-G-C (m. 4), D-F-B, C-E-A (m. 5), B-D-G (m. 6), A-C-F (m. 7),
G-B-E (m. 8), and F (m. 9). M2 recurs, complete with imitation and
varied sequence but in inverted voices, in mm. 21-23. In addition, its first
half, the segment that is related to M1, reappears three times in mm. 13-14
in almost complete parallel motion and a further four timestwice with
neutral accompaniment, twice with a disguised parallelin mm. 30-32.
M3 also comes in two versions: as a trunk, and with an extension. The
trunk begins like M1 and M2 after the strong beat and ends on the next
strong beat (U: m. 5 E-E). Among the eight regular 16th-notes, the initial
four and the last are melodic, describing a turn-figure and, after interrupting escape notes, its resolution. (The melodic idea is thus conceived as
E-D-C-DE.) In this format M3 is sequenced (see mm. 5-61: D-D).
In three further sequences, the non-melodic inner segment is extended and
thus lengthens the motif to full-measure scope (M3a see mm. 6-71: C-C,
7-81: B-B, 8-91: A-A). The dynamic shaping of this motif takes the play
with melodic and non-melodic components into account. The first four
16th-notes sound in a dense legato of considerable intensity, the brokenchord insertion is lighter both in touch (quasi legato) and in dynamics
(diminuendo), and the final note, while soft at the end of the decrease,
picks up the more intense tone color of the melodic notes.
The prelude comprises four sections. The first introduces all motifs
with the upper voice in the lead (mm. 1-91). After the initial M1 statements,
D minor
407
the remainder of the first section is dominated by descending peak-note

lines in both parts creating a protracted decrease in tension (see U: mm.
3-91 the final notes of each motif, G-F-E-D-C-B-A; L: mm. 4-8 the
final notes of M2 and its partial sequences, C-B-A-G-F-E).
Section II matches a twofold set of M1 statements with partial quotations of M2 and a free cadential close. As the second pair of M1 statements
inverts the voices and thus creates the effect of an ascending sequence, the
beginning of this section expresses a tension increase. The M2 sequences
do not respond with the expected relaxation since Bach transposes the last
segment upward (U: m. 14). The increasing tendency is then continued
through the peak notes in the upper voice (mm. 14-15: F-G-A) and only
released from m. 153 onward. The rhythmic density of the cadential close
thwarts true relaxation.
Section III combines all motifs from section I as well as several
features from section II in a new way. Owing to the heightened rhythmic
density of CM2 and its dynamic independence from M1, the beginning of
this section represents the greatest tension so far in this prelude. The alternation of this intense combination with the dynamically fairly weak
expanded M3 creates strong contrasts from one measure to another.
Section IV presents yet another combination of thematic material. This
juxtaposition, while not as dramatic as the previous one, admits even less
relaxation than any of the preceding sections. The development from the
polyphonically less intense M1/CM1 combination to the more active
M1/CM2 match alone would call for a buildup. In addition, the three final
M1 statements are conceived with growing harmonic tension. Even the M2
sequences, which convey a decreasing tendency, are here set in a pattern of
increasing textural density (mm. 30-31). Only the figurative close, which
picks up immediately from the main motif, brings forth a degree of relaxation and thus concludes the prelude on a moderately soft note.
WTC II/8 in D minor Fugue

The scope of this subject does not pose a problem: while there areas
Bach demonstrates masterfullya host of possible harmonizations
throughout the phrase, the return to the tonic occurs unfalteringly on the
downbeat of the third measure. This gives the subject a very balanced
length, with the final note complementing the initial rest to add up to two
full measures.
408
WTC II/08
The pitch pattern also displays a clear direction. Beginning on the

keynote, which is reinforced by a short deviation to the leading note, the
subject features three ascending motions, the step D-E in m. 1 and the
fourths D-G and E-A in m. 2, before ending in a downward motion
(mm. 2-3: A-G-F). The rhythmic pattern includes eighth- and 16th-notes
as well as two dotted quarter-notes, both placed as syncopations. This
rhythmic structure has a bearing on the perception of the melodic shape.
As the second of the fourths is metrically unaccented while the preceding
rising motions are reinforced by the two syncopations, we may hear the
following line underlying the subject: D-E-GF.
The harmonization of the subject shows great variety, as the entirely
different solutions in mm. 7-8, 21-22, and 40-42 demonstrate. The initial
note repetitions can represent the tonic or the subdominant, the leading
note C may stand for the dominant or the tonic relative, the syncopations
for either dominant-ninth + subdominant or dominant relative + dominant,
and so forth. Ample use of chromaticism owing to the secondary materials
artificial leading notes adds to the versatility and beauty of this fugue. This
can be seen particularly well in mm. 21-22, where bass and tenor use ten of
the twelve semitones. Unchallenged in all entries is only the return to the
tonic on the final notenever a fragment earlier. This resolution, as well
as the curve underlying the subjects
contour, defines the
phrase as an indivisible whole. The
harmonic setting in
the example stems
from the analysis
offered in the study
by L. Czaczkes.
The dynamic outline follows the prominent features. The tension grows
through the leading note and the first syncopation all the way to the second
syncopation, which is propelled by the fourth leap. Thereafter, it subsides
throughout the final half measure. There are 16 subject statements.
1 mm. 1-3 A
6 mm. 17-19 A 11 mm. 27-29 S
2 mm. 3-5 T
7 mm. 19-21 T
12 mm. 30-32 A
3 mm. 7-9 B
8 mm. 21-23 S
13 mm. 32-34 T
4 mm. 9-11 S
9 mm. 23-25 A 14 mm. 40-42 B
5 mm. 15-17 B
10 mm. 25-27 B 15 mm. 43-45 S
16 mm. 43-45 T
D minor
409
The subject receives a real answer without any interval adjustment.

The modification of the first step from a minor second to a minor third,
i.e., the kind of change that is expected in the answer, occurs only later in
the fugue in three consecutive entries (mm. 15, 17, 19). Another entry
enlarges the first step to a whole tone while lowering the step above the
keynote (m. 32: C and E). Two entries feature a Picardy-third modification at the end: after a statement in the minor mode, the final note is
raised to the major third (see mm. 27 and 29).
The fugue features one stretto, one parallel, and one inversion. The
overlapping occurs in mm. 23-25 where the entry in the alto is followed
first by a fake entry in the tenor (see m. 24: E E E D E F) and then by a
complete entry in the bass. The inverted subject statement is also the one
that appears in parallel to another entry (see mm. 43-45: soprano with
subject in original shape, tenor with subject in inversion).
Bach has given this subject one companion, but only for the first half
of the fugue. It is introduced in mm. 3-5 against the answer. Characterized
by its ascending three-note figure, it seems like an ornamentation of a
simple scale (mm. 3-5, alto: E-DEFGA). While it shares many of
the subjects traits
(compare, e.g., the
ascending threenote groups with the
rising fourths), it differs significantly in
that it presents only
one gesture: a rise in
pitch and tension,
contrasting with the curved shape in the subject. CS1 recurs, with slight
modifications, in mm. 7-9, 9-11, 15-17, 19-21, and 21-23.
There are eight subject-free passages.
E1 mm. 5-7
E5
mm. 29-30
E2 mm. 11-15
E6
mm. 34-40
E3 m. 23 (eighth-notes 2-5)
E7
mm. 42-43
E4 m. 27 (eighth-notes 2-5)
E8
mm. 45-46
The episodes establish two distinct motifs. M1 is introduced in E1
where it dominates the uppermost voice (from B in m. 5 to F in m. 7). It
recurs in E2 (T. mm. 11-13 and A. mm. 13-15) and in E6 (T: mm. 35-37)
410
WTC II/08
where its second half is sequenced. A four-note figure that accompanies the
motifs first appearance (m. 5: B-A-A-D) also recurs (A: mm. 37-38).
M2 appears in a preliminary shape already in E4 (T: m. 27, from B to C)
but gains its full size only in E6 where it is followed by a sequence (S:
mm. 36-37). Its extension features a descending 16th-note group (S: mm.
38-39) that is imitated and sequenced throughout the remainder of E6.
Four of the subject-free passages suggest a harmonic conclusion, but in
only one of them does this conclusion coincide with the end of the episode:
in E8 the cadence concludes not only the episode but the entire piece. In
E3, the soprano sounds a traditional do-si-do figure, but this formula with
its cadential return to D major at m. 241 overlaps with the subject entry in
the alto. Similarly, the F-major cadential formula in mm. 29-30 ends on
the fourth eighth-note of the next subject entry. In E6, by contrast, the
cadential close occurs after the first 12 measures, in m. 353, while the
episode continues for another five measures, dividing this episode into two
segments (E6a and E6b).
The role each episode plays in the dynamic design is determined by its
motivic and cadential content. Thus E1, E2, and E6b all create a contrasting
color that is used as a kind of negative preparation for the following subject
statement. E3, E5, E6a, E7, and E8 wrap up a preceding entry. They bring
forth a slight diminuendo, but not with a change of color.
The rhythmic variety, together with the high content of leading notes
and other alterations, indicates a rather calm basic character. Yet as the
eighth-notes and longer note values carry the melodic line while the 16thnotes serve as embellishments, the pace should not be too slow. The tempo
proportion between prelude and fugue must be complex: Given the identical time signature in both pieces, a simple translation of the pulse would
give a dull result. One good way of relating the two pieces is to use an
imagined eighth-note triplet (instead of the actual four 16th-notes in each
beat) for the pivot: a triplet eighth-note in the prelude corresponds with a
16th-note in the fugue. (Approximate metronome settings: prelude beat =
80, fugue beat = 60.)1
The articulation in the fugue is predominantly legato. Detached playing
is only required in connection with the leaps in the episode motifs (in M1
after the D and Fin A: m. 6, before and after the higher A in T: m. 5,
1
The translation is achieved by mentally continuing the preludes pulse while dividing
each quarter-note into a triplet. These triplet-beats are then regrouped in pairs and, as soon
as they are perceived as independent pulses, each pair is converted into one of the opening
eighth-notes of the fugue.
D minor
411
and throughout the broken chord in M2) as well as in the cadential-bass

patterns (mm. 30, 35, 43, and 46).
There are four sections in this fugue. The first comprises the four initial
subject statements bridged, between two entry pairs, by the contrasting E1.
The section concludes at m. 111 on the minor dominant (A minor, the alto
with a resolution of the suspension). In the subsequent measures, the
reduced ensemblethe alto rests in mm. 12-17, the soprano in mm. 17-21
indicates the beginning of a new section. The second section begins with
an episode (E2) followed by four consecutive subject statements. The
cadential extension of E3 concludes it at m. 241 on the tonic with Picardy
third. Thereafter the ensemble is once again reduced to three voices. The
third section begins with a stretto whereby the entry in the alto, setting out
earlier, overlaps with the preceding cadence. Because of this irregularity,
the bass entry has more weight with regard to the structure and counts as
the group leader in this stretto. Short, half-measure episodes build bridges
between this stretto and the next soprano statement as well as between it
and the following alto statement, while the fourth entry, in the tenor, follows without interruption. The cadential close provided by E6a concludes
this section on the middle beat of m. 35 in A minor, the minor dominant.
The ensuing episode returns once more to three-part texture. The fourth
section begins again with an episode (E6b) before launching its bass entry.
As a first subject statement in a new section this bass entry avoids the full
four-part texture. Bach does not achieve this by reducing the ensemble but
by accompanying the statement with metrically regular chords and thus
changing the texture to near homophony. The bass entry is followed, after
the short E7, by the parallel statements in soprano and inverted tenor. The
final episode E8 concludes the fugue in D major.
It may already have become obvious from this description that there is
a remarkable symmetry in the layout of this fugue:
Sections I + II encompass 23 measures, as do sections III + IV.
Sections I and III conclude in A minor, while
sections II and IV end in the tonic with the Picardy third.
Sections II and IV both begin with a longer episode, while
sections II and IV both end with similar cadential extensions
(E3 and E8 respectively).
An essential difference between the fugues two halves additionally
underpins the binary effect: almost all entries in sections I + II are supported by the counter-subject, whereas sections III + IV do not quote it
even once.
412
WTC II/08
Bach designs the fugues first half as a graded tension increase. The
first two subject statements are contrasted with one, the following two with
two quotations of the episode motif. The ensuing four subject entries in
uninterrupted succession heighten the tension further before it is rounded
off by a concise cadential formula contributing the expected relaxation.
The second half begins with an overlap and a stretto, thus immediately
propelling the tension upward. Three further subject entries in four-part
setting, interspersed only by half-measure episodes, retain the tension at a
high level that is only abandoned at the cadential close of this section. The
episode opening section IV provides the longest span of contrasting color
in the fugue. The section ends, after the cadential return to the home key,
with the climax in the parallel subject statements.
WTC II/9 in E major Prelude

The E-major prelude is composed in a texture that moves consistently
in three parts, apart from voice splitting in the final measures of each half.
Although closer inspection reveals that the lower voice has a less active
role in contributing thematic material than the upper and middle voices, the
setting is undoubtedly polyphonic. The material does not center on any one
melodic idea but presents a chain of motifs. All motifs are either sequenced
or imitated right away, but only a few also recur later in the piece. This
thematic design, together with the structural layout in two repeated halves,
is reminiscent of movements in a Baroque suite. Given the prevalent rhythmic features and the triple meter, one thinks of a courante.
The distinct motifs are interspersed with cadential formulas that,
although not always typical in their melodic details, stand out from their
surroundings because of the noticeable absence of motivic material. It is
these cadential closes, rather than the ones concluding a four-measure
motif, that shape the harmonic layout of the piece. One can distinguish five
sections:
I mm. 1-181
tonic to dominant (E major to B major)
II mm. 18-24
dominant confirmed (B major throughout)
III mm. 25-321 dominant to tonic relative (B major to C minor)
IV mm. 32-461 return to the tonic (C minor to E major)
V mm. 46-54
tonic confirmed (E major throughout)
Among other structurally relevant features, symmetries and pedal notes
deserve attention. Correspondences can be found three times:
mm. 1-51 . mm. 5-91 . mm. 25-291 (varied),
mm. 18-211 . mm. 46-491, and
mm. 21-241 . mm. 43-461.
Pedal notes occur, directly or indirectly, in 19 of the preludes 54 measures:
mm. 1-3: E
mm. 18-20: B
mm. 46-48: E
mm. 5-7: B
mm. 28-30: G
mm. 51-54: E
The rhythmic pattern is simple. Melodically relevant notes are either
16th- or eighth-notes. Longer notes occur only in accompanying voices
(see particularly mm. 1-3 and 5-7 etc.) or in complementary patterns (see,
e.g., the syncopations in mm. 18-20 and 23-24). The pitch pattern features
ornamental figures in the 16th-notes (see particularly the frequently recurring turn) and broken-chord figures (L: mm. 3 and 7, U: mm. 9 and 10,
413
414
WTC II/09
etc.). The eighth-notes move mainly in leaps, in broken chords, or in

octaves. The basic character can thus be identified as rather lively. The
tempo is paced in a swift triple-meter pulse. Ornamental 16th-note figures
should be perceived as groups projecting a single gesture. The articulation
requires legato for the 16th-notes and non legato for most of the eighthnotes. Exceptions occur only where one of the voices in the three-part
texture splits to form a complementary rhythm and the detachment thus
separates the notes of each layer but not the succession itself (see, e.g.,
mm. 23-24).
Three ornament symbols appear in the score. Two of them indicate
mordents (L: mm. 21 and 43). As both are preceded by an appoggiatura,
they begin on the main note and contain only a single three-note shake.
The trill in m. 26 also begins on the main note with a 16th-note, followed
by seven 32nd-note pairs, the last of which is the suffix A-B.
In terms of the material characterizing this prelude, the first eight
measures are dominated by the theme and its answer on the dominant. The
theme consists of three and a half measures. Harmonically, a tonic pedal is
followed by a short brushing of the dominant chord (m. 33) before the tonic
is reinstated. After a 4/16 link establishing the dominant as a new center of
reference, the theme recurs a fifth higher, with its upper and middle voices
inverted.1 The most prominent features are the eighth-note upbeat to a
weak beat (see, e.g., m. 1: U upbeat to beat 2, M upbeat to beat 3) and the
turn figure. With regard to texture, the theme involves two voices in free
imitation and a pedal that the third voice sustains and to which it returns
after a short-lived attempt to take part in the polyphonic play. At the
beginning of the preludes second half, the theme can be heard in the
original order of voices but with slight modifications. The lower-voice
notes B-C in mm. 26-27 abandon the pedal-note effect without disturbing
the main outline of this voice, while U: mm. 26-27 and M: m. 28, setting out
from analogous notes and ending in the turn figure on the expected
pitch, feature different curves in between. This statement of the theme
modulates from B major to C. The turn figure alone recurs in several
other instances. These include the preparation for the first section-ending
cadence, the preparation for the preludes concluding phrase, and the
beginning of two motifs.2
1
Note that there is no immediate modulation. In mm. 5-6, B major is still the dominant of
E major. B major as a new tonic is only established after the F7 chord in m. 7.
2
Furthermore, L: mm. 3-41 can be recognized in L: mm. 40-41.
E major
415
The theme is complemented with nine motifs or figures. They are of

varying extension (two measures long, one measure, or shorter) and display different textures.
M1 appears in U: mm. 9-111. Its distinctive features are, rhythmically,
the three syncopations, and melodically, the broken diminished-seventh
chord on the second beat and the diminished-seventh interval at the end.
This motif is accompanied by a contrapuntal line (see M: mm. 9-111).
Dynamically, the motif contains two climaxes. The weaker one falls on the
first syncopation at m. 93 and, after the relaxation up to A at m. 102, is
followed by a stronger buildup through the second broken chord up to the
final powerful diminished-seventh interval. The contrapuntal companion
reaches the moment of greatest tension on its syncopation (i.e., the tied
note), thus describing a single curve. M1 recurs immediately. As in the
theme before, the imitation is set harmonically one fifth up from its
original position (see L: mm. 11-131). The companion is shortened and
simplified (see M: mm. 11-12), while the third voice (L: mm. 9-11 G-F
corresponds with U: mm. 11-13: D-C) appears much more elaborate.
M2 is also first presented in the upper voice (mm. 13-141). Its metric
organization and melodic design are very straightforward and create a
simple one-measure tension increase, in which the climax falls on the final
downbeat. M2 is imitated twice (M: mm. 14-151 and L: mm. 15-161). Harmonically, the imitations descend in the circle of fifths (m. 141 = B major,
m. 151 = E7, m. 161 = A7). This motif has no steady companion.
M3 is introduced after the B-major cadence. It consists of a two-part
complementary pattern spanning one measure (see U + M: mm. 18-191).
With only a minor adjustment in the initial interval, the complementary
motif is sequenced twice in descending direction (mm. 19-201 and 20-211).
The original motif is accompanied by a pedal in leaping octaves, as are its
two sequences. Dynamically, the initial figure (U: F-B-A-B) represents
a dynamic increase followed, in both voices in conjunction, by a relaxation
up to the next downbeat. On a larger scale, the descending sequences cause
an overall diminuendo throughout mm. 18-20. The third sequence omits
the upper-voice resolution. Instead, the lower-voice B turns into an appoggiatura to A (m. 211), thus linking this motif to the next.
M4 is even shorter. Its lower-voice leader consists of eight 16th-notes
(L: mm. 21-221) sequenced twice in descending order. The upper and
middle voices accompany the ascent of this motif with two metrically
accented and harmonically active double notes, while the descent is met
with the passive gesture of a note repetition in the upper voice alone. To
observe these details in the accompanying voices is essential since they
416
WTC II/09
emphasize the metric grouping of two-quarter-note length reminiscent of

the conspicuous hemiola characteristic of the section endings in Baroque
courante movements (see the climaxes in L: mm. 213, 222, and 231). The
preludes first half is then rounded off by a 1-measure cadential close
with abundant syncopation and, in the final measure, the elaborate key
confirmation typical for allemandes and courantes.
M5 (U: mm. 29-301, sequenced in mm. 30-311) follows the restatement
of the allemande theme at the beginning of the preludes second half. It
effectively combines features of several motifs heard previously: the first
four 16th-notes are reminiscent of the opening of M3, the following eight
16th-notes are similar to those in M2, the middle-voice indirect pedal
recalls the accompaniment of M3, and the metrically fixed, harmonically
active upper-voice notes place the originally hemiolic accompaniment of
M4 into a non-hemiolic pattern. Like M2, this motif describes a simple
dynamic rise. The climaxes in each statement are enhanced by the accompaniment, so that this motif appears almost homophonic.
M6 is launched after the one-measure cadential close that establishes
the tonic relative at m. 321. This motif consists exclusively of stepwise
motion describing a long curve in 16th-notes (A to A) followed by a
shorter one in 32nd-notes (B to B). From its original position in the middle
voice, M6 moves to the upper voice where it is subsequently sequenced
(see U: mm. 33-341 and 34-351). The accompaniment with its sequence
(L: mm. 32-33 . 33-34) suggests a contrapuntal voice, but later recurrences reveal that only the final four-eighth-note descent is characteristic
while the two initial notes may be changed or omitted. (See the recurrences
of M6 in U: mm. 41-421 and M: mm. 42-431.)
M7 is introduced in U: mm. 35-361 and imitated in M: mm. 36-371.
This motif sounds particularly charming if the hidden two-part content of
the final eight 16th-notes is emphasized: it thus presents itself as an initial
turn (U m. 351: C-B-A-B) answered by its expanded retrograde in the
higher register (see U: mm. 352-361, ABA-G). The climax falls, at least
in this interpretation, on the highest pitch on beat 3.
M8 is another motif launched with the turn figure from the theme
(mm. 37-381, sequenced in mm. 38-39 and 39-40). It is conceived in three
voices. The leading upper voice is complemented by the middle voice,
while the lower voice creates a contrapuntal contrast with its descending
scale. M8 contains four syncopations in each measure. After the link from
the end of the theme and reminiscences (M6, M4, part of the cadential
close that ended the first half of the prelude, and M3), m. 49 brings forth a
one-measure preparation for yet another cadential close on the tonic.
E major
417
The very last four-measure phrase of the prelude presents M9 in mm.

51-521 (see in U: the ornamented E major chord, complemented in M by
four descending 16th-notes). Together with its two sequences, this motif
represents the chords of the final cadence spread over a final tonic pedal:
m. 51 = I, m. 52 = IV, m. 53 = V9, m. 54 = I. The texture is gradually
expanded from three voices to four (mm. 51-52), five (mm. 52-53) and six
(m. 54). In keeping with the motifs original shape, the climax falls each
time on beat 3, followed by a relaxation to the next downbeat.
To sum up: In each of the preludes halves the length of the motifs
shrinks gradually while the harmonic development progresses in waves
through the circle of fifths.
theme (4 + 4 mm.)
E7
B
M1 (2 + 2 mm.)
theme (4 mm.)
M5 + cadence (1 + 1 + 1 mm.)
M6
(1 + 1 + 1 mm.)
B-C-G
C
M2 (1 + 1 + 1 mm.)
M7
(1 + 1 mm.)
E7
A7
cadence (1 + 1 mm.) B
M3 (1 + 1 + 1 mm.)
M4 (b+b+b mm.)
close (1 mm.)
= 24 measures
B
F
M8
(1 + 1 + 1 mm.)
end of theme (1 m.)
A7
M6
(1 + 1 m.)
B
M4
(b+b+b mm.) E
close ( mm.)
E
M3
(1 + 1 + 1 mm.)
B7
= 24 measures
+ cadence + M9 (3 + 3 mm.)
E
WTC II/9 in E major Fugue

The beautifully simple curve of the E-major subject leaves no room for
doubt about its length or phrase structure. Beginning on the downbeat of
the first measure it builds an almost symmetrical arch through four halfnotes before concluding in the middle of m. 2. The pitch pattern in the
subject features only small intervals: a single minor third is surrounded by
seconds. The rhythmic pattern encompasses the initial whole-note, four
half-notes and the final quarter-note. In the course of the fugue, Bach adds
eighth-notes and syncopated half-notes as well as dotted half-notes and
418
WTC II/09
various tied notes. The time signature is somewhat puzzling since we are
used to reading the as 2/2 time. Here Bachs indication refers to the
half-note as the relevant beat, without specifying that he is fitting four halfnotes, and not just two, into each measure. The subjects harmonic
background describes the simple progression IiiI64 V7I. The dynamic
development follows harmony (rising from the tonic to the subdominant,
falling on the way back to the tonic) as well as melody (increasing from
the keynote through ascending step and skip, and diminishing with the
descending line).
The fugue comprises 31 subject statements in a great variety of shapes
and sizes, but nevertheless in strikingly regular order.
1. mm. 1-2
B
17. mm. 26-27
S
2. mm. 2-4
T
18. mm. 27-28
A
3. mm. 4-5
A
19. mm. 28-29
T
4. mm. 5-7
S
20. mm. 28-29
B
5. mm. 9-10
A
21. mm. 30-31
B
6. mm. 9-11
T
22. mm. 30-32
A
7. mm. 10-12
B
23. mm. 30-31
T var 2
8. mm. 11-12
S
24. mm. 31-32
S var 2
9. mm. 16-17
A
25. mm. 31-32
T var 2
10. mm. 17-18
S
26. mm. 35-36
A
11. mm. 19-20
B
27. mm. 35-36
S
12. mm. 20-21
T
28. mm. 35-37
T
13. mm. 23-24
S var
29. mm. 36-38
B
14. mm. 23-25
A var
30. mm. 37-39
S
15. mm. 25-27
B var
31. mm. 40-41
B
16. mm. 25-27
T var
The answer is conceived without adjustments in the interval pattern.

At the same time, Bach makes fullest use of almost all modifications that
can traditionally occur to a subject. Besides abridgments of the initial and
final notes the concluding resolution is postponed and replaced by a dosi
do formula (mm. 21-22). The subjects minor-third interval is repeatedly
bridged by an inserted passing note, the climax often prolonged with a tie,
followed by a descent accelerated to quarter-notes (S/A: mm. 23-24; B/T:
mm. 25-26, at var in the table above). A diminution of all notes to half
their value occurs in mm. 26-29 (S, A, T, B) as well as in mm. 30-31 (B).
E major
419
Moreover, the diminution may be inverted. In conspicuous modification,

the initial note is additionally shortened and the initial interval increased;
this variant appears in mm. mm. 30-313 (T), 31-321 (S), 31-323 (T) and
35-361 (S). In the two statements that conclude the fugue, the final notes do
not find their expected harmonic resolution and (therefore) continue
their descent in half-notes (mm. 37-42: S, B).
Strettos occur frequently, involving statements in original rhythm,
diminutions, or a mixture of both. The follower may chase the leader at a
very short distance, thus creating an extended overlap (mm. 9-11), or enter
only after the leaders climax (mm. 16-17 and 19-20). Chain strettos, in
which a third entry overlaps with the second but not or hardly with the
first, appear twice in this fugue and involve all four voices (see mm. 10-12
and 35-38: A/T/B/S). While parallel statements are not used, transitory
parallels of subject segments do occur, particularly in the context of the
diminished entries, and serve to heighten the tension (see mm. 28-29:
A with T and m. 29: T with B).
As if the various modifications of the subject had not already provided
a rather complex picture, further color is added by a number of regular
counter-subjects. CS1 is introduced in the bass against the answer. It
begins after a little buffer or link, the inverted-mordent figure followed by
two leaps (B: mm. 2-3). The leaps, major third down and perfect fourth up,
recur several times in the course of the fugue, jointly or separately. The
counter-subject proper thus begins on a weak beat with an ascending
tetrachord. Its ending comes in two versions: as a syncopated, flattened
suspension (as in mm. 3-4) with belated final resolution (G) or a closing
formula (as in m. 5) with the penultimate note as leading-note resolving,
on time and upward, into the keynote. CS1 recurs as the subjects
companion in mm. 4-5 and 6-7, varied and late in mm. 11-12, varied yet
on time in relation to the subject statement in the alto in mm. 31-32, and
regularly again in mm. 36-37, 37-38, 38-39, and 40-41. The last two
entries are extended in keeping with those of the subject. Variations of the
counter-subject are also used as episode material; see below.
CS2 is first presented in the tenor of mm. 16-17. It begins likewise on
a weak beat and contains, after the initial fourth leap, two syncopations and
two inverted-mordent figures before a final three-note descent. This
counter-subject recurs, with variations that include a widening of the initial
leap to a perfect fifth, only in A: mm. 17-19 and S: mm. 19-20. It is thus
only of local importance. In m. 17, where it sounds as a companion to the
second entry in a stretto, it changes its position (entering a half-note late)
and its initial interval (to a fifth).
420
WTC II/09
CS3 appears closely linked to CS2; they only ever materialize together.
The third counter-subject is introduced in the bass of mm. 16-17. It
consists of four ascending semitones in a short-long, short-long rhythm
pattern, and a concluding do-si-do figure that recalls the ending of the first
counter-subject. Its recurrences coincide with those of CS2: mm. 17-19 (T)
and mm. 19-20 (A).
There are six subject-free passages in this fugue.

E1
mm. 7-91
E4
mm. 29-303*
E2
mm. 123-161*
E5
mm. 323-351
E3
mm. 22-231*
E6
mm. 423-43
* These episode endings overlap with the
beginning of the next subject statement.
The material of these episodes can be traced back to three different

origins. The CS1 variants were already briefly mentioned. They may
remain rhythmically faithful to the model (as toward the end of E1),
rhythmically modified (as in E2), or shortened in the middle (as in the
tripartite bass sequence in mm. 31-34). A second component consists of an
ascending perfect fourth and an inverted-mordent figure. i.e., of CS2
segments that have already been heard in inverted order in the buffer.
Finally, descending scales play an important role in the episodes. Together
with the various closing formulas, they facilitate the listeners orientation.
The only subject-free passage that does not feature any of these three
components is the final 12-measure cadence.
The dynamic gestures within these components are very distinct and
shape the episodes significantly: the CS1-variation forms a dynamic curve,
the ascending fourth leaps raise the tension, the inverted-mordent figures
sustain and the descending scales and closing formulas release the tension.
All episodes thus end with a release of tension. E1 and E2 begin with an
E major
421
active passage, in E5, the gradual buildup through the ascending sequences
of the CS1-variation begins already under the umbrella of the preceding
subject statement, and E3 and E6 create nothing beyond a cadential close.
Only in E4 is the harmonic close not only diverted into an interrupted
cadence (see m. 303: C minor instead of E major) but also overlaps significantly with the beginning of the following bass statement.
The basic character of this fugue is calm. The four half-note beats3 can
indeed be understood as fairly flowing, while on the other hand, fast
tempo for the half-measures would be incompatible both with the eighthnotes and, more importantly, with the spirit of this fugue. The relative
tempo of the prelude to the fugue is most convincing when larger units, not
smaller beats, are chosen to establish proportion. A good solution is to
make one prelude measure correspond with half a measure in the fugue.
(Approximate metronome settings: prelude beats = 100, fugue beats = 66.)
The articulation is primarily legato. Non legato applies only to quarter- or
half-notes that form cadential-bass patterns or consecutive leaps (B mm.
2-3: D-B-E; T + B mm. 7-8; etc.). Ornaments do not appear in this fugue.
The structure of this composition is determined by the introduction of
new transformations of the primary material. The concluding quality of all
episodes except for E4 supports this concept. There are five sections in the
E-major fugue.
Section I comprises four statements in ascending order followed by
E1. It concludes at m. 91 with a harmonic resolution into B major,
after a one-measure dominant pedal. Except for the unaccompanied
first statement, all subject entries are accompanied by CS1.
3
The meter and, by extension, the tempo require some pondering. This is due to the time
signature. As has already been mentioned, the (alla breve) usually stands for 2/2 time.
Thus performers may wonder whether in measures of two whole-notes each the information should be read to mean only two beats in a measure or rather count half-notes.
A helpful piece of evidence is found in Bachs Partita No.6 in E minor. In the final
version that performers use today, its gigue appears written with measures of double
duration, with the time marked by a slashed circle ( ). That this unusual metrical indication must be read as a 2/1 (= count two whole-note beats in each measure) is supported
by the corresponding time signature found in the earlier version of this partita, included in
Bachs Notebook for Anna Magdalena, where all notes of the gigue are written in double
speed (eighth-notes originated as 16th-notes etc.) so that each measure contains four
quarter-notes. The matching time signature is (alla breve), which unquestionably means
count two beats in each measure. As it is most unlikely that the final version of the gigue
with its augmented note values was not meant to retain this metric order, one may be
permitted to deduce that the symbol stands for two beats in a measure of eight
quarter-notes. By extension, the in the equally long measures of the E-major fugue from
WTC Book II should be read as count half-notes, i.e., four beats to the measure.
422
WTC II/09
Section II also encompasses one subject statement in each of the
four voices followed by a concluding episode (E2). The statements
are grouped into two strettos here, with the follower entering in
both cases on the second note of the statement presented by the
leader. The episode is longer and much more substantial in
material than the one in the first section. One could almost read this
as an attempt to make up for the shrinking of the section as a result
of the stretto overlap. The section ends at m. 161 in C minor, the
tonic relative.
Section III, too, comprises four subject statements and an episode.
As before, the entries are grouped in two strettos, with the distance
between leader and follower now an entire measure. New contrapuntal material is presented in the form of the paired components
CS2 and CS3. The section closes at m. 231 in F minor, the
subdominant relative.
Section IV introduces the variation of the subject, presented in all
four voices, which are once again grouped in two strettos. The distance between the entries of each stretto is smaller than ever before,
and both strettos together cover only four measures. Without an
episode to serve as a buffer, another transformation of the subject is
introduced: its diminution. The distribution (one diminished statement in each of the four voices) and the structure (two strettos) are
not surprising, even less so since these strettos are built by the
voices that were paired just before in the strettos with the varied
subject (mm. 23-24 S + A and mm. 25-26 B + T: subject variation;
mm. 26-28 S + A and mm. 28-29 T + B: subject diminution). The
distance in the stretto resembles that of section III, albeit in half size.
After E4, which serves as an interruption without closing, the fifth
stretto combines one voice from each of the two groups in a combination of two different versions of the subject (mm. 30-31 B:
diminution, A: original values). One could thus describe the entries
as presenting two intertwining rounds. There are 4 + 1 entries of the
varied subject and 4 + 1 entries of the diminution. The redundant
stretto combining the two different shapes is further distinguished
by the surrounding material: it is accompanied by three entries of a
diminished and inverted subject variation (mm. 30-32: T, S, T) as
well as a variation of CS1 that extends, with its sequences, into the
subsequent episode. The fourth section thus combines features from
all three preceding sections before concluding at m. 351 in G minor,
the dominant relative.
E major
423
424
WTC II/09
Section V presents five subject statements. The first two appear in

stretto, at the same distance as those heard in section II, while the
third and fourth statements overlap only very little. Both the fourth
and fifth statements are extended by a descending scale in halfnotes. Four entries of CS1 relate this section to the initial one. E6
concludes the sectionand the fuguein m. 43.
In terms of the development of tension, section I presents a fairly
strong increase, which is surpassed by that in section II. At the respective
end of each section, E1 grants a gradual relaxation while E2 provides a
distinct color contrast before it concludes, with a cadential close in the
tonic relative, on a particularly soft note. In section III, the minor mode,
the softening of the contour with passing notes, and the lessened canonic
density all add up to a result of reduced intensity, while the new countersubjects make a change of sound color seem desirable. Section IV begins
even more softly but then describes a powerful increase that will need the
whole of the fifth section to achieve full relaxation.
WTC II/10 in E minor Prelude

This prelude is written in polyphonic two-part texture. It is based on a
single motif and its variations. This motif is imitated, sequenced entirely
and partially, and developed in different ways. All these features point at
a familiar genre: the invention. A perfunctory glance reveals that the
prelude consists of two halves, both of them repeated. The first half counts
forty-eight measures, the second is longer with altogether sixty measures.
Within the first half there are no outstanding closing formulas that would
immediately hint at structural sectioning. We must therefore consider
harmonic processes in conjunction with particularities of the thematic
development. The second half, by contrast, features two traditional closing
formulas that aid the performers and listeners orientation. The following
list mentions only the harmonic steps, leaving an investigation of the
thematic details for later. There are six sections:
I
mm. 1-111
i-v
E minor to B minor
II
mm. 11-231
v-VI
B minor to C major
III
mm. 23-48
VI-v
C major back to B minor1
IV
mm. 49-721
v-iv
B minor to A minor2
V
mm. 73-103
VII (i) interrupted cadence
VI
mm. 103-108 return to i: perfect cadence
The preludes two halves are related by means of several structural
analogies. The following passages of the first half recur in the second half
in a transposition one fifth down or one fourth up:
mm. 26-29
.
mm. 82-85 (voices inverted)
mm. 30-36
.
mm. 86-92
mm. 413-47
.
mm. 963-102
In addition, sections IV and V (the first and second sections in the
second half of the prelude) begin with analogous passages:
mm. 49-531
.
mm. 73-771 (voices inverted and varied)
mm. 53-571
.
mm. 77-811 (voices inverted, mode changed)
1
The first ending continues with a link to the repetition in m. 1. As a representative of the
dominant chord, this link uses the major third and the leading-note D.
2
This section-ending cadence is followed, as was the end of section III, by a one-measure
link that modulates to D major, the (major-mode) relative of the (minor) dominant.
425
426
WTC II/10
Equally analogous are the first and second sections of the preludes
first half, but here we are dealing not with literal transpositions but with
structural correspondences:
mm. 1-111
.
mm. 11-231
motif, imitation
motif, imitation + sequence
motif + 2 sequences
motif + 2 sequences
To sum up: Both halves of the prelude feature inner correspondences
between their first and second sections. These are not literal but structural;
they concern primarily the leading voice in each measure. The latter parts
of the two halves display a very close analogy. The correspondence mostly
involves both voices and uses literal transposition.
The prelude is based on a simple rhythmic pattern with a predominance
of 16th- and eighth-notes. Exceptional 32nd-notes occur only in mm. 3, 4,
12, and 22 where they represent written-out ornaments (inverted slides)
and in mm. 77 and 97 where they spell out the pitches of the trill suffixes.
The basic character of this invention is thus rather lively. The tempo may
be fairly swift; sensing whole-measure beats is preferable to counting in
eighth-notes. The corresponding articulation consists of a light non legato
for the eighth-notes, a crisp quasi legato for the 16th-notes, and legato only
for the ornamental 32nd-notes. One exception to the non legato in eighthnotes is indicated by a slur (U: m. 51). It is possible to interpret the
following m. 52 as a varied sequence, in which case one would play it with
a similar two-note slur.
There are a host of ornament symbols to be considered: mordents,
inverted mordents, turns, compound ornaments and trills. Mordents occur
in mm. 18, 20, 51, 52, and 71. In mm. 18 and 20, they form an integral part
of a melodic figure. In both cases they begin on the upper auxiliary and
comprise four notes. The mordent in m. 51 is approached in stepwise
motion and thus begins on the main note; a three-note shake is sufficient in
the swift tempo. The remaining two mordents are printed in brackets. As
the circumstances differ, the two cases must be decided independently. In
m. 52, the 16th-note figure C-B-C is itself a written-out ornament, so that
an additional mordent might cause congestion.3 Conversely, the mordent in
m. 71 appears in a typical cadential formula that, in Bachs time and style,
would always have been played with an ornament, whether or not this is
indicated. Inverted mordents appear in mm. 37-41, 92, 95, 96, 102, and 107.
In all cases, the lower neighbor notes are taken from the E-minor scale. As
3
A further consideration for omitting this mordent is that this would be the only case in the
entire prelude where a 16th-note is ornamented. All other ornaments decorate longer notes.
E minor
427
this ornament is used twice in closing formulas (see mm. 102 and 107), it
may be a good idea to play the corresponding note in m. 47 (C) with the
same inverted mordent.
Turns are found in mm. 47, 57-59, 78, 102, and 107. The first begins
on the upper neighbor B and uses the sharpened lower neighbor note G.
Structurally identical turns occur in mm. 102 and 107. Both also begin on
the upper neighbor, and the first also requires a sharpened lower note (see
m. 102: D-turn with C). The turns in mm. 57-59 all appear in the context
of scalar motion and therefore begin on the main note. Within their fivenote figures, the lower neighbor repeats the pitch of the preceding eighthnote (e.g., in m. 57: G). As this ornament is integrated into a motivic
figure that continues in sequences, it should be transferred to the second
eighth-notes in mm. 60 (L), 61 (U), 62 (L), and 63 (U). The turn in m. 78
repeats the same five-note ornament once again.
Trills are required in mm. 29-32 and, correspondingly, in mm. 33-37,
86-88, and 89-92, as well as in m. 97. The four earlier trills serve to
prolong the sound in a sustained pedal note rather than taking part in a
melodic line. They had therefore best begin on the main note and launch
their shakes in 32nd-notes right away, so that the main note and not the
auxiliary meets the melodic notes in the other voice.4 These extended trills
all end either with a tie or before a rest. None of them requires, or even
allows for, a suffix. Instead, they all come to a halt on the last main note
before the bar line. The remaining trill in m. 97 is most straightforward as
it is an ornament with melodic content: it begins on the upper auxiliary E,
shakes in regular 32nd-notes and ends with the suffix as marked by the
composer.
Compound ornaments appear in mm. 43 and 77. The former begins (as
indicated by the vertical stroke at the symbols left) with an appoggiatura
E that resolves after an eighth-note, i.e., against the B in the lower voice.
The shake that follows is that of an interrupted trill: it consists of four fast
notes and stops short in a metrically weak position. This ornament may be
transferred to the corresponding note G in m. 98. The second compound
4
18th-century style distinguished melodic and non-melodic trills: The purpose of melodic
trills is to heighten the interest of the ornamented note by repeated appoggiaturas; thus an
approach from the upper neighbor note is generally the rule. Wherever the melodic surroundings bring about a beginning on the main note, the initial note is lengthened so that all
other beats in the measure are met by the upper neighbor note, giving the impression of a
repeated appoggiatura. The purpose of non-melodic trills is usually, as in this piece, to
sustain the sound of a note that, on a keyboard instrument, would otherwise fade far too
early. In this case the sound itself, i.e., the main note, is the issue. It thus falls on the beats.
428
WTC II/10
ornament begins, as the convex curve preceding the symbol indicates, from
the lower neighbor C. As it is followed by a written-out suffix, it should
be interpreted as a note-filling trill. It therefore shakes in regular 32ndnotes (C-D-E-D-E-D-C-D).
The main motif of this invention spans two measures. It begins after
a downbeat rest and ends at m. 31. The motif can be regarded as consisting
of two symmetrical halves: the first, from G to B, contains a four-note
ascent, a skip in the opposite direction, and a final ascending step; the
second half, beginning again on G, correspondingly encompasses a
four-note descent followed by a skip in the opposite direction and a final
descending step. In terms of tension, the motifs first half presents an
increase with the climax at m. 21. The second half adds the complementary
decrease. In its complete and unvaried shape (but including inversions), the
main motif recurs fifteen times:
1.
mm. 1-3 U
9. mm. 17-19 L
2.
mm. 3-5 L
10. mm. 19-21 L
3.
mm. 5-7 U
11. mm. 21-23 L
4.
mm. 7-9 U
12. mm. 49-51 Linv5
5.
mm. 9-11 U
13. mm. 53-55 Uinv
6.
mm. 11-13 L
14. mm. 73-75 Uinv
7.
mm. 13-15 U
15. mm. 77-79 Linv
8.
mm. 15-17 U
The initial statement of the main motif is accompanied by the octave
leap E. This does not form part of the polyphonic pattern but only serves as
a harmonic support. It should therefore be played in neutral touch. Later
statements of the main motif are accompanied by various counter-motifs
(CM). The unifying feature in all of them appears in the second measure:
CM1 accompanies the climax of the main motif with an inverted slide
followed by an ascending seventh leap. Preceded by varying figures, it
occurs exclusively in the upper voice: see mm. (3)-4, (11)-12 and (21)-22.
CM2 features three ascending eighth-notes followed by a cadential step
(fifth down or fourth up) to the final downbeat. Equally preceded by
varying upbeats, it occurs exclusively in the lower voice; see mm. (5)-7,
(7)-9, (9)-11, (13)-15; also, inverted and with varied ending, in mm.
(53)-55, (73)-75. CM3, the only melodically distinct companion of the
principal motif, appears only twice, in U: mm. 17-19 and 19-21.
Each of the four inversions is followed by two partial sequences; see L: mm. 49-53,
U: mm. 53-57, L: mm. 73-77, U: mm. 77-81.
E minor
429
Beginning in m. 23, the central motif forms five distinctly identifiable

variants, M1a-M1e.6 Their dynamic impact is determined both by their
textural density and by the direction of their sequences. The picture is quite
simple: M1a, M1c and M1d appear exclusively in stretto texture and in
ascending sequences. Their dynamic value is therefore that of tension
increase. (The fact that the final upper-voice sequence of M1d sounds an
octave lower is owed to the limited range of the 18th century keyboard.)
M1b appears accompanied by pedal notes, thus in a polyphonically sparse
setting. All sequences move in descending direction, so that M1b induces
a decrease in tension. M1e plays a special role both because of its extended
scope and because of the almost homophonic effect created by the mirrored parallels. The overall dynamic movement is a very slight increase in
the first statement and an equally gentle decrease in the second.
The tension develops in the preludes first half as a large crescendo
diminuendo curve (sections I + II) followed by two smaller curves (III:
mm. 23-29/29-37 and 37-41/41-48). Section IV combines a concave (mm.
57-63) with a convex curve (mm. 63-72), while section V resembles
section III. The prelude ends with a final dynamic curve (mm. 103-108).
WTC II/10 in E minor Fugue

This subject is long and multifaceted. Spanning nearly six measures,
from the upbeat to the middle beat of m. 6, it stretches over more than an
octave and includes a surprising number of distinct little motivic patterns.
When examining its phrase structure, one quickly discovers a rough
division into two contrasting segments. The first ends on the middle beat
of m. 2 with a return to the keynote. Its outline can be described as a
slightly leaning curve (leaning because it ascends slowly but falls back in
less than half the time). The second segment is characterized by three syncopations that describe a gradual descent (see m. 2: C, m. 3: A, m. 4: F).
After some suspension, they also resolve into the keynote (m. 6: E). Within
6
M1a in mm. 23-251 (U), 24-26 (L), 25-27 (U), 26-28 (L), 25-27 (U), 81-83 (L), 82-84
(U) as well as divided between mm. 28-29 + 23 and mm. 84/85 + 81;
M1b in mm. 29-32, 32-33 (U), 33-35, 35-37 (L); 85-87, 87-89 (U), 89-91, 91-93 (L);
M1c in mm. 37-39 (L), 38-40 (U), 39-41 (L), 40-42 (U), 94-96 (L), 95-97 (U) as well as
modified in mm. 93-95 (U);
M1d in mm. 57-59 (L), 58-60 (U), 59-61 (L), 60-62 (U), 61-63 (L), 62-64 (U); and
M1e in mm. 63-671 (L) and 67-711 (L) as well as inverted in mm. 67-701 (U).
430
WTC II/10
each of these larger segments, smaller subphrases are established by way

of sequences. These smallest components set out with almost minimal
length (the first subphrase is less than two quarter-notes long as the upbeat
features only two out of three triplet eighth-notes), then grow longer (two
quarter-notes), and longer (four quarter-notes in the third, fourth, and fifth
subphrases), until the final subphrase spans nearly two measures.7
Pitch and rhythm in the subject are as intricate as the phrase structure.
There are ornamental steps as well as melodious seconds (C-B and A-G in
mm. 2-4), a broken chord, two leaps created by the hidden two-part texture
and the diminished seventh that, in m. 5, links the indirect extension to
the subjects main body. Furthermore there are 16th-notes, triplet eighthnotes, dotted eighth-notes, quarter-notes, a quarter-note tied to a triplet
eighth-note, and syncopated half-notes. In spite of the stunning variety of
six subphrases, five kinds of intervals, and six different note values there
is nonetheless a unifying force. If one strips the subject of its written-out
ornaments (the pre-beat turns in m. 1), its broken chord (m. 2) and its
charm notes (the D-E in mm. 3 and 4 that, while adding beauty, do not
contribute to the basic line), the skeleton reads like this:
This line constitutes an important guide for the performer who might
otherwise lose sight of the whole in the face of so many enchanting parts.
It is also helpful when we set out to determine the subjects dynamic
shape. The local climaxes in the three sequencing subphrases of the beginning form a crescendo while the three syncopated climaxes initiating the
remaining subphrases describe an overarching diminuendo. The most expressive of these climaxes is the C in mm. 2-3. As the first syncopation it
brings a rhythmic surprise, as the peak of the simplified line it marks the
natural turning point of the tension, and as the bearer of the subdominant
chord it also highlights the active harmonic step. The following example
shows the combination of harmony and dynamic outline:
The scope of this final subphrase is achieved by an indirect extension. It is indirect because
it cannot be omitted without leaving the subject incomplete. However, one might wish to
try playing the final subphrase (from F in m. 4) with the eighth-notes in the first half of m. 5
transposed one note lower and the middle beat bending back and resolving onto E: the
subject might convincingly have ended here. It thus does not come as a surprise that Bach
later uses the curved triplet run (from C down to E) as a motif independently of the subject.
E minor
The subject appears nine times in the course of the fugue.

1
mm. 0-6
U
6
mm. 41-47
2
mm. 6-12
M
7
mm. 49-55
3
mm. 12-18
L
8
mm. 59-65
4
mm. 23-29
U
9
mm. 71-77
5
mm. 29-35
M
431
L
M
U
L
Bach invents one counter-subject for this fugue. It proves a faithful

companion to the subject. CS is introduced against the subjects answer, in
the upper-voice part of mm. 7-12. The last five eighth-notes of m. 6 recur
more often than not (see M. m. 12 and 41, L: m. 49, and U: m. 71), so that
one might ask whether they do not form the upbeat of CS. Yet, as they
sequence the sixth subphrase and are also widely used as episode material,
their connection with CS cannot be ascertained. We therefore prefer to
regard the descending scale as a link between the end of the subject and the
beginning of the counter-subject.
In contrast to the subject, which remains unchanged throughout the
fugue, CS develops certain modifications. In several later statements, it is
either split between two complementary voices (mm. 13-18: M + U and mm.
24-29: L + M), or it switches voices halfway through the phrase (mm. 30-35:
L/U, mm. 42-47: M/U, and mm. 72-77: U/M). The only CS entry closely
resembling the original shape occurs in mm. 50-55 (L), while the entry in
mm. 62-64 recalls only a fragment. Such frequent variations of a countersubject invite examination of what might be regarded as its basic format.
To determine its simplified line is not at all difficult if one takes as
guidelines the modifications occurring in mm. 13-18, 42-47, and 72-77.
The splitting into two voices and the very free variation of the unaccented
eighth-notes indicates that the primary line is concentrated in the main
(half-note) beats. These half-note beats present a surprise: what appears at
the surface as a fairly independent contrapuntal line turns out to be an
embellished parallel (in thirds or sixths respectively) of the subject. The
432
WTC II/10
main dynamic outline will therefore have to be drawn in parallel to that of

the subject; only within the subphrases is independent shaping possible.
The following examples give first the juxtaposition of simplified subject
and simplified counter-subject, then a possible interpretation of the tension
processes in the two thematic components.
The E minor fugue encompasses six subject-free passages.

E1 mm. 182-23
E4
mm. 552-59
E2 mm. 352-41
E5
mm. 652-71
E3 mm. 472-49
E6
mm. 772-86
Several of these episodes are divided into distinguishable segments: A
change of material and a completed modulation to the tonic relative G major
separates E1a (mm. 182-202) from E1b (mm. 202-23). E5a (mm. 652-702) is
followed after the sudden homophonic elements of fermata and general
pause by E5b (mm. 702-71). The final episode comprises three segments:
E6a (mm. 772-812), E6b (mm. 812-83 fermata) and E6c (mm. 832-86).
None of the episodes can be described as just a cadential close. Instead, all
of them use material from the subject8 and also develop a motif character8
The subjects final measure is taken up in E1a (L: 2x), E2 (M: 1x mm. 35-36), E3 (L: 2x),
E4 (U: 3x and M: 1x), as well as in E5a (M: 1x and L: 1x). The five-note descent alone
recurs separately in E1b (M: 3x), E2 (M: mm. 36-37, all voices alternating mm. 37-41),
E5a (M: m. 68), E5b (L: 2x), and E6c (U: m. 84). Another derivation, also featuring six
eighth-notes in this metric position and in scalar motion, is the ascent with varying endings
appearing in E1b (U: 3x), E5a (U/L 3x), E6a (M: mm. 78-79), and E6b (U: mm. 84-85).
E minor
433
ized by a dotted rhythm that comes either in an inverted-mordent figure

plus leap (M1a, introduced in mm. U: 18-19) or in a simple ascent (M1b,
introduced in M: m. 19). Other noteworthy features include the sequence
model in the lower part of E1b (not imitated in another voice or taken up
anywhere later in the piece), the three-part complementary pattern over a
dominant pedal in E6a (mm. 79-81), and the voice-splitting to a four-part
chord in m. 83 and to a four-part texture all through the measure in m. 86.
A most obvious relationship exists between E1a and E3.
The role played by each episode in the development of tension largely
follows the direction of the sequences. The crescendo/diminuendo in E1a
+ E1b gives this episode the shape of a closed curve. E2 describes a similar
curve, despite some ambiguity in the second half (L: falling, U and M:
rising in mm. 39-41), as does E4. E3 recalls only the rising first half of E1
and thus suggests that it serves to connect two subject statements. E5,
which features a rhythmic interruption with fermata and general pause,
nevertheless also conveys a sense of linking entries. Its short segment E5b
effects a compelling rise of tension as a preparation for the final subject
statement. E6 is the longest and most diverse episode. It contains true
toccata features (mm. 79-83) but is united by the pedal note B supporting
the episodes first and third segments (mm. 78-81 and 84-85). The texture
and complementary rhythm of the final measure, together with the ending
on a weak beat, are reminiscent of the key confirmations found in most
allemandes and some courantes of Bachs suites.
As suggested by the alla breve time signature as well as the ornamental
and broken-chord figures in the surface structure, the basic character of the
E-minor fugue is rather lively. That makes the wedges on the quarter-notes
all the more interesting. In a lively character, these quarter-notes would be
played detached anyway. The wedges transform the otherwise passive, soft
detachment into an active, energetic approach to the note values. In all
subject statements throughout the fugue this concerns the final notes of the
first five subphrases as well as the broken-chord figures and the notes
concluding the subphrases in the counter-subject. By contrast, Bach never
once marks the broken chords in the episode motif with wedges. They
should thus be executed in a distinctly gentler non legato. Legato applies to
all longer note values that represent an appoggiatura, i.e., particularly the
first note pairs in the subjects fourth and fifth subphrases.
A short word on the execution of the various rhythmic values is in
order. Both the four 16th-notes in the written-out turns of the subjects first
measure and the triplets in the subjects final subphrase are of course to be
played exactly as noted. Throughout the fugue, the two rhythmic figures
434
WTC II/10
never once coincide, thus the problem of polyrhythm does not arise. The
dotted-eighth-note figures, however, regularly sound against triplets. As
Bachs manuscript displays his writing of note head against note head very
clearly (see, e.g., mm. 18ff), these dotted-note figures are intended to be
read in gigue rhythm, i.e., in a proportion of 2 + 1 instead of 3 + 1. While
most performers intuitively choose this reading in m. 12 and from m. 18
onward, many are unaware of having played a different, strictly dotted
rhythm in the three initial statements of the subject. Consistency requires
that a chosen rhythmic rendering of a recurring segment should be retained
throughout a composition. This means that the dotted-note figure in m. 3
should already be given the rhythmic shape it is going to have throughout
the piece. (A slightly odd though basically corresponding case occurs on
the final beat of m. 83: the 16th-note upbeat in the middle and lower voices
also makes more sense if read in triplet rhythmwith the dotted note here
substituted by rests.)
The tempo does not permit much individual variation. It must be fast
enough to convey the alla breve pace but moderate enough to allow for
transparency in the ornaments 32nd-notes. Thus unhurriedly swinging
half-notes give a good measure. The tempo proportion between the prelude
and the fugue gives best results if one measure in the preludes 3/8 time
corresponds with half a measure (a half-note) in the fugue. (Approximate
The score contains both thematic and cadential ornaments. The first
thematic ornament appears in the subject at m. 11. It is an inverted mordent
printed in parentheses on three occasions (mm. 1, 7, 13). The question is
not so much which notes to play but rather whether or not to play it at all.
The three consecutive markings indicate that the ornament is to be treated
as an integral part of the subject; omitting the symbol in further subject
statements was usual practice. The parentheses mark it as a later addition,
the full print as stemming from Bachs own hand. (Additions found in
copies by other hands are distinguished in the Urtext by small print.).
Performers who can technically master the required speed and middlevoice acrobatics (in mm. 30 and 50) should not forgo this embellishment.9
The second thematic ornament decorates the counter-subject. It is printed
only once, though without brackets (m. 10). Placed against a half-note, a
note that must be sustained and linked, it poses a much bigger problem.
The option is therefore to play it only onceor not at all.
9
The statement in mm. 41-47 features the mordent on the first syncopation. This seems
somewhat arbitrary and could probably be silently adjusted.
E minor
435
The non-thematic ornaments, by contrast, are straightforward. The

mordents under the fermatas in mm. 70 and 83 begin on the main note and
can easily accommodate five notes as there is ample time. The same holds
true for the turn that, after the rhythmic break caused by the fermata and
the melodic break caused by the octave leap, appears as the beginning of a
new phrase (D-E-D-C-D). The mordent in m. 85 is launched from the
upper auxiliary as is the turn in m. 37, both comprising four notes.
In the absence of conspicuous cadential closes in episodes and of
subject statements appearing in reduced ensemble, the harmonic outline
and the detailed use of material must be examined in order to determine the
scope of each section. Section I does not pose a problem. Its three entries
sound on the tonic, the minor dominant, and the tonic respectively. As the
first episode develops from the preceding statements ending by way of
sequences, cutting it off would be quite unthinkable. As a result, section I
ends with the conclusion of E1, at m. 241, in a brief overlap with the next
subject entry. Section II is distinguished from section I by a change to the
major mode. It contains two subject statements, in the key of the tonic
relative and its dominant respectively, plus another episode. As before, the
end of the section (at m. 421) overlaps with the upbeat of the subsequent
lower-voice statement. Section III returns to the harmonic realm of E minor,
with two entries on the dominant and tonic respectively. The linking
episode E3 and the concluding episode E4 together make up six measures,
thus matching the length of each of the episodes in sections I and II. In
terms of structural balance, E3 can thus be regarded as an anticipated
segment of this sections concluding episode. Section IV presents the statement on the subdominant and thus truly confirms the return to the home
key. There are again two statements but, as expounded above in detail, the
episodes are much extended and expose material that might seem alien to
the fugue. Yet Bach has composed these 27 measures in such a way that
they clearly form one large inseparable unit, despite fermata and general
pause. In fact, the entire section represents one protracted cadence in E
minor. Shortly after the end of the subject entry on the subdominant in
mm. 60-65, the lower voice features a chromatic descent to the dominant
pedal (see mm. 66-70: DC-C-BA-B). Suspended for the duration of
the final subject entry, the dominant pedal is resumed in mm. 78-81 and, in
mm. 84-85, prepares the final tonic bass.
After an exposition with statements in all three voices, the three
subsequent sections present the three possible pairings. In this manner, this
fugue is perfectly balanced with three subject statements in each of the
parts.
436
WTC II/10
All four sections describe an increase in tension from their initial to the
last statements. In sections I and II, these increases are not interrupted by
episodes, and E3 in section III with its rising tendency also maintains the
drive from one entry to the other. Only section IV reverses these facts. The
initial (subdominant) entry appears more intense than the second (tonic)
entry because of the harmonic release from subdominant to dominant and
tonic, a prevailingly relaxing attitude in E5, and the three-measure thinning
of the ensemble in mm. 702-732, which makes this subject statement the
only one in the fugue to appear in reduced surroundings after the full threepart texture has been established. The final episode then describes a full
dynamic curve analogous to those in the other three section-concluding
episodes.
Among the four sections, the second one in the major mode may be
regarded as most outgoing in character. The third section is still on an
elevated level, and only the final section returns to the minor-mode level
and intensity of the fugues beginning. The differences, however, are not
dramatic. This is another playful fugue in which the thematic material as
such captures the attention more than its manifold modifications.
WTC II/11 in F major Prelude

The F-major prelude is a meditative piece. It is determined by metric
features, particularly the uninterrupted eighth-note motion, the continuous
presence of half-note beats in at least one of the complementary voices,
and the abundant use of ties combining selected eighth-notes into chords.
There are up to five voices, but the part-writing is not fully consistent.
Although the main melodic motif moves between the players two hands,
it can hardly ever be allocated to any one voice. It may be preferable not to
speak of voices or parts in this composition but to perceive the texture as
eighth-note lines embedded in four- to five-part homophonic chords. The
initial phrase of the prelude could thus be represented as follows:
The initial harmonic progression concludes in the middle of m. 8. As

one voice after the other resolves, the tonic is regained in a metrically
weak position on the third beat. This is a structurally relevant cadence. The
subsequent eight measures modulate to the dominant key C major. The
striking correspondence of mm. 17 and 1 suggests a new beginning on the
dominant and thus reveals a larger structural grouping of two harmonic
progressions into one section. The prelude comprises altogether four such
compound sections:
I mm. 1-8-16
I, V
II mm. 17-24-32
V, iii
III mm. 33-40-47-56 vi, V, iii
IV mm. 57-64-72
IV, I
There are several extended analogies: mm. 1-111 . 17-271 (transposed),
mm. 1-63 . 57-623 (identical), mm. 1-32 . 33-352 (transposed and varied),
mm. 11-162 . 67-722 (transposed and varied). Shorter correspondences are
composed as full-ensemble sequences: mm. 41-42 . 43-44 . 45-46.
437
438
WTC II/11
The preludes meditative character requires a calm tempo and overall

legato playing. As the slurs in m. 1 indicate, the eighth-notes must not be
perceived as active individuals. Instead they should be grouped in such a
way that only the metrically relevant ones carry some weight. Dynamic
shaping can easily be misinterpreted in this piece. The information is not
encapsulated in melodic eventsin whatever voicebut in the interplay
of extended harmonies woven into arabesque-like threads. Note-to-note
increases or decreases would therefore mislead listeners. Gentle curves do
occur, built by the metrically emphasized notes within the eighth-note
threads and by the harmonic development within each phrase.
The prelude features no ornaments except for a single grace-note in
m. 66. Its appearance in brackets indicates that Bach may not have notated
it in his original manuscript. Since it adds greatly to the smoothness and
beauty of the measure, it can safely be recommended. The grace-note
represents an appoggiatura and should thus be played for half of the main
note value (i.e., appoggiatura B = half-note, resolution A = half-note).
The weaving eighth-note lines build primarily three patterns. All begin
on the first eighth-note after a strong beat with a three-eighth-note ascent
that serves as an upbeat to what follows. Furthermore, all patterns consist
of little four-note groups and include a varying number of sequences. The
metrically strong notes of each four-note group may or may not be tied
over to form sustained chords. These patterns will be referred to as P1, P2,
and P3 to remind us that we are not dealing with melodic motifs. The three
patterns in conjunction account for the entire prelude. On page 439 below,
the relevant features of the structural layout are listed in table format, so as
to allow easy comparison and overview.
P1 is introduced in mm. 1-2. An inverted turn (E-F-G-F) leads into
four descending turn-figures (E-D-C-D, C-B-A-B, A-G-F-G, F-E-D-E).
At the end of m. 2, a renewed upbeat seems to trigger an imitation that,
however, sounds much more like a continuation of the same pattern. The
line of metrically enhanced notes in mm. 1-3 (F, D, B, G, E; C, A, F, C) is
extended with two further imitations of the turn-figure in a higher register
adding G and E. Dynamically, P1 represents a protracted decrease.
P2, introduced in mm. 5-71 and sequenced one tone lower in mm. 7-91,
features three inverted-turn figures in ascending sequences complemented
by two turns in descending sequence. (The third turn is extended and varied:
for a simple A-B-C-B mm. 5-6 present A-B-CA-B). The metrically
highlighted notes form a curve: D, F, (C)-B, G, E, as does the tension.
P3 is first heard in the cadential close of the first section (see the upper
right-hand part in mm. 143-161). It is characteristic only at its beginning,
F major
439
mm.
Section I
1-4
5-71
7-91
9-111
11-143
143-161
16-171
pattern
clue notes
structure
P1
P2
P2
P1
P1
P3
P1
F-D-B-G-E-C-A-F-C + G-E
D-F-(C)-B-G-E
C-E-(B)-A-F-D
D-B-G-C-A-D
F-D-B-G-E-C-A-F-D-B + E
C/AC/C
G-E-C
complete phrase
two-measure model
two-measure model sequenced
short version, phrase 1
variation of phrase 1
closing formula
link
Section II
17-20
21-231
23-251
25-271
27-293
29-322
32-331
P1
P2
P2
P1
P1
P1
P1
C-A-F-D-B-G-E-C-G + D-B
A-C-(G)-F-D-B
G-B-(F)-E-C-A
A-F-D-G-E-A
D-B-G (parall. F-D-G) + C-A-D
G-E-C-A-F-D-B-G
A-F
complete phrase
two-measure model
short version, phrase 1
development of phrase 1
link
Section III
33-37
37-381
38-401
40-411
41-431
43-451
45-471
47-491
49-511
51-541
54-56
P1
P2
P3
P1
P3a
P3a
P3a
P1
P1
P1
P1
D-B-G-E-C-A-F-D-B + D-G
E-B-C
A/EE/D
F-D-B
B / G A-F-D
D / B C-A-F
F / D E-C-A
F-D-B-G-E-C
E-C-A-F, F-D
A-F-D, F-D-B, D-B-G
E-C-A-E, B-A-E-C
development of phrase 1
development of pattern
link
two-measure model
free imitation
partial sequences
cadential close + link
Section IV
57-60
61-623
623-642
643-671
67-702
703-721
P1
P2
P3
P1
P1
P3
F-D-B-G-E-C-A-F-C + G-E
D-F-(C)-B-G
B / E D / B
E-E, G-E-C, F-D-G
B-G-E-C-A-F-D-B-G-E
F / D F / F + A-F
complete phrase, see 1-4

two-measure model, see 5-6
free development
free development
closing formula
where it features not a written-out ornament but a straight four-note ascent.

The immediate imitation can be distinctly perceived as a different level in
the polyphonic structure, unlike the imitations in P1 and P2, which appear
as integral parts of a single complementary line. The relaxation following
the twofold ascent abides by no pattern. A variation of P3 occurs also in
440
WTC II/11
the third section. The pattern establishes a model in which the initial ascent
and its imitation are complemented by a triple turn (see mm. 41-431). This
variant is referred to in the table above as P3a. It is followed by two
sequences. In both cases this pattern, like that of P2, forms a curve. The rise
and decline in P3 are slightly steeper than their counterparts in P2.
WTC II/11 in F major Fugue

Before one considers the scope of this fugues subject, the time signature requires a comment. Bachs 6/16, like many other compound time
signatures, does not indicate that performers should count six beats in each
measure. Rather, it is a way of notating music based on triplet groups
without the tedious bracketing that would normally require. A conductor
would beat 2/4 time. This should be kept in mind when considering
questions of tempo, phrasing, small-scale dynamics, etc.
The subject is simple only at first glance. It spans four measures. In
m. 5, the relevant return to the tonic occurs already on the downbeat. In
many later cases, however, the two ensuing 16th-notes must be included as
a passive extension or as a completion of the larger beat. In yet other
cases, these two notes form the beginning of the subsequent melodic unit.
Finally there are subject endings where the same two 16th-notes do not
remain within the harmony and therefore cannot conclude the phrase. We
are thus dealing with a subject conceived with a male (strong-beat) and
a female (weak-beat) ending. Its rhythmic pattern is simple, comprising
only 16th-notes and eighth-notes (or: triplets and whole beats). A third,
very characteristic rhythmic figure that is introduced outside the subject
(see, from m. 5 onward, the eighth-note + 16th-note groups) gives the fugue
the distinctive touch of a gigue.
The subjects pitch pattern includes inverted-mordent figures (the threenote groups in mm. 1 and 2), two larger intervals, and consecutive seconds.
While the direction is upward in mm. 1-4 and downward in mm. 4-5, the
former is interrupted by two restarts. Both follow cuts in the melodic flow
that are emphasized by the combination of rest + wedge at mm. 21 and 31.
The cuts combine with the subsequent restart to deliver an explicit message
regarding phrase structure: the subject consists of the structural pattern of
1+1+2 measures. Bachs
harmonization progresses with every half-bar
beat.
F major
441
Any decision regarding the dynamic shaping in this subject should take
the allusion to the gigue into account. In light of this virtuoso dance and
the exemplary simplicity of phrase structure that is typical of it, the three
consecutive upward thrusts (F-C, A-D, and C-F) take the lead over any
considerations of harmonic tension. The subjects outline thus consists of
three short crescendos, each surpassing the preceding one, followed by an
unbroken decrease through the octave from F to F.
This fugue features eight statements of the subject.
1
mm. 1-5
U
5
mm. 52-56
M
2
mm. 5-9
M
6
mm. 66-70
L
3
mm. 14-18
L
7
mm. 85-89
U
4
mm. 21-25
L
8
mm. 89-95
L
The subject undergoes few changes. In the tonal answer, the first leap
is modified from a fifth to a fourth. In three instances, the female ending of
the initial statement (i.e., the melodic ending on a weak beat) is substituted
by a male ending (see entries 4 and 8, mm. 25 and 89 respectively, and m.
95 where the harmonic resolution occurs on the downbeat and is followed
by a new active step in the 16th-notes.) No inversions, strettos, or parallels
of the subject are used in this fugue. In addition to the modification of the
ending, the two final statements feature further changes. The penultimate
entry fluctuates between the major and minor modes: its beginning is
surrounded by the F-minor chord but the subjects fifth note is A. This is
followed by D, which in turn gives way to D (see mm. 85-87). The same
entry also surprises with its thickened texture: mm. 86-87 contain two
chords with five and one with six parts. The final entry is expanded not
vertically but horizontally, by way of additional sequences of its head
motif (see mm. 89-93). It is further set apart from the remainder of the
composition by an accompaniment in 32nd-note runs that does not occur
anywhere else. Bach does not give this subject any regular companion or
counter-subject.
As the list of the subject statements reveals, by far the largest portion
of this fugue is taken up by passages in which the subject is absent. Only
34 of the 99 measures feature the subject, while the remaining two thirds
fall into six episodes, some of them surprisingly long.
E1 mm. 9-14
E4 mm. 56-66
E2 mm. 18-21
E5 mm. 70-85
E3 mm. 25-52
E6 mm. 95-99
442
WTC II/11
In the construction of these episodes, material derived from the subject

plays a vital role. Its third subphrase, which we will call Ms, is used
frequently, with or without the ascending half-measure upbeat. In E2 we
find three imitations of Ms without its upbeat: mm. 18-19 (M), 19-20 (U),
and 20-21 (L). In E3, Ms appears five times extended by an additional
interval of a fourth: mm. 29-30 (L), 30-31 (U), 31-32 (M), 32-33 (U), and
36-37 (U). Still in E3, Ms (now without the additional leap but with its
original upbeat) occurs four times in the lower voice: see mm. 44-52. The
last recurrence of Ms can be heard in E5, mm. 70-72, in the lower voice as
a partial sequence of the preceding subject statement.
Other characteristic episode material includes one motif and several
sequence-models.1 M1 is introduced in E1 (mm. 9-10: B-E) and subsequently fills the entire episode with its imitations. It recurs in the middle of
E3, both in its original shape (M: mm. 37-38 and U: 38-39) and in a variation with a suspension of the third note (M: mm. 39-44 and U: mm. 39-45).
This variation of M1 is sequenced twice in ascending direction in both the
middle and upper voices. At the same time, the lower voice also presents a
group of notes (see mm. 38-40: E-F) that recurs twice in rising sequences.
The three voices together thus form what we may call Model 1.
E4 contains two models, both of them short and occurring only in
U + M. In mm. 57-58, the middle voice imitates the upper voice in a threenote ascent. This combination is sequenced twice in rising direction. Similarly, the same two voices establish another short model in mm. 62-63 that
is sequenced three times in falling direction. E5 also contains two such
models. A very generous eye might even detect a relationship with those in
E4, last but not least because the two episodes share the pedal note feature.
The first model in E5 appears in mm. 72-73. It, too, is conceived as an
imitation between U + M with three notes but including a note repetition.
Three falling sequences follow here. The second model in this episode
materializes in mm. 78-79. It is more complex since the middle-voice figure
is longer. Once again ascending sequences follow. Finally there is one
figure that only ever appears in the lower voice but presents its sequences
without the support of coinciding sequences in other voices. It is thus
neither a real model nor a true motif. This sequential figure first appears in
mm. 25-26 where it is followed up to m. 281 by two ascending sequences.
Exactly the same order of events recurs in mm. 95-981.
1
As regards this distinction: it is common to call a melodic unit a motif if it recurs in more
than one place and in more than one voice. Sequence models, by contrast, are short-lived,
usually involve more than one voice, and remain unchanged in their texture.
F major
443
The fugue contains only one significant cadential close. It resembles

the final close to a remarkable degree (compare mm. 25-29 with 95-99)
and must thus be interpreted as dividing the longest episode into two
segments: E3a and E3b.
Summing up, one can distinguish three types of episodes in this fugue:
motivically determined episodes with imitation (E1 and E2), episodes built
mainly on sequence-models (E4 and E5), and cadential episodes (E3a and
E6). Then there is the composite construction of E3b with motifs and
models that, with its twenty-three measures, is almost as long as the four
initial subject entries together.
The character of this fugue is determined by its metric and rhythmic
allusion to the gigue. Like all gigues, it is lively. Most of its 16th-notes
constitute ornamental figures or runs. They should be articulated accordingly: legato for the written-out inverted mordents in the subjects first
and second subphrases but quasi legato or even leggiero for runs like the
one in the subjects third subphrase. The typical eighth- + 16th-note pattern
is decidedly bouncing, with a hint of heavy-light, heavy-light. Also derived
from the gigue is the combination of the two rhythmic patterns: a triplet
followed by an eighth- + 16th-note group (see, e.g., U: m. 8) in which the
shorter note is traditionally unaccented. Dotted eighth-notes are non legato,
with exceptions occurring only where a tie forms another pattern typical for
the gigue: the suspension followed by 16th-notes (see, e.g., M: mm. 23-24).
In such cases, the suspension acts as the first of a 16th-note group and is
therefore not separated from what follows.
The tempo of this fugue can be as lively as the 32nd-notes toward the
end allow. The score features no ornament symbols. The tempo proportion
between prelude and fugue is simple: a half-note in the prelude corresponds
with a full measure in the fugue. Approximate metronome settings: prelude
beats = 60, fugue beats (dotted quarter-notes) = 120.
The two blocks of four subject statements each make a first assessment
of the structure easy. E1 and E2 act as bridges between consecutive entries
in the first section. E4 and E5 appear in structurally analogous position in
the second round of entries but differ on account of their bass pedals. But
then there is a larger open question that has a decisive impact on performance. The conspicuous correspondence between E3a and E6 suggests
that the first section might end in m. 29. As, however, the second round of
subject entries definitely begins not earlier than in the middle of m. 52, we
find ourselves in the following predicament: A reading of this fugue as
comprising three sections whereby the central section is completely devoid
of subject statements is unorthodox both with regard to a fugue (whose
444
WTC II/11
sections are essentially defined by subject statements) and with regard to a

gigue (a dance conventionally laid out in binary form). The assumption
that Bach may have balanced a first section of 28 measures with a second
section of 70 measures is equally unconvincing and does not have
precursors in either genre, fugues or gigues. Conversely, if one decides that
the fugues initial section must be meant to include the entire third episode,
the result is a layout in approximately equal halves (51 + 48 measures). Yet
this decision comes at the expense of ignoring the cadential close in m. 29
and the analogy of E3a and E6. The color sketch shows only this latter
option, without intending to forestall the conclusion.
The gigue-like character of this fugue determines the mood to such an

extent that increases and decreases of tension play only a subordinate role,
Furthermore, the dynamic layout of this fugue is closely related to the
interpretation of the design and thus reflects back on the predicament
mentioned above. A compromise that evades the problem (without solving
it, though) is to play the first section as consisting of two halves. The first
half is determined by subject entries and episodes that connect (E1, E2) or
close (E3a) them. This portion therefore sounds in the usual touch for
gigues (mf-f, or register I). The second half of section I, comprising the
remainder of E3, would then consist exclusively of secondary material and
therefore sound like a shadow, or musing afterthought, of the preceding
(p, or register II).
Consistency would then require that the second section be interpreted
as an intertwining of the two levels. Thus E4 and E5 would be played as
afterthoughts following each subject statement separately, and therefore
soft (register II), unlike E1 and E2 with whom they correspond structurally
but not in material or, in this option, intention. Of the episodes in the
second section, only E6 will then sound on the dynamic level I.
WTC II/12 in F minor Prelude

Several features make access to this prelude easy for listeners. Its
phrases are regular and neatly contrasted with regard to texture, melodic
figures, rhythmic features, and intensity. Moreover, almost all of these
phrases recur several times in the course of the composition, so that a high
degree of familiarization is possible, particularly since both halves are
repeated. It is thus safe to state that the F-minor prelude is determined by
its material. The reluctance to say by motifs is owed to the fact that one
expects motifs in Bachs compositions to be single-voiced units used
within a polyphonic texture, with imitations, sequences, and partial developments. While the phrases here represent a variety of textures, imitative
play is not among them.
The harmonic progression in this prelude is much more largely scaled
than the design in four-measure phrases might suggest. In fact, many of the
phrases do not establish independent harmonic clauses. The overall design
encompasses mainly the familiar functions: a progressive step away from
the tonic in the first half of the prelude, and the corresponding return from
the subdominant to the tonic in the second half.
I
mm. 02-122 tonic to tonic relative (F minor to A major)
II
mm. 122-242 tonic relative confirmed
mm. 242-282 tonic relative confirmed again
III
mm. 282-402 modulation to the subdominant B minor
IV
mm. 402-482 return to the tonic with imperfect cadence
V
mm. 482-602 interrupted cadence in the home key
mm. 602-622 plagal cadence in the home key
mm. 622-662
home key confirmed
mm. 662-702
home key confirmed again
The rhythmic pattern is based on a regular use of eighth-notes and
16th-notes and can thus be called simple. The pitch pattern, on the other
hand, contains much variation in the different phrases: there are stepwise
passages, many broken-chord figures, as well as large intervals, the latter
occurring both in the 16th-notes and in the eighth-notes.
The character must be determined as a borderline case. Rhythm and
intervals point to a lively character, but the abundant use of appoggiaturas
indicates an emotional intensity that places a restraint on the tempo.
445
446
WTC II/12
Articulation and touch must be mindful of the various components of

the material: A neutral tone quality is appropriate for all quarter-notes that
represent a cadential bass or should be distinguished from rhythmically
interlocking yet melodically more relevant note-groups in another voice and
for all eighth-notes followed by rests in complementary-rhythm patterns.1
Non legato with a more singing touch is required for all quarter-notes and
eighth-notes that form melodic lines. Legato with high-intensity touch is
indispensable in all appoggiatura-resolution pairs and also applies in the
closing formulas.2 Quasi legato with a light touch is a good choice for the
16th-notes in the toccata-style complementary patterns (see mm. 4-8 etc.).
The score contains five ornaments, all of them mordents. Those in mm.
40, 42, and 69 are approached in stepwise motion and therefore begin on
the main note. A simple shake with three notes is sufficient here. The ones
in mm. 45 and 46 are preceded by a large leap. They therefore begin on the
upper neighbor note and feature a four-note shake.
The F-minor prelude is built not on single-voiced units in a polyphonic
texture but from homophonic cells that develop into short phrases. Each is
at first presented in its own unique texture before there is a vivid exchange
and interplay of accompanying features. (Although these cells are thus not
motifs, the customary M will be used for easy reference.)
M1 is presented in the first four measures. The lowest of its three voices
features quarter-notes in cadential steps. The upper and middle voices run
in parallels. Their core consists of an appoggiatura-resolution pair in
eighth-notes preceded by an upbeat in the shape of an anticipation of the
appoggiatura. Owing to the melodic parallel and to the appoggiaturas, each
of the one-measure patterns is high in emotional intensity. The entire
four-measure motif forms a dynamic curve. The tension increases to m. 31
where the subdominant harmony coincides with the sudden leap in the
pitch level of the upper voice. The ensuing decrease is expressed, in the
lower voice by the chromatic transition to the dominant, and in the two
higher voices by a gradual descent.
M2 presents a toccata-pattern. The left-hand strong beats are complemented by three 16th-notes in the right hand. The pitches of both voices
interlock, giving the impression of a single layer in broken-chord figures.
1
Neutral non legato: mm. 1-4 (L), 4-8 (M), 8-16, 20-24, 28-32 (L), 52-56 (U+M), 56-58 (M).
Appoggiaturas requiring intense legato may come in parallels (U+M: mm. 1-4, 9-16, 28,
29-32, 40, and 57), grow single-voiced out of a suspension (M: m. 34 and U: mm. 36),
appear single-voiced without a marking (U: mm. 37, 38, 45, and 46, L: 49, 51, and 52, and
U+M: m. 70) or slurred (U: mm. 42 and 44). For a closing formula in legato see U: m. 39.
F minor
447
There could hardly be a more obvious contrast between the textures of M1

and M2. The melodic content of M2 is negligible: the left-hand notes owe
their melodic value more to their metric position on strong beats than to the
consistency of the line they build. This is not surprising since they are,
after all, primarily parts of broken chords. This lack of melodic expression
is reflected in the somewhat subdued intensity of this cell and in its lack of
direction. Harmonically, these four measures make no progress but extend
the C-major chord reached in the middle of m. 4 in an ornamental fashion.
The broken-chord figures seem to circle without a goal (see mm. 5-62 ,
which recur identically in mm. 7-82). Correspondingly, the dynamic tension
is as if suspended. Neither build-up nor relaxation is called for.
M3 appears in mm. 202-242. The texture displays once again three
voices, each of them essentially independent of the other. What makes the
interplay nevertheless so simple is the regular rhythm and the one-measure
sequential pattern. Each of these one-measure units contains a rising step
in quarter-notes in the lower voice, a syncopated falling third ending on a
weak beat in the middle voice, and a curve made up of rising and falling
broken-chord 16th-notes in the upper voice. Harmonically, each middle
beat represents an inverted seventh-chord followed in the ensuing measure
not by its resolution but by another extended triad, now in root position
(mm. 202-241: A7-D9, g7-c9, f7-B9, E7-A9). The entire phrase conveys the
impression of a long release of tension.
The next four measures seem different at first glance. They can,
however, be regarded as a combination of M2 and M3. The complementary rhythm of the lower and middle voices appears in a different guise
from M3, but the broken-chord figures in the upper voice resemble those
in M2. Harmonically, the phrase is distinguished by a high content of
accidentals (see particularly the F and C in mm. 26-27). Although the
melodic direction, as in M3, is descending, this variant expresses an
increase to m. 27 followed only then by a relaxation.
M4 only emerges in mm. 402-422. The upper-voice contour consists of
a two-measure unit, the longest unit in this prelude and also the one that is
melodically the most eloquent. The dynamic shape shows a curve in which
the climax falls on the appoggiatura. The middle and lower voices are
combined here in the hidden two-part structure of the left-hand part. The
melodic lower voice contrasts the eloquent upper-voice unit with a descent
in eighth-notes (C-B-A-G-F-E), while the middle-voice part of the
hidden two-part structure adds a pedal on D followed by a broken G-minor
chord as harmonic backdrop. M4 is followed by one complete and two
varied partial sequences, and complemented by an imperfect cadence.
448
WTC II/12
The following table lists the preludes material in order to show what
correspondences hide behind the many variants. Strongly diverging
variants are marked with an asterisk, developments with the plus sign.
mm. 02-122 M1, M2, M1..............
inserted:
mm. 122-202 M1, M2
..............
inserted:
mm. 202-282 M3, M3/M2 ..............
mm. 282-402
mm. 402-482
mm. 482-562
mm. 562-622
mm. 622-70
M1, M2, M1*

M4
M1*, M2*
M1+
M3, M3/M2
f/p/f
mf
f/p
f
mf
WTC II/12 in F minor Fugue

The material of the F-minor fugue has a dance-like character. Bach
achieves this primarily through the persuasive rhythm and pitch pattern in
the subjects head: upbeat + triple repetition, upbeat + triple repetition.
These two immediately remembered initial measures are complemented by
two measures of unobtrusive running notes. The end of the subject is
reached, as can be expected of such a regular phrase, after precisely four
measures. The only irregularity arises in the two possible endings Bach
invents for this subject. There is the strong-beat ending (a male ending in
the terminology taken from Greek poetic meter) in which the melodic line
concludes with a return to the tonic harmony on the downbeat of a
statements fourth measure (see, e.g., in m. 32). Much more frequent is the
weak-beat (or female) ending, which extends this tonic through two
more eighth-notes and thus achieves a more perfect complement to the
subjects 1/8 upbeat.
While the first upbeat + triple-repetition group should be separated from
its varied sequence by phrasing, the final note of the sequence (the third E
in m. 2) doubles as the beginning of the running notes and does not allow
for another cut. The subject thus consists of a short first subphrase and a
much longer second one. The rhythmic pattern, in the subject itself as well
as in the fugue as a whole, is simple, consisting primarily of eighth- and
16th-notes. In the pitch pattern, the most noteworthy characteristics of this
subject are the leap and broken chord respectively linking the two upbeats
to their subsequent downbeats (F-C = perfect fifth, D-B-E = extract of C9
chord) and the triple note repetition. The remaining 16th-notes move
almost exclusively in stepwise motion.
F minor
449
The subjects harmonic layout is also straightforward: the tonic on the

downbeats of mm. 1 and 4 surrounds the dominant in mm. 2 and 3. The
subdominant appears in a metrically weak position (m. 1, last eighth-note)
and thus does not gain momentum. Yet as it marks the only fast harmonic
change in this subject (with only one eighth-note between its appearance
and the change to the dominant), it endows the beginning of the second
subphrase with a sense
of slightly heightened
urgency. The dynamic
development reflects
these features. In the short initial subphrase, the downbeat, prepared by the
bouncing upbeat, is distinctly accented. The relationship between upbeat
and downbeat is enhanced in the sequence, where the first E is even more
prominent than the first F had been. While the note repetition in the first
subphrase brings about an immediate release, the length of the second
subphrase now requires a gradual diminuendo. (Emotional waves in m. 3
should be avoided so as not to blur the subjects structure and character.)
The F-minor fugue comprises nine subject statements.
1. mm. 0-4 U
4. mm. 24-28 U
7. mm. 50-54 M
2. mm. 4-8 M
5. mm. 28-32 M
8. mm. 71-75 U
3. mm. 11-15 L
6. mm. 40-44 L
9. mm. 74-78 M
Apart from the interval adjustment in the answer (as is common in

fugues whose subject begins with a fifth, this interval is converted into a
fourth to retain the tonality) and the above-mentioned female and male
endings, the subject remains unchanged. No inversions occur, and the only
overlap of consecutive statements (mm. 74-75) is so minimal that it is not
perceived as a stretto. Moreover, Bach has not invented a single countersubject for this fugue. There are six subject-free passages.
E1 mm. 82-112
E3 mm. 321-402
E5 mm. 541-712
E2 mm. 152-242
E4 mm. 442-502
E6 mm. 781-85
Four of these episodes are subdivided. We should distinguish E2a/b
(mm. 152-171-242), E3a/b (mm. 321-331-40), E4a/b (mm. 442-462-502), and
E5a/b (mm. 541-661-712). The subjects tail figures in episodes that begin
as subject prolongations, as happens in E2a (descending sequences), E3a
(twofold imitation), and E4a (a rising two-measure sequence). The
subjects head with extending sequences is used as a motif in E5. Genuine
episode material is limited to one sequence model and one small motif.
450
WTC II/12
The motif is first heard with sequences in the lower voice of E4b and taken
up once at the outset of E5a, with an imitation. The sequence model, introduced in mm. 17-18, quotes the subjects note repetition in double thirds
with an ornamental extension in the upper voice against quasi-melodic
curves in 16th- and eighth-notes in the lower voice. E2b continues with
two descending sequences, one faithful, the other varied and extended. In
the structurally corresponding E3b the upper and middle voices are inverted
and the extension of the second sequence is slightly modified. Another
analogous episode is E6, while E5b is shortened in its second sequence.
The relationships between the episodes are manifold and revealing for
the structural layout. Note particularly the four larger passages that correspond with one another while being distinguished from their surroundings
by texture owing to the three-part model: E2b . E3b . E5b .E6. All have
harmonically concluding endings. These are thematically open in mm. 24
and 71 but characterized by closing formulas in mm. 39-40 and 84-85.
Performers may wish to consider setting the distinct sequence models in
the four corresponding blocks apart from the remainder of the fugue by
choosing a different color.3
E2b and E3b are prepared by analogous episode segments that extend
a preceding entry, whereas E5a consists of two corresponding halves.
Dynamically, E5a is the only episode that represents truly active increases
in the ascending sequences of the subject head. (One might have a point in
arguing that these make up for the missing third entry that the balance with
section I would require). Only E1 and E4, which are not so distinct in their
material and structure, serve as links between two subject statements
within a section. Both begin with a slight increase before giving way to
relaxation, while all other episodes represent a decreasing tendency.
The thematic leaps and broken chords as well as the simplicity of the
rhythm indicate a rather lively basic character. The tempo should be fast
enough to create the effect of a true 2/4 meter, with only one strong beat in
each measure, and to guarantee the light character in the 16th-notes. The
relative tempo of the prelude to the fugue needs careful planning. A simple
proportion of quarter-note = quarter-note (and thus: measure = measure)
may sound dull. A complex proportion, while more difficult to imagine,
gives more satisfactory results for each of the two pieces. It implies using
3
Having the second manual of a harpsichord in mind may help. On modern pianos, one can
shade the fingers soft touch in these segments by depressing the left pedal. This gives an
effect that is both convincing in terms of the structure of this fugue and in keeping with
Baroque performance practice. Exact timing of the left-foot pedal is, of course, essential.
F minor
451
a note value that does not literally occur, the triplet eighth-note, as a base:
one triplet eighth-note in prelude tempo corresponds with one eighth-note
in the fugue. (Approximate metronome settings: prelude beats = 66, fugue
beats = 100.) The appropriate articulation includes non legato for the
eighth-notes and a crisp legato for the 16th-notes. The non legato may
sound bouncing in the subject itself, slightly less assertive in the secondary
material, and particularly gentle in the sequence models.
The only ornament suggested in the score appears, in parentheses, on
the first down-beat note of the first subject entry. If played, it must be
added to all other complete entries. An ornament that is not indicated even
in brackets but which almost all performers will intuitively add is the
(four-note) cadential mordent on the dotted G in m. 84.
The structure of this fugue is very clear, as can already be seen from
the outline and function of the episodes. The four corresponding episode
segments each conclude a section. As they end with different degrees of
melodic conclusiveness, the impression arises that sections I+II and
sections III+IV form the two larger blocks of a binary design. This concept
is supported by the observation that the first and third sections contain
bridging episodes between their entries, while the second and fourth
sections do not. The following table visualizes the correspondences.
section III:
section I:
subject in L + E4 (linking)
subject in U, M, + E1 (linking)
subject in L + E2a (short opening)
subject in M + E5a (long opening)
E2b (sequence model;
E5b (sequence model;
no closing formula)
no closing formula)
section II:
section IV:
subject in U, M
subject in U, M
E3a (short opening)
E3b (sequence model +

E6 (sequence model +
closing formula)
closing formula)
The harmonic plan supports the binary layout in that both the first and
third sections set out from the tonic. The developments inside each half of
the fugue, however, are different: Section I contains three statements in the
field of the tonic, after which E2 modulates to E major (the dominant of
the tonic relative A major). The second section is held in the major mode,
with its two entries in A and E major respectively. E3a then modulates to
C minor, and E3b closes in C major (the dominant). The two statements of
the third section are both on the tonic, after which E5 ends on an F7 chord
(the dominant of the subdominant B minor). The fourth section features
entries on the subdominant and tonic and thus closes the circle.
452
WTC II/12
The dynamic development from one entry to the next is not the foremost intent in this fugue. In section I, the tension rises throughout the three
entries (the dynamic curve in the bridging episodes constituting no interruption but just a suspension) and then diminishes gradually through the
concluding episode E2. Correspondingly in section III, the tension
increases through two entries (suspended during E4) and again in the
sequences of the subject head, before it subsides gradually throughout the
concluding E5. In both cases, the harmonic ending on the dominant of the
key in which the subsequent section will set out creates further release and
thus mollifies the beginning of sections II and IV. The reduced texture
supports this, as the entry in mm. 24-28 appears manifestly (and that in
mm. 71-75 practically) in two-part setting. In both cases, the final subject
statement is placed in the middle voice, the most subdued of the three
possible positions. Any growth in tension that might be brought forward
by the increase in texture is thus annihilated, and sections II and IV remain
much less outgoing than sections I and III. The concluding episodes E3
and E6 end both halves of the fugue on a soft note.
WTC II/13 in F major Prelude

This prelude is determined by several rhythmic features. There is not a
single measure without at least one dotted eighth-note pattern, and in many
measures these patterns are present on each of the three beats. The aural
impression is slightly different from measure to measure only because the
dotted note may sound either shortened by an appoggiatura or ornamented.
Moreover, the triple time in this prelude displays a particularly large
number of syncopations on the second beat in the measure (see mm. 1, 2,
12-14, etc.). This is reminiscent of the typical sarabande rhythm. A third
feature regards what could be called the large-scale rhythm, the longand-short patterns created by the phrases. A first glance at only the most
conspicuous elements reveals that mm. 1-16 (in other pieces, particularly
in sarabandes, commonly structured regularly as 4 + 4 + 4 + 4) are laid out
in an irregular grouping of 3 + 3 + 5 + 5 measures. The F-major prelude
can therefore be regarded as determined, both on the smaller and the larger
scale, by rhythmic features.
The first harmonic progression concludes at m. 41 with a return to the
tonic. This harmonic conclusion follows the end of a three-measure motif
and marks the beginning of another. It should therefore not be regarded as
relevant for the overall structure. The subsequent modulation to the dominant is completed at m. 71, its close occurring again after the end of one
motif and on the beginning of the next. A cursory glance at the remainder
of the prelude shows that all small-scale cadences coincide with motifs and
are therefore not to be relied on for an understanding of the overall design.
For a first overview it will be necessary to identify recurrences of the first
motif and explicit cadential formulas. M1 (mm. 1-41) is restated in mm.
17-201, 20-231, and 57-601, and developed in mm. 16, 42-451, and 65-681.
Closing formulas occur twice, in mm. 44-451 and 67-681. The prelude is
thus composed of five sections each comprising several phrases.
section
I
II
III
IV
V
mm.
1-16
17-44
45-56
57-67
68-75
phrase structure
phrase length
3+3+5+5
16
3 + 3 + 3 + 3 + 5 + 4 + 4 + 3 28
4+4+4
12
3+5+3
11
5+3
8
453
454
WTC II/13
There are two structurally analogous passages in this prelude. They

underline the corresponding beginnings of the first and the forth sections:
mm. 1-3 . 57-59 (same key, hardly varied)
mm. 7-13 . 60-66 (transposed from V to I)
The tempo of this prelude is slowly swinging: slow enough to accommodate 64th-notes without any haste. Articulation and touch should be
considered very carefully, with regard to each of the rhythmic and melodic
features:
Legato in melodically intense touch and with expressive dynamic
shading should be adopted for all notes in thematic lines characterized primarily by stepwise motion (U: mm. 1-2, L: mm. 17-192, etc.).
Legato in neutral touch and low intensity is appropriate for the
regular 16th-note figures serving as an accompaniment in mm.
4-11, 17-41, 45-56, 60-64, and 68-72.
Quasi legato should be attempted in dotted-note groups that are
designed in hidden two-part structure (e.g., L: mm. 4-6). Within
this texture, the harmonic background, featured here by the repeated
indirect pedal note C, sounds most convincing if played with the
idea of a rhythmically vibrating sustained note in mind. The melodic
part of the texture (F-A-E-G-D-F-C) deserves warmth and,
therefore, weighted playing in each note.
non legato in melodically intense touch and with expressive dynamic
shaping applies for all the other dotted-note groups (L: mm. 1-3,
7-11, etc.). The degree of detachment can be increased in consecutive leaps and broken chords, and the melodic intensity may give
way to a more neutral color in cadential-bass patterns.1
The score comprises a number of ornaments: grace-notes, mordents,
inverted mordents, and trills. Grace-notes appear in mm. 1, 15, 41, 43, 44,
and 67. Where they precede dotted-note values, they are played as one
third of the main note, with two thirds left for the resolution. In the case of
very long notes signifying a momentary retreat from the melodic field, the
appoggiatura may even be resolved as early as harmonically possible.
Note that appoggiaturas were regarded as part of the melodic flow and
therefore not meant to be played in note values otherwise not used in the
1
Bachs dotted notes sometimes confuse players. In the early 18th century, the dot after the
note did not yet have the rigid meaning it was to acquire later, i.e., to lengthen the note by
exactly half of its value. In this era it meant a little longer, the following notes determining
in each case how much extension is implied. Whenever the actual value of the dot is only
a 32nd-note (as in mm. 1 and 12-16), the same legato applies between the prolonged note
and its complement as would be used between other melodic 32nd-notes in this piece.
F major
455
piece. The grace-notes in mm. 15, 41, 43, and 441 should therefore be
played as 16th-notes followed by eighth-note resolutions. In m. 12 the
appoggiatura precedes a dotted quarter-note. Most performers would argue
that the melodically relevant portion of this note, here as elsewhere in the
piece, is the rhythmically predominant dotted eighth-note, and therefore
render the grace-note as a 16th-note. Reading the almost identical closing
formulas of mm. 44 and 67, by contrast, is unambiguous since a specific
rhythmic rendition is established for this variant of the dosido figure:
| ).
Inverted mordents appear, as so often in Bachs music, on notes that
are thereby specifically enhanced; see, e.g., m. 12. As the figure presented
in this measure is sequenced twice in mm. 13 and 14, it is recommended to
transfer the inverted mordent to the corresponding notes. By the same
token, the ornament may also be added on the syncopations in the analogous phrases, in mm. 65 and 66. In m. 22, by contrast, the printed inverted
mordent seems less appropriate: the note it embellishes is not melodically
significant and has remained unornamented in identical circumstances just
before (compare U: m. 22 with L: m. 19). The two mordents in mm. 28 and
74 adorn typical closing formulas and begin on the main note. The first
contains only one three-note shake while the ornament in the final cadence,
appearing in ritardando, may be played with five notes.
Trills occur in mm. 26-27, 29-32, 38, 44, 52, 56, and 67. Most of them
begin regularly on the upper note; only the three trills that are preceded by
appoggiaturas (mm. 44 and 67) or approached stepwise (m. 52) begin on
the (prolonged) main note. The speed of the shake depends on ones
interpretation of the fast note values in the piece. Performers who feel that
all 32nd-notes represent written-out ornaments may decide to shake in the
same values. There is, however, good reason to regard the 32nd-notes as
fully weighted melodic notes; this is supported particularly in figures like
the one in mm. 1-2 etc. Another reason for reconsidering is the fact that the
final cadential formula features written-out 64th-notes, which should not
sound faster than the trill notes. The suffixes must in any case be rendered as
spelled out by Bach and thus as slower than the 64th-note shakes, a
common practice particularly in dances and other non-polyphonic music of
the era. The examples
show possible executions
for the trills in m. 26
(with m. 27 accordingly),
m. 32, and m. 44 (with
m. 67 accordingly).
m. 26
456
WTC II/13
Bach invents three

melodic phrases as well
as a standard accompaniment pattern. The
melodic components
recur both with small
m. 32
variation and in freely
developed format. They
occasionally even swap
their accompaniment
patterns. The result is
pleasant for the listener
m. 44
since there are always
enough familiar features to be recognized but it never sounds quite the
same. Even when a kind of recapitulation appears, the principle of subtle
variationbe it concerning details or the order of eventsis maintained.
The standard accompaniment pattern2 is introduced in the right-hand
part of m. 4. In the attempt to describe it in such a way that all further
variations can easily be recognized, one can state the following features:
The two initial 16th-notes may vary in each occurrence depending (among
other reasons) on the connection to the preceding measure. The third to
ninth 16th-notes move within the confines of a broken chord in which they
represent steps 1-2-3-1 followed by the falling chord. Lastly, the three final
16th-notes are joined to the previously described triad by a variety of
intervals, but always moving in a descending line.
The principal melodic phrase extends over a little more than three
measures (see mm. 1-41) and is characterized by a line that, although interrupted by rests, features neither sequence nor repetition. Its most striking
feature is the syncopation on the second beat. Its rhythmic hallmark, the
dotted eighth-note on the downbeat, is complemented in m. 1 by three
notes, in m. 2 by two, and in m. 3 by one note only: a gradual shift from a
very smooth to a more and more pronounced rhythm. Dynamically the first
subphrase presents a simple diminuendo, the second climaxes either on the
downbeat (in a lyrical interpretation) or on the syncopation (in a rendering
emphasizing the metrical aspect, i.e., the allusion to the sarabande), and the
third consists of an upbeat with subsequent relaxation. The characteristic
accompaniment of this phrase exerts a secondary melodic power.
2
For different shapes compare mm. 4, 18 and 45.
F major
457
The second melodic phrase answers in the left-hand part of mm. 4-71.
It is conceived in hidden two-part structure, with a background featuring
a repeated indirect pedal on C that briefly gives way to the leading-note at
the very end of the motif, and a foreground manifested in the downbeat
notes consisting of a descending line (F-E-D-C) that is paralleled in
thirds on beat 3 of each measure. The dynamic equivalent is a diminuendo
throughout the melodic notes, while the background is best kept very soft
and evenly colored throughout, imitating as much as possible the effect of
a sustained pedal note. The right hand, in this first statement of the phrase
as always in this prelude, realizes the standard accompaniment pattern.
The third melodic phrase emerges in the left-hand part of mm. 7-8. Its
prominent feature is the combination of a descending broken V7 chord with
the harmonic resolution, represented by steps 8-5-1 also in a descending
pattern. Both the harmonic and the melodic shape point to the seventh (see,
e.g., at m. 72) as the climax of this two-measure motif. The model is
sequenced in mm. 9-10 and partially again in mm. 11-12. Together with its
two sequences it thus triggers a stepwise modulation (m. 8: F major;
m. 10: G major; m. 12: A minor).
The preludes first section begins with the original statements of the
three phrases followed by a free development of the characteristic rhythm
in which dotted-note groups with one 16th-note and three 32nd-note complements are set against each other in a quasi-contrapuntal texture (mm.
12-16). The standard accompaniment pattern is suspended during this
development. The dynamic shape is determined by ascending peak note
lines in both voices (mm. 121, 131, 141, L: A, B, C; mm. 122, 132, 142,
U: F, G, A). A reminiscence of the accompaniment pattern in m. 14,
descending lines in m. 15, and a cadential close in mm. 16-171 complete
the tension curve.
In sections II and III Bach develops the same material with multiple
variants and new combination. In section II he concludes with a cadential
formula in D minor (see mm. 44-451). Section III is rounded off with four
measures of gentle transition in which a parallel version of the standard
accompaniment pattern is interspersed with reminiscences of the very first
development in mm. 12-16.
Section IV shows traces of a recapitulation: mm. 57-601 review the
principal phrase with its original accompaniment and an almost identical
melodic line, mm. 60-651 are a transposition of the third phrase, and mm.
65-66 recall mm. 12-13, equally transposed. The section concludes with a
closing formula similar to the one heard at the end of section III.
458
WTC II/13
The short final section completes the recapitulation of the thematic

material with the one element Bach had omitted in the previous section:
the second melodic phrase appears in a generously extended version of
five measures and leads into a lavishly embellished cadential formula in
four-part homophonic texture.
WTC II/13 in F major - Fugue

This three-part fugue, counting 85 measures in the score, develops
from a very unusual subject. The beginning on the leading-note is certainly
unique, as is the flattening of this leading-note in its next appearance (see
m. 21: E), which seems to call a tonality into question that has not yet been
fully established. The launch with a half-measure upbeat is typical for the
gavotte with which this fugue shares the time signature. Yet while compositions featuring a metric organization with repeated extensive upbeats
create a strong expectation for a phrase ending on a downbeat, none of the
strong beats before the entrance of the answer offers itself as a convincing
conclusion. Instead, the arrival on the tonic occurs in the weak second half
of m. 41 where Bach postpones the downward step from the fourth scale
degree to the third (B-A) by inserting an appoggiatura that resolves in
unaccented position on the third eighth-note. The F on the middle beat of
m. 4, although melodically qualified as the keynote, cannot conclude the
subject because by then, the tonic harmony has already been abandoned:
this F, together with the B on which the upper voice enters, belongs to a
V7 chord in C major. Should further proof be needed, it should be noted
that Bach gives six of the eleven subject statements in this fugue an ending
concluding with the appoggiatura-resolution.
Having determined where the subject ends, one immediately faces the
next question. As a comparison with later statements proves, the characteristic counter-subject motif is launched on the fourth quarter-note of the
measure; see the upbeat in the sigh motif group F | F-E. This leaves
two notes between the subjects conclusion and the counter-subjects
beginning. The function of these notes is to smooth the transition and at the
same time to serve as a harmonic support of the answers leading-note
beginning.3
3
Similar harmonic fillers can be found in several other fugues where a delayed entry of
the counter-subject would otherwise leave the beginning of the answer unaccompanied.
F major
459
The subject features five different note values: 16th-notes, eighthnotes, quarter-notes, dotted quarter-notes, and a half-note (this is the actual
value of the upbeat, hidden behind the trill and its written-out suffix). The
pitch pattern displays mainly stepwise motion, with only a few skips that
are confined to m. 3. The phrase structure, however, is anything but simple.
Whether or not all the skips in m. 3 are in fact interval leaps in a melodic
context, or whether one of them marks the line between two subphrases, is
a question of eminent importance for the performance as well as for the
understanding of the piece. As so often, Bachs harmonization delivers the
clue. It consists of four essentially analogous harmonic gestures in a metric
organization of perfect regularity. Each weak beat represents a V7 chord
that resolves on the following downbeat. In the initial upbeat as well as in
m. 4, this V7 is instituted on the middle beat, while in mm. 1 and 2 it only
materializes on the final quarter-note and eighth-note respectively.
7
7
7
Guided by the harmonic pattern one detects that the melodic line contains an equivalent structure: a triple sequence followed by a sigh motif,
all of it wrapped in linking and embellishing notes:
The value of these insights for decisions regarding dynamic shaping is

nevertheless limited since Bachs rich surface structure knits the second
and third sequences so closely together that a caesura between them would
sound artificial. For performers, the subject thus has three relevant
subphrases: the first consists of the trill on the unusual seventh-degree
beginning and its resolution onto the tonic, the second comprises two
measures from C in m. 1 to G in m. 3, and the third features the sigh
motif with upbeat. In the first subphrase, the ornamented leading-note
captures all the tension. The subject thus begins with a diminuendo. In the
longer second subphrase the flattened seventh degree remains comparably
subdued while the tension rises toward the subdominant representative (the
460
WTC II/13
D in m. 31), after which the rising phrase ending relaxes gradually toward
a relatively soft G. Finally, the dynamic shaping of the sigh motif
follows the established pattern of upbeat / heavylight.
The fugue comprises eleven entries of the subject.
1 mm. 0-4 M
5 mm. 32-36 L
9 mm. 64-68 L
2 mm. 4-8 U
6 mm. 36-40 M
10 mm. 70-74 M
3 mm. 8-12 L
7 mm. 40-44 U
11 mm. 76-80 U
4 mm. 20-24 U
8 mm. 52-56 M
The subject undergoes few modifications. The answer does not need
any interval adjustment, and inversion, stretto, or parallel are not used. The
only variation occurs where the trill on the first subject note is replaced by
a written-out ornament in a similar pitch pattern but a different rhythm (see
m. 20). The same variation is even more disconcerting in mm. 70-71 where
the original trilled ascent appears as a parallel in the lower voice.
Bach invented two counter-subjects for this fugue. CS1, introduced
against the answer (M: mm. 4-8), has particularly close ties to the subject
since its initial sigh motif with upbeat appears like an echo of the subjects
final subphrase. This figure is then sequenced in chromatic descent, complemented by a third subphrase that culminates on a long trill before
concluding with its resolution on the following downbeat. Once introduced
the first counter-subject remains a faithful companion of the subject.
CS2 enters against the third subject statement (M: mm. 8-12). Like
CS1 it begins with a one-measure unit that is sequenced a whole-tone
lower. The twofold figure leads to a climax on the syncopation, after which
it resolves downward onto the tonic. The second counter-subject can be
found several times with small modifications (in mm. 32-36 and 64-68 it
swaps endings with CS1). In other instances variations are so significant
(see M: mm. 20-24 and U: mm. 37-40) that they make recognition quite
difficult, particularly for listeners.
F major
461
The fugue includes seven subject-free passages.4

E1 mm. 12*-202
E3 mm. 44**-522
E6 mm. 742 -762
E2 mm. 24*-322
E4 mm. 562 -642
E7 mm. 802 -84
E5 mm. 682 -702
Two categories of material are used in the subject-free passages of this
fugue: subject-derived motifs with lower-voice accompaniment occur in
E2, E4, E5, E6, and E7, while independent episode motifs in a highly
complex imitative pattern are found in E1 and E3. The only episode
segment not to answer to either of these descriptions is the final twomeasure cadential close. E2 introduces the four final notes of the subject as
a partial (descending) sequence that follows the subject in the upper voice
(mm. 24-25). After another sequence the motif and its sequence are
imitated first in the middle voice (mm. 26-28), then in the upper voice
(mm. 28-30), and again in the middle voice (mm. 30-32). The parallel to
the sigh motif has to be colored very softly or else the imitation pattern
will easily get lost. The lower voice accompanies this imitative pattern
with a two-measure figure whose peak notes build sequences from the
second half of the harmonic minor scale (see mm. 24-36: A-B-C-D,
D-E-F-G, B-C-D-E, E-F-G-A). E4 is an exact transposition of
E2, while E6 and the first half of E7 (mm. 80-82) only quote one quarter
(with a free version of the lower voice figure), and E5 recalls this segment
with inverted voices.
The two remaining episodes are made up of material that is unrelated
to both the subject and its counter-subjects, and is intricately polyphonic.
These episodes thus create a distinct color contrast to the primary material.
They encompass three motifs. M1 (U: mm. 12-142) begins with an upbeat
to an ornamented falling D-minor triad that is sequenced with a varied
ending one note lower. The motif is imitated, with or without its upbeat,
three more times in E1 and taken up with four analogous entries in E3. M2
begins, contrapuntally to M1, with a syncopated half-note followed by an
inverted-mordent figure and two cadential notes (if in the lower voice) or
another tied note (if in the upper or middle voice). Like M1 it is imitated
three times in E1 and taken up another four times in E3. M3, presenting a
further counterpoint, features a quarter-note upbeat leading to a three-note
ascent in half-notes.
4
The episodes marked with one asterisk begin as follows: in the voice carrying CS1: after
the downbeat; in the voice carrying the subject: after the resolution on the third eighth-note.
At the double asterisk, the episode begins after the downbeat in the middle voice but after
the middle beat in the upper and lower voices.
462
WTC II/13
The role played by the episodes in the fugues dynamic design follows
from their different content. The color contrast in the two episodes characterized by independent motifs was already mentioned. The descending
sequences in the two longer episodes determined by sequences from the
subjects tail as well as in the shorter E6 imply a gradual diminuendo. By
contrast, their two other relatives, E5 and E7, quote the sequences in
ascending direction and thus build up tension. The cadential close of the
final measures features the rhythmic pattern and melodic formula typical
for gavottes, traits one associates not with a release but with a triumphant
close, thus concluding the fugue on a confident note.
The highly emotional quality of the subject in conjunction with the
complex rhythmic pattern determines the basic character as rather calm. In
the alla breve meter, the half-notes are gently flowing but far from hasty.
The tempo proportion between prelude and fugue can be simple as the
preludes rhythmic features guard against monotony: thus a quarter-note in
the prelude becomes a half-note in the fugue. (Approximate metronome
settings: all beats = 56.)
The articulation that corresponds with this character demands legato
for all notes in melodic context. Non legato occurs only where quarternotes form cadential-bass patterns (L: mm. 55-56, 83-84) or consecutive
leaps (L: mm. 13-14, 19-20, and 45-46, in M2). There are, however,
passages in this fugue that require a different approach. The five episodes
that develop the subjects tail display features that would normally indicate
a rather lively character: simple rhythmic patterns with only eighth- and
quarter-notes shaping lines with frequent large intervals. The leaps in
quarter-notes result from Bachs decision to use in these episodes not just
the subjects last subphrase but also the sixth that precedes ita leap that,
in the subject, divides two subphrases and thus is not really conceived as
an interval but becomes one here. The accompanying eighth-notes feature
regular written-out inverted mordent figures, another indication of lively
character. While the tempo in these episodes may certainly not divert from
that chosen for the subject-dominated passages, their material invites a
different touch quality. In the quarter-notes, lightly bouncing non legato is
interrupted only for the appoggiatura-resolution pairs; in the eighth-notes,
quasi legato alternates with tight legato in the ornamental figures.
To sum up this point: the F-major fugue thus encompasses three fairly
different colors: one is reserved for the subject-dominated passages, a
second color applies in the motivically determined episodes E1 and E3,
while a third distinguishes the other five episodes.
F major
463
Ornaments occur in several instances. The trill on the initial subject

note begins on the main note in mm. 1, 4, 8, 52, and 76, probably also in
mm. 36 and 40, but on the upper auxiliary in mm. 32 and 64.5 It shakes in
16th-notes and ends in a suffix as indicated by the composer. Whenever it is
launched from the main note, this initial note is held for the duration of an
eighth-note. The mordent on the penultimate subject note begins on the
upper auxiliary and includes four notes. While the rhythm in a mordent is
not necessarily fixed, it helps to imagine this ornament as placing a 16thnote triplet C-B-C against an eighth-note.6 The trill at the end of the first
counter-subject is identical with the one launching the subject: beginning
on the main (eighth-)note, it continues with six 16th-notes including the
suffix (C-B-C-B-A-B). It should be transferred to the end of the CS1
statements in mm. 11 (E), 23 (E), 35 (B), 39 (E), 43 (C), 67 (E), and
79 (E). A cadential mordent may be added on the middle beat of m. 83.
This kind of closing formulas would customarily have been graced with an
ornament beginning on the upper note and containing two shakes. By
contrast, the mordent indicated only in one instance for the first appoggiatura of CS1 (L: m. 37) seems odd and sounds misleading for listeners who
will assume it to be a sequence of the subject ending. It may be preferable
to omit this ornament for the sake of both clarity and consistency.
The fugues layout is very regular. Of the three sections, the first two
(mm. 1-32 and 32-64) correspond with one another in their details: each
comprises three consecutive entries, one motivically determined episode,
5
One of the exceptions to the rule for trills in 18th-century music generally beginning on the
upper note concerns notes opening a new phrase. This applies in all cases where the voice that
presents the subject sets in newly (at the beginning of the piece or after a rest). Even in mm.
36 and 40, the considerable leap preceding the subject entry supports the interpretation of
phrase end and new phrase beginning, while the lower-voice figure found in mm. 31-32 and
63-64 seems to lead into the subject entry, so that phrasing here would sound artificial.
6
While Bach, apparently out of consideration for his contemporaries, omits this ornament in
the lower-voice entries and in the context of large intervals (see L: mm. 12, 36, 68, and M: m.
56), it is preferable to retain the mordent in each statement. The same holds true for the five
episodes built on the subjects tail. Bach signals ornamentation only in comfortable positions but not in double sixths and in the lower voice. Given todays standard of pianistic
technique, such inhibition is no longer valid. What is more, this mordent can contribute
tremendously to the effect of imitation. It should therefore be added on the downbeat notes
in M: mm. 27, 28, 31, 32; 57, 58, 61, and 62; U: mm. 59, 60, 63, and 64; and L: mm. 69
and 70. Note that in all cases, the mordents upper neighbor note must match the harmony
realized by the other voices, which may be different from the home key and its key
signature. Thus the mordents on D in M: mm. 271 and 571 must both begin on E, that on
A in U: m. 301 on B, that on G in U: m. 591 an A, and that on A in U: m. 631 on B.
464
WTC II/13
a redundant entry and an episode derived from the subjects tail. The third
section (mm. 64 84) is shorter and contains only three entries linked by the
latter type of episode. Harmonically, the first section is in the key of the
tonic, the entries in the second section are on the dominant, tonic, tonic
relative, and subdominant respectively, and the statements of the third
section return to the realm of the tonic.
Owing to the very carefully laid-out contrasts within this fugues

material, contrasts of color and/or of levels of intensity rather than overall
dynamic developments are the expressive aim of this fugue. A helpful
approach is to think in terms of different registers; not so much in the sense
of the registers of an organ or harpsichord, but in terms of different instrumental colorse.g.: woodwind for the primary material (the subject, the
counter-subjects, and the notes accompanying them), solo strings for the
motivically determined episodes E1 and E3, and two recorders accompanied by a bassoon for the episodes E2, E4, E5, E6, and E7.
WTC II/14 in F minor Prelude

The F-minor prelude is conceived in a texture of three parts assuming
different degrees of melodic significance. The upper voice dominates by
far the largest number of measures. Occasional imitations take place
between upper and middle voice but never involve the lower voice, which
is limited to scalar motions in neutral note values, cadential accompaniment, some ornamented broken chords, and occasional links. Although the
main melodic component of the prelude recurs twice in the course of the
piece, complete with its accompaniment and imitation, one cannot speak of
a motivically determined composition. This component is used in an almost
improvising manner, as are other models spawning sequences. It would
thus be fitting to speak of an improvisation around a few thematic models,
connected by passages in free play.
A cursory glance at the beginning of the piece reveals that each of the
three initial measures contains a perfect cadence in the home key. These
harmonic progressions thus do not serve structural purposes. Bach allows
for a first caesura at the modulation to the dominant key (see m. 72 where
the C-major chord is preceded by a partial representation of the G9 chord,
B-D-F-A). As this harmonic conclusion takes place on a weak beat, the
truly convincing section ending may still be pending. This occurs with the
reconfirmation of the modulation to C and the return of the first
component in m. 12.
The prelude can be divided into four sections:
I mm. 1-72-121 tonic to dominant
F minorC minor
II mm. 12-211
dominant to rel. major C minorA major
III mm. 21-291
rel. major to dominant A majorC major
link 29-301 return to tonic
IV mm. 30-43
tonic confirmed
There are no substantial analogies beyond mm. 1-22 . 12-132 . 30-312.
The prelude emanates a sense of calm, a floating unearthliness. This
effect is achieved by a combination of various means. There are, first of
all, a great number of different note values and used alternately in melodic
lines, as part of hidden two-part structure, and in ornamental patterns:
half-notes with or without tie prolongation, dotted quarter-notes and
quarter-notes, dotted eighth-notes and eighth-notes, normal 16th-notes and
16th-note triplets, and even dotted 16th-notes and 32nd-notes all take part
465
466
WTC II/14
in the melodic processes of this piece. In addition to numerous ties that

often only have a passive function in complementary-rhythm patterns,
there are also syncopations occurring in active positions (see, e.g., the
syncopated quarter-notes in mm. 1 and 2, and the syncopated eighth-notes
in mm. 3, 5, etc.).
The pitch patterns in this prelude suggest constant changes in the
melodic value of the single note. The first two measures may serve as an
example for what exerts crucial influence on the mood and character.
In the upper voice, the notes in beats 1 and 2 are melodic, the
16th-note triplets are an ornamentation of the simple line C-B-A,
and even the regular 16th-notes in m. 21-2 represent a written-out
turn figure around the A that is the target note of the little melodic
component. The 16th-note C is a link, while the following eight
notes are a parallel to the middle voice that is now in the lead with
its imitation of the main component. The essential melodic
content of these two measures thus consists merely of the following
line: F-CA-B-CBA. Subtle touch control and dynamic
coloring is crucial to bring out all the nuances.
In the lower voice, all notes in m. 1 represent melodic steps of
secondary importance. The keynote at m. 21 qualifies serves as a
harmonic support.
In the middle voice, the three notes forming the rising F-minor
triad do not represent a melodic event but, together with the initial
F of the lower voice, provide harmonic support. The high F marks
the beginning of one of the traditional harmonic closing formulas
(dosido) and thus acquires secondary melodic quality. Its resolution doubles as a beginning of the principal components imitation.
This F, together with the subsequent C-A-B, are melodic notes of
primary importance, while the triplets concluding the measure are
an ornamentation of C-B-A, as was explained above for the initial
presentation of the melodic component in the upper voice.
The articulation in this prelude is uncomplicated. Basically, legato
applies to all notes that belong either to a melodic line, an ornamented
melodic line, or a broken chord representing a vertical effect (i.e., a harmonic support, as explained above for M: m. 1). Non legato is appropriate
only for the slower note values that progress melodically in skips (see, e.g.,
L: mm. 14-151, 16-171, 18-191, mm. 13-21 A-C-F, mm. 22-231 G-C-F-B,
mm. 24-251 A-D-G-C, etc.), and for cadential-bass patterns as in mm.
11-12: G-C and mm. 20-21 E-E-A.
F minor
467
The score features three ornaments in mm. 9, 23, and 25. All appear as
mordents in brackets, suggesting that they do not stem from the main
source todays scholars regard as authoritative for Bachs intentions. If
played, all begin on the main note and comprise a simple three-note shake.
It might, however, be a good idea to consider abstaining from additional
ornamentation in a piece so rich in written-out embellishments.
The musical events in this prelude can be viewed from two entirely
different viewpoints. One approach is to trace a hidden large-scale framework that, while only the skeleton of the prelude, might provide important
orientation for the integration of small-scale events into a larger design.
The large-scale processes, together with the dynamic outlines in the main
components, are better shown visually than explained verbally.
468
WTC II/14
The other, more conventional approach traces the melodic patterns and
sequence models. The principal component, together with is contrapuntal lower voice, the filling broken chord, closing formula, and subsequent
melodic imitation in the middle voice, recurs in mm. 12-13 (imitation
shortened) and in mm. 30-31. The syncopation figure introduced in U: m.
3 reappears (always in the treble) in a large variety of pitch patterns
throughout the prelude (see mm. 5, 22, 24, 28, 32, 33, 34, 36, and 41). In
mm. 132-152, a three-part model is presented that is sequenced one note
lower in mm. 152-172 (with small variations in the lower voice) and then
followed by two contracted sequences (see mm. 172-182 and 182-192) and
a cadential close. A second three-part model is introduced in mm. 21-22
(U: from m. 212 E; M+L from m. 211). This model is sequenced one note
higher, also followed by three contracted sequences (mm. 25, 26, and 27)
and a harmonic conclusion in an imperfect cadence with fermata and
general pause (m. 29). The preludes final section, having begun once
more with the main idea, does not establish any new models but freely
alternates the syncopation figure with patterns formed from 16th-note
triplets. An interrupted cadence at m. 391 diverts a formula that seemed
headed for C minor in favor of an arrival, in one voice after another, at the
tonic with Picardy third .
WTC II/14 in F minor Fugue

Beginning on the fourth eighth-note of m. 1 and concluding at m. 41,
this subject spans a little less than three measures. The characteristic trill
on the penultimate note, which Bach uses so often in the first volume of
the Well-Tempered Clavier but only this once in the second volume, marks
the end of the phrase. The pitch pattern comprises a broken chord and three
large intervals (sixth, fourth, and fifth) as well as two written-out inverted
mordents. A glance across the remainder of the fugue shows that leaps and
broken chords are a constant feature. With regard to rhythm, the subject
features 16th-notes, eighth-notes, three quarter-notes prolonged by ties to
form syncopations, and the ornamented half-note. This rhythmic complexity
is not, however, typical for the entire piece but later gives way to long
stretches in a much simpler pattern.
The subjects phrase structure allows for two interpretations, depending
on whether an interpreter wishes to emphasize the rhythmic particularities
or the melodic backbone. The subjects rhythmic pattern can be regarded
F minor
469
as consisting of two similar halves, with the second slightly extended: see
upbeat ( ) / syncopation / inverted mordent / long note varied to become
upbeat ( ) / syncopation / inverted mordent / short note + tail (F-G-F).
In this light, the subject presents itself as divided into two subphrases. The
pitch pattern presents a different message. It can be described as combining a broken-chord upbeat followed by an ornamented descending scale.
This scale features three syncopated half-notes leading, after a quarter-note
inserted to correct the meter (so to speak) to a strong-beat half-note that
resolves onto the keynote: upbeat-DCBAGF. In this view,
subphrasing is inconceivable and the subject appears as one indivisible
unit.1 The subjects harmonic layout (which can be observed most easily in
mm. 16-19) emphasizes the syncopations by allotting each of them a shift
from a primary function to a secondary dominant:
The dynamic development in the subject obviously depends on ones

choice with regard to phrasing. The overall climax, however, is undisputed.
The D combines all tension-enhancing aspects: as the first syncopation, it
is metrically most exposed; as the representative of the subdominant chord,
it is the target of a harmonically active step; as the sixth degree of a minor
scale, it holds special melodic tension; and in terms of interval structure, it
is approached in what is the largest leap in the subject and a high-tension
interval.
The two possible interpretations thus differ primarily in what happens
in the middle of m. 2. In an indivisible phrase, the Cs are not too relaxed
as they hold tension that will be released throughout the descent to the
final note. The eighth-note F following the second C is played lightly, as
a kind of charm note interrupting the (more important) line from C to B.
By contrast, in a divided phrase the syncopated C is the ending of the first
subphrase and thus very relaxed, while the ensuing F acts as a new upbeat
and is thus an expressive, actively forward-bouncing note.
1
There is no right and wrong. At this early point in material appreciation, the preference
depends entirely on the view adopted by each individual interpreter. Later in the fugue we
may find that other components work better with one of these concepts than with the other,
but so far, the choice is one of taste only.
470
WTC II/14
The complete subject appears ten times in the course of the fugue.
1
mm. 1-4
M
6
mm. 34-37
L
2
mm. 4-7
U
7
mm. 51-54
M
3
mm. 8-11
L
8
mm. 54-57
U
4
mm. 16-19
U
9
mm. 60-63
L
5
mm. 28-31
M
10
mm. 66-69
U
Inversion, stretto, and parallel do not occur, and the changes in the
subjects shape are minor. The characteristic adjustment of the initial interval in the answer occurs only in m. 4 as the two further subject statements
appearing in the harmonic position of an answer (mm. 54 and 60) both
feature a varied beginning. The subjects ending is modified in mm. 36-37,
where the ornamented step is replaced by a cadential-bass pattern, and in
m. 69, where the resolution of the trill is delayed.
Bach invents no regular counter-subject for this fugue. Instead, components of the subject itself appear in different combinations as contrapuntal
material.2 Nonetheless, this is by no means all there is to be said about the
primary material, for as one discovers only later, this F-minor fugue is a
triple fugue: a fugue with three subjects. The second and third subjects are
introduced separately before they are combined with the first. Moreover, as
with the first subject, several of their components appear as contrapuntal
material, eventually accompanying any of the three subjects.
The second subject makes its first appearance in L: mm. 20-22 after an
A-major cadence. Beginning thus in the major mode, its initial descent
with the characteristic dotted-note group leads to a raised fifth that implies
a modulation back to the original tonic: E, the leading note to F, leaves A
major. The ensuing fourth leap to the syncopated F seems to introduce a
traditional closing formula, and indeed this dosido figure constitutes the
ending of the second subject in several of its later statements (see e.g. M:
mm. 56-57 and mm. 62-63; L: mm. 68-69; with slight variation also L:
mm. 24-25). In the first few presentations of this second subject, however,
the resolution is omitted, replaced by a chromatic descent to E. This is tied
and resolves belatedly (outside the subjects actual scope) onto a note that
initiates a cadential-bass pattern in B major. The rhythmic pattern of this
2
See mm. 4-7. What begins like an independent voice (diatonic ascents interrupted by a
tritone leap downward and various rests) ends with a partial imitation of the subjects end.
In mm. 8-11 the secondary voices both imitate the subjects initial broken chord.
F minor
471
second subject is thus extremely intricate, with a 16th-note, eighth-notes,

a dotted eighth-note, and a syncopated half-note. The salient feature in the
pitch pattern is the fourth leap that confirms the initial modulation. This
leap also focuses dynamic tension, shaping the second subject into a single
curve with the climax on the syncopated half-note.
The following table lists the complete statements of the second subject.
(Its head motif particularly is used freely in-between). Interestingly, the
number of these statements is again ten and thus the same as that of the
statements of the main subject.
1
mm. 203-221 L
6
mm. 27-29
M
2
mm. 213-231 U
7
mm. 30-31
L
3
mm. 223-241 M
8
mm. 56-57
M
4
mm. 231-251 L
9
mm. 61-63
M
5
mm. 27-28
U
10
mm. 68-69
L
The third subject is introduced in M: mm. 36-37. Apart from its eighthnote upbeat it consists exclusively of 16th-notes in an ornamental pattern.
Its structure is simple as the second half is a sequence of the first, so that
the phrase can be described as upbeat + half-measure with 16th-notes +
half-measure sequence + final note. (The ensuing eighth-note leaps, while
included in the first imitation, do not form part of this subject later on.
Moreover, the harmonic conclusion is reached at the D on the middle beat
of m. 37.) The dynamic shape consists of a long decrease after a very
active upbeat.
This subject, although entering so late, manifests eleven times and thus
in even more entries than S1 and S2:
1
mm. 36-37
M
6
mm. 42-43
M
2
mm. 37-38
U
7
mm. 44-46
L
3
mm. 38-39
L
8
mm. 45-47
U
4
mm. 39-40
M
9
mm. 55-57
L
5
mm. 40-42
L
10
mm. 60-63
U
11
mm. 67-69
M
The short component undergoes two kinds of modifications: its beginning is frequently varied3 and its extension is expanded to twice the original
length in the last three entries where its scope is adjusted to that of S1.
The presentation of the three subjects in this fugue follows a pattern
that allows for an almost ideal appreciation of each one. S1 appears in its
3
See M: m. 42 = fifth instead of second; L: mm. 44-45, U: mm. 45-46 and L: m. 55 = three
16th-notes instead of one eighth-note.
472
WTC II/14
three initial statements plus one redundant entry (M, U, L, U) before S2

first emerges. S2 is then given space for an equal number of statements
(see mm. 20-251: L, U, M, L) plus an additional tight stretto (mm. 27-29).
The redundant lower-voice entry of S2 in mm. 23-25 and the stretto both
hint at the re-emergence of S1 (see the broken chord followed by a syncopation in M: mm. 24-25 and L: mm. 28-29). The juxtaposition of S1 and
S2 in mm. 28-31 is thus gradually prepared. It retreats equally gradually in
that the following lower-voice entry of S1 is contrasted only with the head
motif of S2 instead of its full phrase. S3, as if eager to make its appearance,
enters in overlap with both the conclusion of the S1 entry and the melodic
closing formula in U: mm. 36-37. After thus taking the stage somewhat
impatiently, S3 is given room not only for its first round of three-plus-one
statements (mm. 36-40: M, U, L, M, concluded by a melodic closing
formula in U: mm. 40-41) but for two more entry pairs (L, M: mm. 404-433,
L, U: mm. 444-471). The first of these pairs is accompanied by two
reminders of S1: the inversion of the S1 answers head (see the parallel in
U + M: m. 41) and the syncopation + inverted-mordent figure in M: mm.
41-42. The two quotations may be interpreted as an anticipation of the
impending juxtaposition of S1 and S3. The announcement is not, however,
immediately followed up, and further S3 statements unfold without any
interference of the two earlier subjects.
The re-emergence of the main subject announces the great synthesis in
three three-subject juxtapositions:
mm.
54-57 60-63
66-69
U
S1
S3
S1
M
S2
S2
S3
L
S3
S1
S2
The music example shows the dynamic juxtaposition:
F minor
473
The fugue encompasses ten subject-free passages.

E1 mm. 71-84
E6 mm. 433-444
E2 mm. 113-162
E7 mm. 471-514
E3 mm. 191-204
E8 mm. 573-602
E4 mm. 251-271
E9 mm. 631-664
E5 mm. 313-342
E10 mm. 693-70
These episodes feature fragments of the three subjects, elements from
the non-thematic components that accompany the subjects in lieu of
counter-subjects, as well as three largely independent episode motifs.4
Material from S1 is heard primarily in the exposition of S1 (i.e., in E1, E2,
and E3), the S2-head enriches the secondary material in the context of the
exposition of S2 (i.e., in E4), and the motif derived from S3 plays a
corresponding role in the exposition of the third subject (i.e., in E6 and
E7). The S1/S2 juxtaposition is followed by E5 with material from the first
two subjects. The greatest independence of secondary material is reached
in the two episodes linking the three three-subject juxtapositions (i.e., in
E8 and E9).
A fugue with such a wealth of thematic material requires more than a
simple answer when it comes to characterization. Contour and rhythm in
the three subjects suggest different color shades within the same piece,
with a distinct easing of the complexity from the first to the third subject.
The third subject is lively. Its rhythm pattern is simple and its line
ornamental. Under its influence (mm. 37-51), the surrounding material also
displays a predominance of eighth- and 16th-notes with frequent leaps.
The appropriate articulation for this subject is quasi legato (or very crisp,
light legato) in the 16th-notes and equally light-weighted non legato in the
S3 upbeat and the accompanying eighth-notes.
S2 is the most melodious component in the fugue. Its beginning and
end, i.e., the four-note descent and the conclusion in either closing formula
or chromatic deflection, are best expressed in legato. The consecutive leaps
in the middle demand detaching. Performers may choose whether to treat
the thus separated C as an active force in preparing the climactic F (and
therefore play it only slightly softer than the syncopation) or as an escape
from the line consisting of the stepwise progression A-G-F-E-F etc.
(and thus give it a much softer tone color than its surroundings).
4
M1 cf. U: mm. 113-131 A-A, imitated in M: mm. 13-14),

M2 cf. M: mm. 11-131 F-E, imitated in U: mm. 13-14), and
M3 cf. L: mm. 12-141 E-D, sequenced in mm. 14-16; partially taken up in M: m. 24,
L: m. 28, and M: mm. 32-33).
474
WTC II/14
S1 contains features of both characters. The complexity of the rhythm,

particularly in the two syncopations, needs time (and thus calm), while the
overwhelming content of leaps as well as the two ornamental figures
epitomize liveliness. The secondary material that derives from this subject
and often accompanies it also prioritizes leaps. The practical difference for
the articulation in this subject is, however, minimal. In both the calm and
the lively characters, the initial broken chord with the ensuing leap and the
consecutive leaps in m. 2 are non legato. As there is a harmonic change
within the first and third syncopations that converts the tied portion into an
appoggiatura, these long notes must be tightly linked to the ornamented
figures that follow (legato in mm. 14-21: D-C and in mm. 24-31: B-A).
Only the three notes in m. 3 (A-F-G) are legato or non legato depending
on the performers approach to the basic character.
The best solution for the entire fugue is to assume a rather lively basic
character in a moderately flowing tempo. The relative tempo of the prelude
to the fugue is complex. The easiest way of perceiving the proportion is to
convert a triplet 16th-note in the prelude into a 16th-note in the fugue.
The articulation of the secondary material in mm. 1-20 requires primarily lightly detached notes, with legato only in the two melodic closing
formulas (U: mm. 10-11 and L: mm. 15-16) and the neutrally colored
cadential-bass notes in L: m. 20. S2 and the motif derived from it sound
legato but are often surrounded by non legato notes. In the sphere dominated by S3, the only longer notes requiring legato are those in melodic
closing formulas (U: mm. 36-37, 49-51, 51-52, and 70). Finally, in E8 and
E9 it is possible to suggest relationships between M2 and S1 by playing
M2 non legato, and between M3 and S2 by playing M3 legato.
The prominent ornament in this fugue is the trill in the subject. As it
forms part of the thematic material, it must be played in each entry, regardless of whether or not Bach notes this. The trill is approached stepwise and
therefore begins on the main note. Its shakes are twice the speed of the
fastest note values, meaning: the trill is launched from a 16th-note before
it continues in fourteen 32nd-notes, the last two of which constitute a
suffix. There are only two S1 statements that either preclude the ornament
owing to a varied ending (mm. 36-37) or feature a trill that ends without
the suffix in a tie (m. 69).
The structure of the F-minor fugue is basically simple despite what
could be described as a three-dimensional layout. We can distinguish four
larger blocks, dominated respectively by S1, S2, S3 and the three-subject
juxtapositions. These are the details:
F minor
I
II
III
IV
475
consists of the three initial entries of S1 (the third concluding in
a closing formula), the short linking E1, the more substantial E2,
a redundant entry, and the concluding E3 whose cadential-bass
pattern confirms the modulation to the tonic relative (A major)
on the middle beat of m. 20 (S1: M U L, U).
encompasses two segments. IIa presents three initial statements
of S2 plus a redundant entry. This fourth entry introduces the
final version of the second subject with its resolving ending and
thereby concludes this segment at m. 251, still in A major
(S2: L U M, L). IIb begins with an episode featuring three
entries of the S2-derived motif. It is followed by a S2 stretto, a
S1/S2 juxtaposition, a longer episode, and an S1 entry accompanied by the S2-derived motif. The section concludes with a
cadential formula in C minor, the minor dominant, at m. 371
(S2: U M L, S1: M L).
encompasses three segments. IIIa is launched prematurely in a
one-measure overlap with the previous section. It presents the
three initial statements of S3 plus a redundant entry. This
redundant entry is accompanied by a melodic closing formula
that marks the modulation to G major, the secondary dominant
V/V (S3: M U L, M). IIIb is very short, containing only two S3
entries and a reiteration of the cadence in G major (S3: L M).
IIIc begins, similarly to the last segment of section II, with an
episode featuring the S3-derived motif. It is followed by two
further entries of S3 and a longer episode concluding in m. 52
with a melodic closing formula in the subdominant B major5
(S3: L U).
begins almost like section III with an overlap (here of eighthnotes only). Its opening S1 statement represents the subdominant
key, but the second half of m. 54 modulates back to the tonic.
There follow the three three-subject juxtapositions, linked by
two episodes and rounded off by a short cadential conclusion.
For the four S1 entries of this section Bach has chosen the same
voices as in the four statements of the first section, thus closing
a large bracket (S1: M U L, U; S2: M M L; S3: L U M).
Note that there is a misprint in the score here: in mm. 49-50, all voices cancel the G to G
except for the lower voice on the middle beat of m. 50 where the natural is almost certainly
omitted by mistake.
476
WTC II/14
As is obvious from the layout, the issue in this fugue is the separate
presentation and eventual juxtaposition of contrasting subjects. Superimposed dynamic developments are therefore not likely to be the composers
intention. While the exposition of S1 constitutes a gradual build-up of the
ensemble from single-voiced to three-part texture, the lack of enhancing
contrapuntal features limits the dynamic development. The expositions of
S2 and S3 present a similar picture. Section IV is the most intense in the
fugue. This applies to the entire section rather than to any single entry. The
fugue thus presents itself as a composition with four sections in four
different shades of emotional involvement. The only color contrasts within
each section are provided by the episodes.
WTC II/15 in G major Prelude

This prelude can be categorized as a playful piece. It features a simple
rhythmic pattern with almost constant 16th-note motion including conventional figurations in hidden two-part structure over repeated indirect pedals.
All thematic components share a surprisingly low melodic profile.
The first harmonic progression announces a conclusion in D major in
m. 13 but is diverted in the manner of a deceptive cadence to B minor. A
perfect cadence in the expected dominant is delayed until the repeat sign.
The second section concludes in m. 28 with a cadential formula different
in detail but analogous in function to the one that was interrupted before
(mm. 27 . 13). The key reached here is E minor, the tonic relative.
Although this is a perfect cadence that does not need correcting, it is followed in mm. 31-32 by another harmonic close over a similar bass figure,
confirming a modulation to the subdominant C major. The return to the tonic
seems imminent from m. 37 onward. Yet the cadential formula in m. 45
the same formula that has been heard twice beforeleads again into a
deception (m. 46: E minor instead of G major) and is only in mm. 47-48
followed by a perfect cadence in the home key.
The prelude encompasses several crucial structural correspondences:
mm. 1-6 . 17-22 . 37-42 (varied), mm. 9-11 . 23-25 (transposed, varied),
and mm. 11-16 . 43-48 (transposed).
The basic character of this composition is clearly rather lively. The
tempo may be swift and flowing. The appropriate articulation distinguishes
quasi legato for the 16th-notes in the hidden two-part structure and a
denser legato for the 16th-notes in ornamental figures1 from a gentle non
legato, which is adequate for all eighth-notes.2
Several ornaments appear in the score, some used consistently, others
remaining local. The turn on the last note of the upper voice in m. 13 recurs
in mm. 27 and 45. In all cases, it begins on the upper neighbor, contains
1
See, e.g., the written-out inverted mordent on the first beats of mm. 7-11. If played in the
same touch as the surrounding 16th-notes, the middle notes of these ornaments might appear
as belonging to the melodic line in the hidden two-part structure, while the melodically
relevant tone is actually the main note of this mordent, the measures second 16th-note.
2
The only exception, owed to a particular convention, occurs in m. 16 and correspondingly

in m. 48. Scalar runs in 16th-notes launched from an eighth-note were a standard figure of
the era and always played with their legato including the initial longer value.
477
478
WTC II/15
four notes and must be played fast particularly in front of the octave
displacement of its resolution (e.g., as a 32nd-note triplet plus 16th-note).
In two cases, the same cadential formula is additionally embellished. In m.
45, there is a mordent on the third 16th-note, enhancing the virtuoso
character of the piece. It can easily be integrated, again as a 32nd-note
triplet. Performers may consider transferring this mordent to the corresponding note in m. 13. The additional turn in m. 26, on the other hand,
occurs in a surrounding of melodic eighth-notes not adorned elsewhere.
Playing it underscores the preludes playful side but obscures the parallels
of the two hands. If included, this turn begins on the main pitch and
encompasses five regular notes. The cadence concluding the first section
also features two ornaments. Both the accented bass note of the V7 chord
and the final I carry inverted mordents. In the corresponding cadential close
in mm. 47-48, the first of the two ornaments is omitted. If played in m. 15,
it should also be added here. As to the pitches of the lower neighboring
notes, only the D in m. 16 requires an artificially raised leading note since
the modulation has brought us to the key of D major. All other inverted
mordents use notes from the G-major scale.
Finally, there are four purely melodic ornaments. Three are taken up by
the inverted mordent that enhances the E in m. 32 and its two sequences.
All are meaningful ingredients both with regard to the character and the
structural clarity of the piece. The fourth, a mordent in m. 20, is also very
listener-friendly: it emphasizes a new beginning within the otherwise
almost confusingly long chain of eighth-notes unfolding before the background pedal in the hidden second part.
Thematically the prelude is based on several homophonically conceived
units with a distinct texture. Three of these models define the first two of
the preludes three sections. Model 1 consists of two pedals: a sustained
pedal on the first scale degree and a repeated indirect pedal D on the weak
beats of the hidden two-part structure in the other voice. In between, there
are two lines moving in eighth-notes: one is written in eighth-note values,
while the other forms the melodic strand of the hidden two-part structure.
These lines move in parallel sixths. In the ensuing measures, the model
recurs in a mixture of sequence and imitation: The sustained pedal and the
repeated indirect pedal have exchanged hands, as have the same two lines in
eighth-notes, which now move in compound thirds, the inversion of parallel
sixths. When Model 1 recurs in mm. 17-22, the two pedal notes sound in
octaves. The third recurrence of M1 inverts the perfect-fifth interval of the
pedal notes so that they form compound fourths. As if slightly uneasy with
this interval, the pedals move from measure to measure (see mm. 37-40:
G major
479
D/G, E/G-E/A, F/A-F/B, G/B). The parallel eighth-note lines are in

compound thirds but display small irregularities owing to the shift from
measure to measure.
Model 2 is similar in texture (see L: from m. 7, U: from m. 8). Once
again there are one or two pedal notes providing a backdrop for melodic
parallels. While the pedal pitches are enlivened by a written-out inverted
mordent on the downbeat, the melodic parallels now start out from a
quarter-note before resuming their eighth-note motion. When the pattern
recurs in mm. 23-24, the left-hand part is considerably changed. The closing
measures of each section develop from the previous four strands through
three strands (with only one pedal) to a two-part texture without any
pedal-note layer. Although they share few surface details, their analogous
textural development earns them the label Model 3.
At the beginning of the third section, Bach adds two further models.
Model 4 (mm. 29-30, partial sequence in 31-321) combines the repeated
pedal with eighth-note lines whose parallel is only a hint. Model 5 (U:
from m. 32, L: from m. 33) is the only component in this prelude to present
some polyphonic independence of voices.
To sum up, the preludes structural layout can be depicted as follows:
section I
section II
section III
M1
M1
M4
M1
M1
M5
M2
M2
M1
M3
M3
M3
cadential close
cadential close
cadential close
WTC II/15 in G major - Fugue

All who think they know what to expect of Bachs fugues will be
surprised by this subject, which is unusual both in its extremely regular
rhythm and in its non-melodic pitch pattern. An uninterrupted string of 16thnotes spans way beyond the confines of the subject itself and, were it not
for the harmonic full stop reached with the return to the tonic at m. 61,
would leave the listener at a complete loss.
The subject, then, is exactly five measures long, the upbeat with five
16th-notes being complemented by the final note. Each measure presents
a broken chord pattern. In fact, the interval pattern contains only a single
second (see mm. 2-3), and even this is not perceived as a melodic step.
480
WTC II/15
While mm. 1 and 2 play within the tonic chord, m. 3 along with the first
harmonic shift establishes a figurative pattern that is sequenced in mm. 4 and
5. No subphrasing is needed or even desirable in this virtuoso play of
broken chords. The harmonic outline describes an active step mm. 2 to 3,
thus enhancing the outset
of the above-mentioned
sequence. The dynamic
7
line can only follow the
simple layout expressed
in harmony and pitch. The result is a two-measure increase up to m. 31,
followed by a gradual decrease to the final note.
The fugue comprises only six subject statements:
1 mm. 1-61
U
4 mm. 33-381
L
2 mm. 8-131
M
5 mm. 40-451
U
3 mm. 15-201
L
6 mm. 65-701
M
Apart from the adjustment of the first interval in the answer, the
subject remains unchanged throughout the fugue. It never appears in either
inversion or stretto.
Bach invents two counter-subjects for this fugue and uses them very
regularly albeit with some variation. CS1 is introduced against the subjects
answer (U: mm. 8-13). In its original shape it consists of a five-16th-note
upbeat followed by slower note values in an overall descent including two
syncopations. It ends in m. 13, in the weak position on the third 16th-note
where the suspended E resolves indirectly into F. CS1 accompanies all
further subject statements. It recurs in M: mm. 15-20 in a simplified
pattern: the descent seems as if cleaned, i.e., freed of all ornaments and
escape notes, and the final resolution occurs at m. 201. In M: mm. 33-38
the initial 16th-note run is replaced by a dotted-note group, and in mm. 4041 the upbeat is moved to another voice (L instead of M). Finally in U:
mm. 65-70 the main rhythmic pattern is substituted by a chain of suspended
notes, with the last broken into a chord. Only the descending direction may
still remind listeners of the original CS1.
CS2 is introduced in mm. 16-20, i.e., against the subjects third statement. Entering one measure late, it is characterized by a group consisting
of a syncopation followed by a fifth leap down to a pair of rising 16thnotes. The little group is sequenced twice in descending direction. The
final syncopation resolves belatedly on the weak beat after the completion
G major
481
of the subject (just as the final note of CS1 had done 7 measures earlier).
CS2 accompanies the following two subject statements but is absent in the
final entry. In the course of its two further entries it, too, is slightly varied:
the one-measure rest before the belated beginning is filled in with 16thnotes (see U: mm. 33-34 and L: mm. 40-41), and the ending is diverted
(see U: mm. 37-38 and L: mm. 44-45).
In view of so
consistent a use of
the contrapuntal
material, it may
come as a surprise
that, upon closer
inspection, these
counter-subjects
are found lacking
a crucial quality
required in true polyphony: they are not
independent of the subjects pitch pattern
but actually ornamented parallels! The
examples show the subject with its two
companions as found in mm. 15-20, first in
Bachs setting and then in skeletal simplification revealing the dependency, which
extends to the dynamic design.
In this fugue, each subject entry is separated from the subsequent one.
There are thus as many episodes as there are subject statements.
E1 mm. 6-81
E4 mm. 38-401
E2 mm. 13-151
E5 mm. 45-651
E3 mm. 20-331
E6 mm. 70-72
Bach has invented three episode motifs. The first is introduced between
the dux and the comes. It is related to the beginning of CS1, which it thus
anticipates (U: mm. 6-71 . mm. 8-91) and recurs in similar function in E2
and E4 as well as varied in E3a. The two other motifs are derived from the
subject. Ms1 (L: mm. 22-251 . mm. 1-31) dominates the ten-measure E3b,
Ms2 (L: mm. 45-471 . mm. 3, 4, 5) the even longer E5a. The two episode
motifs are accompanied by remotely analogous companions.
The two long episodes E3 and E5 are subdivided by D-major cadences
whose melodic closing formulas in mm. 22-23 and 61-62 suggest section
endings. The harmonic development on the inner sides of these cadences is
482
WTC II/15
fascinating insofar as the second is designed as a free retrograde of the

first: in E3b the imitations of Ms1 are in D major, G major, E major, A
minor, and B major. In near mirror symmetry the imitations of Ms2 in E5a
pass through B minor, E major, A minor, D major, G major, D major, and
G major. The only episodes not to be characterized by motifs but by runs
are E5b and E6. They affirm the fugues virtuoso character by doubling its
so far fastest note values.
The tempo chosen for this very lively fugue may be fast, limited only
by the performers technical skills to produce crystal-clear and very regular
32nd-notes in the runs and trills. The relative tempo of the prelude to this
fugue sounds and looks simple: an eighth-note in the prelude corresponds
with an eighth-note in the fugue. (Approximate metronome settings:
prelude beats = 108, fugue beats = 72.) The actual effect for the listener is,
however, one resembling a hemiola: 1 2 3 4 5 6 in one measure of the
prelude turns into 1 2 3 4 5 6 in two fugal measures. The corresponding
articulation requires a light quasi-legato touch for the 16th-notes and a
detached but not too short non legato for the eighth-notes and quarternotes. There are two exceptions. The first occurs in the two eighth-notes
that Bach explicitly marks staccato (mm. 10 and 12): they should be
distinguished from the surrounding non legato by particular lightness. The
second appears in suspended notes that create harmonic tension awaiting a
resolution. Just like appoggiatura-resolution pairs, these may under no circumstances be separated but must be played in tight legato (U: mm. 20-22;
M: mm. 25, 27, 29, and 31; U+M: m. 38, U: m. 39, and U: mm. 65-69).
The fugue features five kinds of ornaments. The two inverted mordents
in mm. 10 and 12 pose no problem. In the two trills in mm. 50 and 52, the
ornamented note finds no harmonic or melodic resolution. As a result, the
trill does not end in a suffix but stops short before the bar line. Both trills
begin on the upper neighbor and shake in 32nd-notes. Each trill thus contains 12 notes the last of which is tied to the following downbeat. The three
trills with suffix and resolution in mm. 57, 59, and 61 begin like the other
trills on the upper auxiliary and end after five shakes in the suffix indicated
by Bach. The mordent in m. 62 is approached stepwise and therefore
begins on the main note. Its rhythm is a 32nd-note triplet. Lastly, the final
note in the upper voice is preceded by a grace note. This C represents an
appoggiatura that, played on the beat together with the notes of the middle
and lower voices, resolves onto B after one eighth-note.
The entering order of the voices together with the design of the episodes
and the harmonic organization of the piece add up to a clear picture of the
fugues structural layout. The six subject entries appear in two complete
G major
483
rounds: U M L and L U M. The episodes also appear as 3 + 3. In the first

half of the piece there are two two-measure episodes (E1 and E2) which
have no other purpose than to link subsequent entries. They are complemented by the long and complex E3. In the second half of the piece, E4
corresponds with E1 in length, material, and purpose, while the long
episode, rather than following the third subject entry, now precedes it for
a more powerful closing effect. The final episode is again two measures
long and thus completes the analogy.
The first group of subject statements relates to the home key, with
entries in G, D, and G major. The entries in mm. 33-38 and 40-45 are in
E minor (the tonic relative) and B minor (the dominant relative). The
change of mode marks the beginning of a new section. The final entry
returns to D major and concludes the second section on the tonic.
The dividing line between the two sections is somewhat hidden. One
might wonder whether Bach did not want the entire fugue to appear as one
virtuoso gesture, without too much obvious structuring. The cadence that
concludes the first section in E minor in mm. 33-34 overlaps with the
beginning of the second section and thus allows not the slightest breathing
either for the performer or for the listeners.
The development of tension in this fugue mirrors the works structural

simplicity. The three subject statements in the first section represent a
gradual dynamic increase, due both to the growing ensemble and the rising
intensity within the two linking episodes. The tension decreases slightly in
the cadential close of E3a, only to rise again in the ascending sequences of
E3b. The relaxation at the end of this episode is short and incomplete, so
that the second section begins on a very elevated level. (A change of color
in the minor-mode portion may be expressed more effectively by means of
touch and articulation than through dynamics.)
484
WTC II/15
In the second section, the first two entries are linked once more by the
dynamically increasing motif derived from the counter-subject. At the
same time, the choice of voices (L, U) and the extension of the pitch range
to the highest octave available on Bachs keyboard instruments support
another growth in tension. After this climactic moment, the long E5a with
its descending sequences brings about a gradual relaxation. Although the
brief E5b presents a powerful rush upward, the final subject statement
enters on a much softer note, owing mainly to its relatively weak position
in the middle of the texture and its reduced polyphonic density, as it is
surrounded by a simplified CS1 and an accompanying bass instead of CS2.
The concluding cadence and particularly the appoggiatura in the final
measure confirm the soft ending.
This playful fugue thus presents itself dynamically in a design that is as
simple as it is capturing:
WTC II/16 in G minor Prelude

This prelude turns around rhythmic patterns. In fact, one rhythmic
figure defines practically the entire composition. This figure appears in
numerous melodic guises. Among them, we can isolate a small
number of recurring models. Now in one voice alone, now in
two complementary voices, they create a continuity of the dotted
rhythm that is almost unbroken throughout the prelude.
The first harmonic progression closes at m. 21 where the initial return
to the home key materializes. This cadential close coincides with the end
of the initial model and thus does not have larger structural bearing. The
modulation launched shortly thereafter leads to the subdominant C minor,
reached with a cadential formula at m. 51. This closure should be regarded
as structurally relevant, not least because of the recurrence of the initial
model thereafter (mm. 5-6 . mm. 1-2).
The prelude comprises six sections grouped into two larger parts:
I
mm. 1-51
tonic to subdominant
II
mm. 5-91
subdominant to dominant
III
mm. 9-111
dominant back to tonic
IV
mm. 11-132
tonic to subdominant
V
mm. 13-201
subdominant back to tonic
VI
mm. 20-21
tonic confirmed
The G-minor prelude is one of only three works among the 48 preludes
and 48 fugues of the Well-Tempered Clavier featuring an initial tempo
indication.1 There must be a good reason for Largo to be marked here. It
could be Bachs warning against treating the continuous dotted rhythm in
too lively a manner. The tempo is thus measured in slow quarter-notes, and
the mood can correspondingly be described as stately, grave, or solemn.
The appropriate articulation is an overall legato. This applies to all voices,
be they melodically relevant or mere accompaniment like the ascending
eighth-notes in m. 2. The usual exceptions in cadential-bass patterns occur
here only in mm. 8-9 and 19-20. In all other cases they seem conscientiously avoided by ties that make detached playing impossible (see mm.
4-5, 10-11, and 12-13).
1
The others are the B-minor fugue I/24 (Largo) and the B-minor prelude II/24 (Allegro).
485
486
WTC II/16
A number of ornaments are to be considered. Two melodic mordents

appear in the first measure. Both are approached stepwise, thus beginning
on the main note and encompassing only a simple three-note shake. These
ornaments recur in the sequence of m. 2. Beyond that, they are not specifically indicated but, as thematically integrated embellishments, should
be transferred to all further entries of the model (i.e., to mm. 5, 6 and 9).
The remaining ornaments appear almost arbitrary as they are neither
melodically induced nor attached to cadential formulas. Their execution
poses no problem: both the inverted mordent in m. 11 and the mordent in
m. 15, which is approached stepwise, are three-note embellishments, while
that in m. 8 is a four-note ornament launched from the upper neighboring
note. The only pre-beat ornament in this composition, the inverted mordent
E-D-E in the middle of the final measure, is written out, as seen in the
32nd-notes in m. 212.
As the rhythmic pattern in this prelude is so uniform, all attention is
captured by the melodic contours and the texture. Two different textural
levels can be distinguished. One is the complementary interplay between
a melodic voice and a short insert in an adjacent voice. These quasipolyphonic features will be dealt with when the models are discussed. The
other level is realized by the pedal notes which, only occasionally hidden
behind rhythmic patterns, pervade the entire piece. Together with the six
cadences, these pedal notes constitute the backbone of the composition.
The only measures not rooted in pedal notes present a feature that is equally
suited to create large-scale contexts: a gradually descending peak-note line
in the uppermost voice linking the root of the subdominant chord (C) to
that of the tonic (G). The following example shows a version of the bass
line that is analytically simplified to facilitate structural understanding, the
section ending cadences, and the peak note line:
I
mm.
II
III
IV
11
13
VI
15
17
19
21
The preludes thematic material is determined by three models. Each of

them consists of two or even three voices that combine in a rhythmically
complementary way. M1 begins after the initial downbeat in the bass and
spans one measure. It consists of a head, a complement, a climax, and a
G minor
487
tail. (Several of these segments recur separately, which is why they are
named here.) The head M1a aims for E , the sixth scale degree that, in
the minor mode, represents high tension. The complement M1b ends with
a descending tritone, one of the high-tension intervals that in Bachs time
often characterized highly emotional pieces in slow tempo. The climax is
marked by the largest interval leap, C-A, which is particularly conspicuous
as it emerges from a descending line. The tail M1c provides the expected
relaxation with a gentle stepwise figure that, were it not for the dotted
rhythm, could be read as a written-out turn. The texture of this model can
be described as two-part polyphony with a pedal in the lowest voice. The
dynamic outline of M1 results from all the details stated. Tension grows in
M1a, is enhanced by the tritone in M1b, peaks on the high A, and resolves
in the gentle curve of M1c. The model is imitated a fifth up in mm. 2-31.2
It also recurs in mm. 5-61, 6-71 (with additional filling voices and the
complement now in the lowest part), and in mm. 9-101. M1a, the head
alone, is quoted several times; see mm. 7-81 (4x), mm. 8-9 (as a link at the
end of the cadence), mm. 11-121 (3x), m. 16 (2x), mm. 173-183 (4x, a varied
transposition of mm. 7-81). M1b recurs in two crucial positions: in mm. 1516 it marks the beginning of the peak note line and in m. 19 its end. M1c
exerts its soothing power as one of the complements to M2 (see below).
Moreover it also appears in connection with the beginning and end of the
peak-note line (twice in m. 15, once in m. 19).
M2 is introduced in mm. 3-41. It is related to M1 in two ways: its final
figure (see mm. 3-4: D-A-C-B) quotes M1a and its second complement
(see m. 3, alto: G-F-E-F) takes up M1c, both transposed into a majormode context. Easily overlooked upon first reading is the rhythm of the
first complement (tenor: m. 31-2), which replaces, for the first time in the
piece, the second 32nd-note of the rhythmic unit with a 16th-note. The two
complements in different voices give the impression that this model is set
in three-part polyphonic texture over the pedal D in the lowest part. The
dynamic design of M2 is once again a simple curve, with the climax here
on beat 3 where C appears as the apex of the leading line. Owing to a lack
in high-tension intervals, this curve is much gentler than that in M1. M2 is
sequenced in mm. 4-51 a fourth up and recurs in mm. 10-111. In these two
measures, the principal voice and its main complement are note-identical
but the former pedal in the bass now takes a more active part.
2
As the notation in m. 2 reveals, the uppermost voice in m. 1 is not the highest to appear in
this piece. A neat distinction of voices is, however, not very meaningful since Bach has not
conceived this prelude in consistent part writing.
488
WTC II/16
M3 first appears in mm. 132-142. In the leading voice, two four-note

figures appear combined to a single broken-chord pattern that climaxes on
the syncopated D. The motif is accompanied by a free parallel and complemented during the syncopation by a little duet in alto and tenor. Its
sequence is even more intense, representing the first and only passage in
the prelude to sound in consistent four-part texture.
The preludes structure can be described as follows:
Section I is thematically active. It presents M1 and M2, both with
an immediate recurrence. Section II recalls, in its first half, M1 with
its imitation. Its second half quotes only segments of the models.
Section III, with two measures half as long as the previous ones,
presents a contracted version of section I and therefore appears
slightly more intense than section II. Section IV, almost as brief,
like the latter half of section II quotes only segments of the models.
Section V appears as a new beginning with a new model. Its importance is underlined with the strongest crescendo in the piece and its
extended scope with a peak-note line. Section VI is but a coda.
The two tripartite blocks of the harmonic design are thus counteracted
by the three binary blocks of the thematic and dynamic layout.
WTC II/16 in G minor Fugue

Beginning on the second beat in 3/4 time, the subject of this fugue
spans a little over four measures. It concludes after an appoggiatura and its
ornamented resolution on the first 16th-note of the second beat in m. 4.
Both the pitch of the first note, the fifth degree of the G-minor scale, and
the unusual rhythmic beginning on a weak-beat quarter-note followed by
a rest create a feeling of elevated tension on the very first note of the fugue
and thus discredit our usual notion of a weak-beat beginning as upbeat.
After this unique opening, the phrase unfolds in the most regular
pattern. The three subsequent downbeats are each prepared by an eighthnote anacrusis in sequencing intervals of ascending fourths (B-E, A-D,
and G-C). While the second of these upbeats opens a subphrase that is the
exact transposition of the one preceding it, the third initiates a unit with
internal extension: the upbeats target C is expanded by means of a sevenfold note repetition up to the next downbeat. There it turns into the abovementioned appoggiatura and then descends to its ornamented resolution
(C[B-A]-B).
G minor
489
The rhythmic pattern within the subject is dominated by eighth-notes

and quarter-notes. While the counter-subjects and all motifs add 16th-notes
and occasional dotted notes, an overall impression of rhythmic simplicity
remains. The pitch pattern is characterized by leaps and note repetitions in
the longer note values, and by ornamental figures and scalar passages in
the 16th-notes (see, e.g., the inverted-mordent figures at mm. 53, 63, 73 and
the chain of written-out mordents in m. 8). Bachs harmonization shows a
swift harmonic change on the
initial syncopation followed by
measure-by-measure cadential
7
9
V -i iv
i
V
i
steps.
The dynamic outline, too, is simple. Launched from the fifth degree,
i.e., from raised tension, the climax falls on m. 21, which represents both
the peak of the descending sequences and the subdominant harmony. The
overall tension then decreases gradually through the smaller climaxes of
the ensuing subphrases (mm. 31 and 41) and finds a perfect release onto the
tonics third at m. 52.
The G-minor fugue comprises the following 17 subject statements.
1.
mm. 1-5 T
9. mm. 45-49
T
2.
mm. 5-9 A
10. mm. 45-49
A
3.
mm. 9-13 S
11. mm. 51-55
A
4.
mm. 13-17 B
12. mm. 51-56
S
5.
mm. 20-24 T
13. mm. 59-63
B
6.
mm. 28-32 A
14. mm. 59-63
T
7.
mm. 32-36 S
15. mm. 67-69 (75)
T
8.
mm. 36-40 B
16. mm. 69-73 (75)
S
17. mm. 79-83
B
The modifications the subject undergoes in the course of its 17 entries

are rather unusual. Instead of strettos there are four statement parallels.
Modifications of shape include common processes like the adjustment of
the initial interval in the answer (mm. 5, 13, and 36) but also the surprising
variation of all upbeats in the final entry. The subjects tail appears in
various forms: shortened to an unornamented beat-one resolution (A: mm.
32 and 49) or emerging from a note repetition with tie prolongation and
subsequent diverted resolution (m. 36), with a varied extension of the note
repetition (A: m. 55) or with a newly ornamented resolution (A: m. 56 and
T: m. 63). Besides the free variations of the ending in the parallel subject
490
WTC II/16
entries (see mm. 55-56: S and m. 63: B), the most drastic modifications
occur in mm. 67-75. The soprano statement, after beginning regularly in
mm. 69-72, features a delayed resolution which, enveloped in the metric
pattern of a hemiola, finds its satisfactory conclusion only at the end of a
cadential formula at m. 751. The preceding tenor statement can either be
read as shortened (i.e. without the entire final subphrase, in which case it
would break off on the second beat of m. 69), or, more likely, as extended
in its middle by two extra sequences (see mm. 67-68 sequenced in 68-69,
69-70, and 70-71) and then complemented by a free and again very much
extended version of the final subphrase.
Bach invents only one true counter-subject for this fugue. Introduced
in mm. 5-9 in the tenor, it displays a phrase structure strikingly similar to
that of the subject without its initial note: a one-measure subphrase is
followed by two descending sequences the second of which is extended to
two-measure length. The brackets in the example mark this similarity:
The almost parallel design of subject and counter-subject permits little

independence in the dynamic layout. All one can do to maintain a certain
degree of distinction is to place the climax in each of the counter-subjects
subphrases on the inverted mordents. The tie in mm. 7-8 supports this
solution which, were it not for the sake of more polyphonic clarity, would
appear somewhat unusual from a metric point of view.
The fugue contains eight subject-free passages.
E1 mm. 17-20
E5 mm. 56-59
E2 mm. 24-28
E6 mm. 63-67
E3 mm. 40-45
E7 mm. 75-79
E4 mm. 49-51
E8 mm. 83-84
The subject-free passages feature three explicit cadential formulas,
ending on the subdominant (E3, mm. 44-45), the dominant (E6, mm. 66-67:
imperfect cadence), and the tonic (E8, mm. 83-84). The short final episode
consists entirely of a cadential close. In E3, the close in C minor displays
a suspension. Its resolution onto the third overlaps with the ensuing subject
entry (see A: m. 45). In E6, the incomplete harmonic progression ends
without any suspension.
G minor
491
E6 is related to E8 in another way, one that distinguishes both from the

fugues other subject-free passages. Instead of moving from the clearly
discernible ending of the preceding subject entries to the typical episode
motif, E6 and E8 unfold almost imperceptibly as extensions of the preceding S/CS parallel. The varied subject tail in the tenor, which substitutes
the note repetition + inverted-mordent figure by a four-note mordent, is
sequenced in such a way as to form an imitation of the threefold mordent
figure characterizing the counter-subjects fourth measure. At the same
time, the bass imitates the preceding soprano descent. The following two
measures mix further imitation of the threefold mordent with a sequence,
in the bass, of the varied subject ending.
In all other episodes, the characteristic opening of the counter-subject,
which will be referred to as Mcs, plays a major role. The three-note ascent
followed by the inverted mordent figure (from T: mm. 5 or 6) is complemented in various ways. The other parts active in each episode present
various figures that are either sequenced or imitated but often not taken up
again at any other instance in the fugue. Phrasing and local dynamics are
essential in shaping these episode components.3
The role each episode plays in the overall dynamic development is
determined by the direction of sequences and imitations. E1 presents a twofold relaxation (mm. 17-191 and, overlapping in the soprano, mm. 19-20).
E2 is conceived with ascending sequences, thus creating a slight increase in
tension. E3 appears as a dynamic curve. An initial increase, which is
straightforward owing to the fact that all voices join in the ascending
sequences (mm. 40-43), is followed by a decrease and complete release in
the first explicit cadential close. E4 links two consecutive parallel statements in increasing tendency, while E5 with its descending sequences and
falling chromatic lines creates a decrease. In E6, the tension grows
throughout the partial sequences of the preceding statement, only letting up
in the approach to the imperfect cadence. After the soprano statements
cadential extension, E7 begins more softly than any of the previous
episodes but builds up even more power, in immediate preparation for the
ensuing final subject statement.
This fugues basic character is conceived as rather lively on account of
the uncomplicated rhythm and the contours featuring leaps, note repeti3
The bass line in E1 begins after phrasing in m. 17 with the 16th-note D. There follow a
climax at m. 173, phrasing in m. 182 after E, another climax in m. 183, and another phrasing
in m. 191, etc. In E2 the climaxes fall on T: m. 243, S: m. 253, T: m. 263, and S: m. 273; in
E7 on B: m. 752, S: m. 753, T: m. 761, A: m. 762, S+T: m. 763, and S+B: m. 772.
492
WTC II/16
tions, and ornamental figures. The articulation requires quasi legato for all
melodic 16th-notes except for written-out ornaments, which should be
legato, and non legato for all eighth- and quarter-notes. It is a good idea to
distinguish the quality of touch in the cadential versus the melodic eighthnotes (e.g., in B: mm. 9 and 24-28). Note-pairs consisting of appoggiatura
and resolution are always legato; this applies also to the tied prolongations
in mm. 14 and 15. Phrasing (e.g., T: in m. 10 after A, m. 11 after G, m. 14
after A, and m. 15 after G) is best expressed through dynamic means.
The tempo of this fugue is moderately fast. The 16th-notes provide the
measure for an appropriate pace: they should be swift enough to convey
their ornamental character but not so brisk as to jeopardize their melodic
quality. Transparency must be guaranteed as well as the possibility for
unhurried phrasing between 16th-notes (as is demanded, e.g., at the
juncture of subject ending and counter-subject beginning). The tempo
proportion between the prelude and the fugue uses larger metric units for
translation: a quarter-note in the prelude corresponds with a whole measure
in the fugue. Approximate metronome settings: prelude beats = 28 (eighthnotes = 56), fugue beats = 84.
The only ornament symbol in the score is found in m. 21. The mordent
is approached stepwise. It thus begins on the main note and consists of a
single three-note shake. Beginning this shake with two 32nd-notes, so that
the return to the main note falls on the bass note C, represents a perfectly
acceptable and clear solution. The rendition of the ornament as a triplet
concluding before the bass motif begins is correct but risks blurring.
In terms of its layout, the G-minor fugue falls into two almost equal
halves (44 + 40 measures). While this major dividing line is beyond doubt,
the distinction between sections in each half appears blurred by contrary
information from within the musical text. Here are the details. The first
half contains eight subject statements in T A S B T A S B. The episode
following the last of these entries describes a self-contained dynamic curve
and concludes with a perfect cadence in C minor. The second half presents
parallel entries in the three neighboring voice pairs T+A, A+S, and B+T as
well as the stretto TS with a varied ending. The entering order of the
leading voice in each pair, T A B S, is almost identical with the order
observed twice in the first half of the fugue.4 The final entry in the bass
4
The leader in a parallel statement is traditionally the one which is placed on the scale
degree identical with that of the original subject. Thus in the first pair which is
harmonically in F major, the tenor begins correctly on the fifth and carries on accordingly,
while the alto, doubling in thirds, does not maintain the original interval structure.
G minor
493
seems the odd one out in several ways: it is the only statement in the fugue
that is considerably varied, the only one to contain a harmonically effective
modification of the pitch pattern (B), and it follows after two conspicuous
homophonic formulas in hemiola pattern, thus appearing more like a coda.
To sum up: Bach introduces a new contrapuntal technique in the second
half of the fugue, which he then uses consistently through four entries and
four different voice combinations. The single section comprising four
parallel statements is then followed by a redundant entry.5
In the first half, the repeated entering order suggests two complete
rounds of statements, but the texture seems to impart a different message.
In mm. 20-24, the tenor statement retains the full four-part ensemble
established in the preceding bass entry throughout the subjects first and
second subphrases. The ensuing subject statements in alto and soprano,
however, appear in reduced ensemble as the bass is resting in mm. 28-36.
Yet since neither the harmonic development nor a distinct cadential close
support one or the other solution, we can only deduce that Bach chose to
leave this boundary ambiguous. What he creates amounts to a double
section in which the dividing line can be determined by careful analysis
alone but does not constitute, particularly for listeners, a structural caesura.
In a way, the composer is thus matching the long section in the fugues
second half with a combined unit in the first.
The development of tension in the fugues first half thus consists of
two increases. Interpreters seeking to emphasize the identical entering
order will tone down the second tenor statement despite its initial four-part
ensemble, in contrast to those who perceive a beginning of the second
section only at the point of textural reduction. The fugues second half
begins with a kind of explosion: in full four-part ensemble and with the
5
From m. 59 onward, Bach introduces a yet more intensified contrapuntal technique:

parallel statements of the counter-subject in addition to parallel statements of the subject.
Some analysts take this fact as an indication for the beginning of a new section in m. 59.
Others regard the imperfect cadence in mm. 66-67, together with the ensuing thinner
texture with only two voices, as a structural caesura. Yet it seems preferable to recognize
both features as means employed to create further intensification. We can state a gradual
buildup of tension through this long section: it begins with the parallel subject statement,
continues in the twofold parallel of subject and counter-subject, and culminates in this same
twofold parallel prolonged by a two-measure anticipation of the accompanying parallel
voices. The reduced texture in mm. 67-68 thus appears as a short repose before the final
triumphant outbreak: the twofold parallel in the home key. The fact that the anticipated
subject statement in the tenor appears already in the expected key (compare T: mm. 67-68
with S: mm. 69-70) creates an effect as if stretto and twofold parallel were combined.
494
WTC II/16
intensity of parallel statements. The interspersed episodes each set out

more softly but soon return to the high intensity level. Only E7 presents a
true relaxation before the redundant entry concludes the fugue on a gentle
note.
WTC II/17 in A major Prelude

This prelude is based on several motifs. Some of them spawn the
imitations and modifications typical for polyphonic material, while others
are presented with an accompaniment or with an answering motif in the
other voice. The texture is conceived in two parts both of which split
regularly to form chordal patterns.
The first harmonic progression ends with the confirmation of the home
key at m. 71. This close marks a structurally relevant point. In view of the
large-scale design of the prelude, however, one might designate it as the
end of a sub-section rather than a fully developed section since the more
convincing section ending appears, complete with a cadential formula, in
mm. 16-17. There are altogether five sections in this piece:
I mm. 1-171 tonic to dominant (A major to E major)
(1-7: tonic confirmed, 7-171: modulation)
II mm. 17-341 dominant to tonic relative (E major to F minor)
(17-231: D confirmed, 23-341: modulation)
III mm. 34-501 tonic relative to subdominant (F minor to D major)
(34-401: tonic relative confirmed; 40-50: modulation)
IV mm. 50-641 subdominant to tonic (D major to A major)
(50-632: modulation to V of A, 632-641: link)
V mm. 64-77 tonic confirmed
There are extended structural analogies in this piece.
mm. 1-4 . 17-20 and 34-37 (transposed and varied)
mm. 1-2 . 50-51
(transposed)
mm. 5-6 . 21-22 and 38-39 (transposed and varied)
mm. 7-91 . 40-421
(transposed and varied)
mm. 11-171 . 44-501
(transposed and varied)
The basic character is rather calm. The rhythmic patterns are complex,
featuring a large variety of note values from dotted and tied eighth-notes to
32nd-notes. The main rhythmic feature is the long on-beat note followed
shortly before the next beat by one, two, or three fast notes. This gives the
piece a swinging momentum, preventing too slow a tempo.
The articulation is mainly legato. Most of the interval leaps occur
either inside 16th-note patterns or between 32nd-notes and the following
longer note, i.e., in instances where detached playing is ruled out. The
495
496
WTC II/17
wedges indicated in mm. 5 and 6, which specifically demand abbreviation

of the long notes, give performers to understand that the dotted notes in
this piece are generally meant to be taken legato. As with all indications
regarding articulation or ornamentation within thematic material, these
wedges should be transferred to corresponding notes. This definitely applies
to the G in m. 393. Whether the same holds true for the A in m. 213
depends on the individual performers interpretation of this measure
which may or may not be perceived as deriving from m. 5.
Conventional exceptions from the general legato touch occur above all
in cadential-bass patterns (mm. 16, 33, 49, 62, and 74). Other exceptions,
in which articulation is explicitly marked, also occur outside the thematic
material in cadential contexts. In m. 62, paired slurring indicates dynamic
shading in an active-passive pattern: these slurs reinforce the harmonic
feature of appoggiatura-resolution, rather than indicating an abbreviation
of the second note. In m. 76, the slur linking the quarter-note A to the
subsequent G serve the same function and additionally guards against any
possible interruption of the dosido formula.
The A-major prelude features three types of ornaments: two gracenotes, an inverted mordent, and several mordents. Mordents occur in two
different contexts. In mm. 52-59 they function as thematically integrated
ornaments appearing before the backdrop of an accompaniment in regular
16th-notes. Each of these mordents is approached in stepwise motion, thus
beginning on the main note and containing only a single three-note shake.
The mordent in m. 76 is a cadential ornament. As it appears with no other
rhythmic features against it and marks the essential step in the final ritardando, this mordent can be more elaborate. Beginning equally on the main
note, it may contain five or seven notes before it stops short. The inverted
mordent highlights the cadential-bass pattern immediately preceding the
final homophonic formula at m. 743. In accordance with the harmonic
context prevailing in this measure it must be played with C as its lower
neighbor note. The two grace-notes both appear in m. 75, as part of the
final homophonic cadence. Both are indicated as eighth-note appoggiaturas
to a main note of quarter-note value. In both cases, the graces are thus
played on the beat, together with the chordal notes in the right hand and
the bass notes, and resolve after one eighth-note into the harmonically
anticipated notes G and E respectively.
The variety of motivic material in this prelude allows for different concepts, and by extension, different ways of labeling. On the one hand, a
distinction of all similar but not-quite-identical shapes would lead to so
large a number of individual shapes that performers might lose track and
A major
497
the benefit of an analysis would thus be thwarted. On the other hand,

recognizing too many different shapes as variations of a single idea might
render analytical observations meaningless. The following discussion tries
to eschew both traps by taking into consideration not only the actual
appearance of a melodic unit but also its texture and context.
M1 is presented in mm. 1-21. It consists of two manifestations of the
broken chord. In the left-hand part, a descending A-major chord appears in
a linear pattern (M1a). The skips are not bridged by the fast notes following
the dotted strong-beat notes but widened. Its homophonic companion M1b
presents the A-major chord in a pattern that focuses on off-beat block
chords preceded by auxiliary notes in the dosido pattern. Dynamically,
M1 describes a diminuendo, caused by the absence of any harmonic activity in this motif and enhanced by the descending direction of the melodic
part M1b. M1 recurs similarly in mm. 3, 17, 19, 34, 36, and 50. A variation
can further be found in m. 10. Moreover, M1b appears separately and in
slight variation in mm. 52, 54, 56, 58, and 59. The motif and its recurrences do not trigger any imitation or inversion of voices: the melodic line
always remains in the left-hand part and the accompanying chords are
restricted to the right-hand part.
M2 is an almost ubiquitous figure in this piece. Introduced in m. 2 and
appearing in a variety of pitch patterns,1 its discerning feature is the singlevoiced contour in a rhythm of regular 16th-notes beginning after a downbeat rest and ending on the next downbeat. All variants of this motif share
the dynamic layout with an initial crescendo to beat 2 followed by a
relaxation to the downbeat. In contrast to M1, which remains confined to
its initially homophonic setting, M2 changes voices and joins a number of
other motifs.
M3, whose original version is introduced in the lower part of mm. 7-81,
consists of a three-32nd-note upbeat followed by a scalar descent in dotted
rhythm. As in M2, the dynamic climax falls on beat 2 of a measure, and
just like M2, M3 comes in many facets.2
1
A list of the relevant appearances could look like this: M2a see mm.2, 4, 18, 20, 35, 51;
M2b see L: mm. 5/6, U: mm. 21/22, L: mm. 38/39; M2c see U: mm. 7-9, 29-31, 40-43, 68,
70/-73 and L: mm. 24, 26, 28, 32, 65, 67, 69; M2d see U: mm. 23, 25, 64, 66; M2e see U:
mm. 53, 55, 60; M2f see U: mm. 112-121, 122-151, 442-481.
M3a see L: mm. 7-101, 40-441 (always as a counterpart to M2c); M3b see L: mm. 12-151,
45-481, 64-671 (the latter against M2d); M3c see L: mm. 52-601 (alternatingly against a M1b
variant and against M2e); M3d see L: mm. 23-241, 25-261, 27-281, and as a partial quotation
in L: mm. 29-31 and U: mm 32-33, then again in U: mm. 65-661, 67-691, 70-741 (as a counterpart to M2c or M2d).
498
WTC II/17
Finally, M4 differs from the three previously mentioned motifs in that

it is chordal and not linear (see mm. 24-251). It is thus closely related to the
homophonic accompaniment within M1. This motif appears not as an
active component in its own right but as a complement to M3c, which thus
grows into a two-measure figure and only drops this complement in the
liquidation process from m. 29 onward.
The following table gives an overview of the structural design in the
prelude as it results from the interplay of the three motifs and their variations, as well as showing the analogy of the first three sections.
Section I
M1a/M1b + M2a
M1a/M1b + M2a
M2b/M2b var
M3a/M2c + sequ.
M3b/M2f + sequ.
cadence
Section II
M1a/M1b + M2a
M1a/M1b + M2a
M2b/M2b var
M3d/M2d, M4/M2c + sequ.
M3d/M2c + sequ. + liquid.
cadence
Section IV
M1a/M1b
M2a
M3c/M1b, M3c/M2e + sequ.
expanded cadence + bridge
Section III
M1a/M1b + M2a
M1a/M1b + M2a var
M2b/M2b var
M3a/M2c + sequ.
M3b/M2f + sequ.
cadence
Section V
M3b/M2d, M3d/M2c + sequ.
M4 var/M2c var
M3d/M2c + sequ.
expanded cadence
WTC II/17 in Amajor Fugue

The subject of the A-major fugue spans exactly two measures. Beginning after an eighth-note rest on the fifth scale degree, the phrase reaches
its cadential conclusion at m. 31, in a gentle ending on the third. The harmonic background is simple.
Its most significant feature is
the progression from the subI V I IV
V
I dominant to the dominant in
the syncopation.
The rhythmic variety in the subject comprises eighth-notes, 16th-notes,
and a syncopation of 5/16 duration. Later in the course of the fugue,
quarter-notes are added as regular features. Yet despite the presence of
these four different note values, the overall impression is one of rhythmic
simplicity. This is due primarily to the fact that while the subject and its
A major
499
answer feature beats that are only once or not at all subdivided, the fugal
voices thereafter complement one another in such a way that they create a
continuous motion up to the general pause in m. 46. In fact, all through
these inner 42 measures the 16th-note motion only misses four beats:
one each in mm. 13, 14, 16, and 19.
The subjects pitch contour features a number of consecutive leaps:
three of them appear at the beginning in increasing interval spans (third,
fourth, and fifth), two more are added before m. 21 (octave and fourth). In
this rhythmic shape, the two ascending tetrachords (m. 1: B-C-D-E,
m. 2: A-B-C-D) constitute popular formulas. Finally, the 16th-note figure
filling the latter part of m. 2 has ornamental character, embellishing the
suspended resolution D-C. It may be interesting to visualize this subject
stripped of all its ornaments (including, for this purpose, the octave split).
The emerging skeleton is striking with regard to both its interval pattern
and its rhythm: the subject turns out to be based on a progression of rising
fourths and falling fifths,
evolving in a pattern of
gradual rhythmic augmentation.
The phrase structure in the subject can be interpreted in two ways,
depending on whether one seeks to emphasize the structural details or the
harmonic and rhythmic features. Performers choosing to stress the fact that
the tetrachord in the first half of m. 2 is a sequence of that in the second
half of m. 1 will render the subject as consisting of two subphrases that are
divided by an octave leap. Its phrase structure may then be described as a
three-note upbeat + ascent followed by a single-note upbeat + ascent + tail
(ornamented resolution). Conversely, performers regarding the syncopation in m. 2 as the subjects salient featurea metrically highlighted note
that is also harmonically alivewill interpret the subject as an undivided
unit encompassing two consecutive upward thrusts that peak in the
syncopation. The dynamic design differs accordingly. If the emphasis is on
the sequence of two tetrachords, the three-eighth-note upbeat leads to a
first climax on B, followed by a decrease on the way up to E. The lower
E then provides the active upbeat to the second climax on the A, which
launches a gradual relaxation throughout m. 2 to the final C. Conversely,
if the emphasis is on the syncopation, there is only one climax, reached in
an unbroken crescendo and followed by a diminuendo of almost equal
extension. Especially for a lively interpretation of this fugue, the second
option for phrasing and dynamic shaping is probably preferable.
The fugue comprises fifteen subject statements.
500
WTC II/17
1.
2.
3.
4.
5.
6.
7.
mm. 1-3
mm. 3-5
mm. 6-8
mm. 8-10
mm. 13-15
mm. 16-18
mm. 18-20
A
S
T
B
B
A
T
8.
9.
10.
11.
12.
13.
14.
15.
mm. 22-24
S
mm. 24-26
A
mm. 32-34
T
mm. 35-37
S
mm. 37-39
B
mm. 41-43
T
mm. 42-44 (45) B
mm. 48-50
T
In addition to the adjustment of the initial interval in the tonal answer,

the subject undergoes several minor changes. The fourth and fifth as well
as the eleventh and twelfth statements end on the keynote instead of the
third degree. This is a modification frequently used by Bach for bass
statements, so the only surprise is that in this fugue it also appears once in
the soprano (mm. 36-37). Furthermore, two statements feature an artificial
leading-note in the first tetrachord that is later corrected (D instead of D
in B: m. 14 and T: m. 41). The final statement, while difficult to read (and
hear), is presented as an unmodified answer.
More significant modifications occur in mm. 42-44. The bass entry not
only overlaps with the preceding tenor entry for half a measure, but also
features drastic harmonic alterations that transpose its final half-measure
onto the key of D minor. The subjects final note does not, however,
provide the expected resolution, so that the subsequent partial sequence
must be regarded as an integral part of this entry. At m. 451, this extended
subject statement concludes with an interrupted cadence (the B-major
chord is VI of D minor). Similar but not quite so drastic harmonic alterations occur in the bass statement of mm. 37-39. Beginning as an implied
D-major entry, the sudden Fs toward the end suggest a turn toward D
minor. But then the final chord, expected to confirm this key, is accompanied by an unresolved voice and allows for subsequent modulations,
here again wrapped into a partial sequence (see at m. 394 the sopranos
closing formula leading to a resolution in A minor).
Bach has invented two counter-subjects for this fugue. CS1, introduced
in mm. 3-5, accompanies all further statements. It consists (after a 3/16
bridge that is not part of the component) of a chromatic descent in quarternotes leaping back up to a dosido formula. Dynamically, CS1 describes
two relaxations, setting in after climaxes on the initial note and on the
closing formulas syncopation. The two diminuendos form a striking
A major
501
contrast to the crescendos dominating the subject. Frequent variants of this

counter-subject, in which the leading-note is protracted by means of a tie,
sometimes additionally flattened or even flattened and diverted downward
(see S: m. 15 and T: mm. 23-24, A: m. 20 and S: m. 26, and A: mm. 36-37)
cause a harmonically unresolved ending. In two cases, CS1 swaps voices
(see mm. 15 and 23-24), in two others, it is reduced to its tail (see A: mm.
33-34 and B: m. 18)an abridgment that, as the structural analysis will
show, underlines the analogous positions of these subject statements within
the design of the fugue. In three further cases we hear only the chromatic
descent (see mm. 38-39, 42-43, and 50).
CS2, introduced in A: mm. 6-8, consists of a long stream of ornamental
sixteenth-notes followed by four eighth-notes. It undergoes a modification
process that soon
leads to forms so far
from the original as
to render it meaningless to retain the
labeling. As its contour is designed as a
veiled CS1 parallel,
CS2 does not represent a dynamically
independent contrapuntal component.
The A-major fugue contains eight subject-free passages.
E1 mm. 5-61
E5
mm. 26-321
E2 mm. 10-133
E6
mm. 34-351
E3 mm. 153-163
E7
mm. 39-411
E4 mm. 203-221
E8
mm. 45-483
Two of these episodes are subdivided. E5 encompasses a cadential
close in C minor that separates E5a (mm. 26-273) from E5b (mm. 273-321).
In E8, the two halves are separated conspicuously by the general pause on
m. 463. The episode material features derivations from subject and countersubject as well as independent motifs. The subject head in 6/8 or 4/8
extension repeatedly anticipates a complete subject statement (see A: m. 5,
A: m. 34, S: mm. 34-35, and B: mm. 203-211), and its 16th-note tail follows
some statements as a sequence or an imitation (see T: mm. 15-16, B: m. 16,
A: mm. 20-21, T: m. 21, A: mm. 21-22, 26, 26-27, 27, T: mm. 34, 34-35,
B: mm. 39, 39-40, S: m. 40, and T: mm. 40-41). The counter-subjects
closing formula is sequenced in T: mm. 20-21, A: m. 21, and T: mm. 21-22.
502
WTC II/17
M1, introduced in S: m. 5, is developed in B: m. 10 with six sequences

(103-133) and in S: mm. 27-28 with sequences and imitations (S: mm.
28-291 and A: mm. 29-303-321). M2 (S: mm. 10-113) recurs in A: mm.
113-131, S: m. 13, and B: mm. 273-291 with imitations in S: mm. 29-303
and A: mm. 303-321. M3, the contrapuntal companion of M2, is heard in A:
mm. 102-111 with imitations in S: mm. 114-123 and in A: mm. 274-283 with
imitations in B: mm. 293-301 and S: mm. 304-313.
Two of the episodes feature cadences with explicit formulas: E3 closes
at m. 173 in F minor and E5a at m. 273 in C minor. In E7, the unresolved
D-minor cadence of the subject entry is corrected half a measure later
with an explicit formula in the soprano toward A minor. The subsequent
sequence of the subjects tail leads into E minor (m. 402), which makes
way for E major, the dominant of the home key. E8 is the only episode not
to use any of the previously introduced motivic material. Its largely homophonic texture sets it apart from the remainder of the fugue. In E8a, a
broken-chord figure in the bass accompanies metric block chords in a
harmonic progression from B major to E9. E8b follows with a two-part
texture in contrary-motion that reconfirms A major but is diverted with
another homophonic pattern into an interrupted cadence.
The role played by the episodes in the dynamic development can be
described as follows: E1 and E6 both bridge between consecutive entries.
Due to the prominent appearance of the subjects head, the level of tension
falls only slightly after the subject statements and rises again for the subsequent entry. E2 and E5b occupy a plane that is contrasted to that of the
subject statements and the other episodes. Therefore, these two passages
are best distinguished from their surroundings by a distinctly different tone
color and touch. E3 and E5a pick up the tension remaining at the end of the
preceding subject statements which, due to the unresolved counter-subject
ending, has not reached complete relaxation. The two episodes then
complete the dynamic decline. Similarly, the falling sequences of the
subjects tail in E7, together with the one-by-one drop out of voices and
the harmonic resolution, also generate the effect of closure. E4 and E8 are
the only subject-free passages to create an increase in tension. In E4, the
basic intensity is generated by episode material stemming exclusively from
the subject and counter-subject, the growth by ascending sequences and
imitations. In E8a, the change of texture from dense four-part polyphony to
homophony creates a strong and sudden outbreak of almost virtuoso intensity, enhanced by the contrary motion and the impact of the harmonic step
from B major to E9. By contrast, E8b, beginning after a general pause,
remains fairly soft.
A major
503
Given the overall effect of rhythmic simplicity combined with the

predominance of interval leaps and ornamental 16th-notes in the pitch
pattern, the basic character of this fugue is rather lively. The tempo should
not be too fast, both in order to avoid the effect of superficial runs without
melodic contours in the 16th-notes and to give listeners a chance to fully
appreciate the melodic closing formulas and their occasional deflection.
Owing to the greatly different character and mood of the prelude and fugue
in A major, a simple proportion can be chosen tempo relation between the
two pieces without risking dullness: an eighth-note in the prelude corresponds with a quarter-note in the fugue. (Approximate metronome settings:
prelude beats = 44, fugue beats = 88.)
The articulation demands careful nuances.3 The shortest and most energetically bouncing non legato should be applied to those eighth-notes in
the subject that do not form part of an ornamental tetrachord.4 A not quite
so energetic and slightly longer non legato applies in the cadential-bass
patterns and in other non-melodic eighth-notes in cadential style.5 An only
gently detached non legato is appropriate for the chromatic descent in the
counter-subject. The eighth-note leaps in M3 and the initial fourth in M2
sound most convincing if their color is lighter than the subjects, graceful
rather than energetic. The chromatic descent in M2 can be taken either in
dense non legato or, perhaps preferable in view of the desired color contrast,
in legato. All melodic formulas are legato. This applies to the subjects two
rising tetrachords, the closing formula in the counter-subject, and the
cadential formulas in S: mm. 16, 39, 44-45, and 50). If the countersubjects closing formula is deflected, it should nevertheless retain the
legato touch to mark its melodic origins (S: m. 15, B: m. 18, A: m. 20 with
imitations in T/A/T, as well as T: m. 23, S: m. 25, and A. mm. 36-37).
While the ornamental 16th-notes in the subject are legato, those in M1 and
M2 benefit from a crisp quasi-legato touch. The cadential trill in S: m. 48
begins regularly from the upper neighbor note, shakes in 32nd-notes, and
ends without rhythmic interruption in the suffix Bach has spelled out.
3
Distinctions may seem small and, particularly to impatient performers, hardly worth the
effort. Yet they do make all the difference in this fugue where color contrastsbetween
subject and counter-subject on the one hand and between primary material and independent
episode material on the otherare essential.
To avoid confusion, here is the articulation in the subject: E-C-F non legato, B up to E
legato, both Es non legato, A up to D and all remaining 16th-notes legato.
5
Cf. B: mm. 15 A-C-D-A, m. 20 G-A-D-E-A, mm. 26-27 B-E, A-D-F-G-G-C, and

mm. 47-50, all eighth-notes and quarter-notes.
504
WTC II/17
The fugues structural layout is easy to grasp. The entering order of the
subject statements, combined with cadential endings followed invariably
by a reduction of the ensemble, creates a particularly unambiguous design.
The only irregularity consists in the position of an episode at the beginning
of a section. These are the details:
Section I comprises the initial round of subject statements (A S T B)
connected by E1. It is followed by the (self-contained) E2, a redundant bass entry, and the closing E3. While all five entries are in
A major, in regular alternation between dux and comes (tonic and
dominant), the closing episode, which concludes the section on the
middle beat of m. 16, modulates to the relative minor key.
Section II features four subject statements, two pairs bridged by E4,
the episode closest to the thematic material. The subject statements
are launched once more from the home key, repeating the modulation to F minor, this time in the final entry. This alto statement is
conceived as redundant in the round (A T S A), thereby announcing
the end of the section, which is confirmed in E5a with the cadential
close in C minor.
Section III begins with a five-bar subject-free passage (E5b). It is
followed by three subject statements (T S B) that are linked by E6
and closed by E7. All three entries are in the minor mode (E minor,
B minor, and D minor), representing the minor dominant, the
subdominant relative, and the minor subdominant respectively).
The return to the home key, prepared in the modulatory process in
E7 and confirmed with the tenor entry in m. 41, marks the beginning of the fourth section. This section encompasses three subject
statements, the last of which is again conceived as a redundant
entry (T B T). The harmonic digression at the end of the bass entry
and in the subsequent E8 enhances the impression of redundancy,
giving the fugue a particularly well-rounded ending. This section is
distinguished not only by its use of texturevirtuoso patterns in
one hand, block chords in the other, a rhythmic surprise in the weakbeat general pause, and a four-part cadence ending deceptively
but also by particularities of its internal structure. The four opening
measures feature the only overlap of subject statements in the
fugue. The counter-subject accompanying the two entries involved
in the stretto is varied and merged into a single prolonged chromatic
descent (see m. 41: A to m. 44: B), and the bass entry is extended
up to m. 451. The impression arising here is that of a superimposed
four-measure phrase.
A major
505
The first and second sections are moderate and fairly similar with
regard to their overall intensity level. Thereafter, the third section with its
consistent harmonization in the minor mode appears comparatively subdued, while the fourth section is even more outgoing than the opening.
Within section I there is a gradual increase of tension due to the growth
in the number of voices. E1 constitutes only a short lessening of tension
without any serious interruption in the dynamic build-up from one subject
506
WTC II/17
entry to another. E2, as mentioned above, brings with it a fundamental

change of color as well as, at the outset of the episode, a radical drop in
tension. While the ascending sequences of M1 generate a certain increase,
the redundant bass entry sounds considerably less assertive than its predecessors, last but not least because, abandoned by the alto, it is left in
three-part setting.
Section II repeats the process with different means. Beginning in
unusually reduced ensemble and thus almost as softly as the first section,
the build-up of tension proceeds unhampered through three entries linked
by the intense E4. The last entry of this round resembles that of the
previous section in that it is both redundant and set in three-part texture.
The harmonization in minor completes the picture, so that section II also
ends on a softer note.
What was left out in these two otherwise almost analogous sections,
the self-contained episode, is now made up for at the beginning of section
III. Setting out very softly, this episode presents a gentle tension increase.
The three ensuing subject statements continue this development in a line
that is only transitorily suspended during E6. After the (still moderate)
climax in the bass entry, E7 provides both dynamic relaxation and release
from the minor mode.
Section IV begins with a boost of confidence in a three-part majormode entry. The above-mentioned harmonic digression provokes a climax
in m. 46 that serves as a culmination point not only of this section but of
the entire fugue, followed by the softer curve of E8a and the triumphant
final statement in five-part texture.
WTC II/18 in G minor Prelude

The predominant texture in this prelude is homophonic. Motifs in one
hand are supported by an accompaniment pattern in the other. Within this
setting, there is hardly any inversion of voices or contrapuntal interplay.
Apart from a few measures that display imitation, the laws of polyphony
do not apply here, and the term motif is understood as denoting a singlevoiced but non-meandering unit.
The prelude carries two verbal indications that draw attention. The use
of the terms piano and forte in mm. 3 and 5 is an unusual sight in Bachs
Well-Tempered Clavier and leaves many interpreters wondering, especially
since neither of them is taken up in the remaining 45 measures of the piece.
One way of reading these dynamic indications is as a hint that this prelude
was written for performance on a keyboard instrument with registers (such
as the harpsichord).1 Whether or not that was implied, it is beyond doubt
that Bach wished mm. 3-4 to sound as an echo of mm. 1-2. This second
interpretation, as the more basic one, should definitely be taken as a guide,
regardless of the particular instrument on which the piece is performed.
Each of the initial four measures is harmonically conceived as i-V-i.
These are not structurally relevant cadences. The same holds true for m. 5,
which features all steps necessary for a modulation but because of the rest
in m. 61 does not qualify as a section ending. That is reached soon afterwards, in the cadential close of mm. 7-81, where the key of the dominant is
successfully established.
The prelude comprises four sections, two of which are subdivided.
Ia mm. 1-81 i-v
G minor to D minor
Ib mm. 8-161 v
D minor to A major/D major
II mm. 16-24 v
dominant confirmed, Picardy-third close
IIIa mm. 25-361 V-VII D major to F major
IIIb mm. 36-411 VII-i F major to D major seventh/G major
IV mm. 41-50 i-i
tonic confirmed
There are three structural correspondences: mm. 3-4 . 41-42 (varied),
mm. 8-113 . 36-393 (transposed), and mm. 113-133 . 34-35 (transposed).
1
The first interpretation entails a register plan throughout the piece in order to convey what
might have been Bachs idea beyond the very scarce indication of mm. 3 and 5. This requires
specialized knowledge about early keyboard instruments.
507
508
WTC II/18
Since the rhythm is simple and the pitch pattern includes many large
intervals, the basic character of this prelude is lively. The articulation
requires 16th-notes in light legato touch and non-legato eighth-notes. The
fact that Bach slurs the eighth-note pairs in mm. 9 and 10 as well as the
similarly built pairs in mm. 37 and 38 indicates something beyond nonseparation. These are characteristic appoggiatura-resolution pairs, a quality
that, without the slurs, might have been easy to overlook in the complex
harmonic setting of these measures. Not only should such pairs under all
circumstances be rendered legato, the slurs add a sense of weightylight.
The similar slurs in mm. 44-45 serve to distinguish metrically analogous
but harmonically different groups: on beat 2, the appoggiatura is to be
linked to its resolution, while beat 4 features two sixths belonging to the
same harmony which must be detached.2
The tempo of the prelude is flowing: fast enough to allow listeners to
hear swiftly moving quarter-notes, but not so rushed as to convert the 16thnotes into virtuoso runs. The appoggiaturas that are distinguishing features
in the prelude must retain some of their sighing quality. Ornaments include the grace-notes and two trills. All grace-notes appear as eighth-notes
preceding quarter-notes. The value of the main note is thus split into equal
halves, with the appoggiatura on the beat (mm. 2, 4, 17, 31, and 42). The
trills embellish 16th-notes that form part of written-out ornamental patterns
(U: m. 18, M: m. 19). They are therefore nothing but suffixless mordents,
beginning on the upper neighbor note and comprising four 64th-notes.
Both the structure of the prelude and its dynamic layout are primarily
influenced by the succession of motifs. Meanwhile, orientation for a largescale design is provided by several indirect lines. Thus the first three of the
preludes four sections are strung together by consistent overall descents in
the outer voices. In section I, the bass moves gingerly downward over 1
octaves, slowly at first, then picking up speed in diatonic steps followed by
more closely spaced chromatic steps, only to rise again in consecutive
fourths. Similar though even more pronounced processes can be detected
in section III (l.h. from m. 25, r.h. from m. 36) where extended chromatic
descents even create rhythmic patterns. Section II is more static with merely
a few bass-note lines. In the fourth section are such lines of no consequence.
The example shows this with an excerpt from the first three sections:
2
Further pairs of appoggiatura-resolution do, however, existunmarked!in the right-hand

part of mm. 16 (B-A), 24 (G-F and C-D), 40 (E-D and G-F), 43 (E-D), and 50 (A-G
and C-B) and should be played accordingly. Appoggiaturas written in the form of gracenotes do not, of course, pose a problem of articulation since they are always legato.
G minor
II
509
mm. 1 2 3 4 5 6 7
mm. 16 17 18 19
III mm. 25 26
20
27
10
11
12
13 14 15
21 22 23 24
28
mm. 36
29 30 31 32 33 34 35
37
38
39
40
This brings us to the motifs characterizing this prelude. M1 is a twobar unit with three elements. In m. 1, a 16th-note line accompanied by two
metrically placed chords describes a twofold curve in the treble. The lines
long descent into the bass register splits into a hidden two-part structure
consisting of a typical bass pattern below a repeated pedal. In m. 2, new
treble voices present two sigh motifs in double thirds. The resulting
impression of a three-strand texture should be underlined: there are the
block chords, the 16th-notes, and the little melodic gestures. M1 is then
repeated (according to Bach: as an echo), with the sigh motifs in inverted
voices as parallel sixths. Variations of M1 can also be found in mm. 16-17,
41-42, and 43-44. In all three cases, the accompanying chords are replaced
by melodic figures based on segments of the other two M1 elements.
M2 is introduced in mm. 5-7. In its accompaniment, it is related to M1.
The melodic double notes are also taken up, together with their particular
metric organization. The double notes describe what appears like an incomplete one-bar curve. Beginning with a written-out inverted mordent
followed by a sudden leap, they embark on a descent that seems to prepare
a conclusion on the next downbeat. Yet the expectation is deceived twice
before the sequence finally reaches its goal (see m. 81). Dynamically the
three measures describe a gentle descent in which the escaping sixth leaps
should sound as charms, i.e., softer than their surroundings. M2 recurs
once, without its second sequence and, more importantly, without the final
resolution (see mm. 21-22).
M3 consists of two contrapuntally conceived voices. M3a, the lower
part, spans a single measure (m. 8). M3b, the right-hand part, begins after
beat 2 and ends in the following measure with the above-mentioned slurred
510
WTC II/18
appoggiatura-resolution pair. M3 is sequenced in falling direction in mm.

9 and 10 and recurs, in inverted voices, in mm. 36-38, where it is followed
by a longer extension leading right up to the end of this section in m. 40.
In addition, Bach invents a half-measure sequence model. It concludes
section I and recurs in section III. Labeled x here in contradistinction to
the motifs, it counters the preceding protracted descent in the bass with an
ascent (m. 11: D-E-F, m. 12: G-A-B, C-D-E, m. 13: F-G-A, B),
and the preceding gradual diminuendo with a forceful crescendo, before
leading to the cadential close in m. 15.
The following table shows the distribution of the material in the four
sections as well as the crucial dynamic developments.
Section I
mm. 1-2
M1
(f)
3-4
M1
p
5-7
8-113
M2
M3
f di mi nu en do
Section II
mm. 16-17
M1 var.
18-20
developmt.
21-22
M2
Section III
mm. 25-33
developmt.
dim.
34-35
x
cresc.
393-40
36-393
M3
expansion
di mi nu en do
Section IV
mm. 41-43
M1 varied + expanded
113-15
x + close
crescendo
23-24
expansion + close
46-49
development + close
WTC II/18 in G minor Fugue

Beginning on the downbeat of a 6/8 measure, the subject extends over
four measures, ending at m. 51 where B, the third of the G minor chord,
concludes the first complete cadence. The phrase structure is perhaps the
most striking feature in this subject, particularly since, as will be seen later,
this structure serves as a model for several other components of the
primary material. Remarkably, the subjects second half is an almost exact
sequence of the first halfan unusually simple pattern for the main
thematic idea of a polyphonic composition. (The final note, in this light,
seems almost like an aborted attempt to add yet more sequences.) This
G minor
511
structure, in conjunction with the extraordinary rhythmic simplicity, might

cause the lulling effect of an ongoing perpetuum mobile if one subject
statement were to follow another without interruption. Bach avoids this by
separating all statements after the initial entry pair of subject and answer
with contrasting secondary material.
Rhythmically, the subject itself consists exclusively of eighth-notes.
Outside the subject, quarter-notes and dotted quarter-notes, i.e., note values
that can be expected in any 6/8 time, predominate. Eventually, however,
syncopations also become a frequent feature. This begins with long tied
notes that can be identified as the beginning of large-scale closing formulas
(mm. 5-6, 7-8). Soon thereafter, metric accents shift from the middle beat
to the second or to the third eighth-notes and from the downbeat to the fifth
or to the sixth eighth-note (see, e.g., mm. 9 + 10, 15-16, 17 + 18, and 45 + 46,
respectively). Finally, Bach even adds some 16th-notes. These are found
only rarely on the first two pages of the score (mm. 39 and 62) but more
frequently from m. 71 onward.
The pitch pattern is equally ambiguous. The string of consecutive leaps
at the subjects beginning (see mm. 1-2: A-D-D-G) is contrasted with
a high percentage of chromaticism throughout the fugue. The harmonic
background implied in the unaccompanied subject is straightforward: a
simple progression is
interspersed with ornamenting steps at the
6
6
7
7
outset of each half.
i V i
iv5 i iv5 V i
Bachs actual harmo(i - - - - - - - - - - iv - - - - - - - V7 i)
nization of the subject
in the course of the fugue covers a wide range of alternatives. Apart from
the final V-I, every chord is at some instance substituted by another
cadential step. This gives the subject an uncanny versatility.
The dynamic shaping brings up the question of the relative melodic
value of the eighth-notes: are all eighth-notes equal, i.e., on the same level
of intensity, or can some of them be regarded as secondary? While this
question certainly enters way into the sphere of individual interpretation, it
is important that it be answered in detail as it influences many aspects,
including that of conveying the phrase structure. A possible concept is the
following: One can regard the sudden upward leaps in mm. 1 and 3 (the
fifth eighth-notes each) as
escape notes and the narrow wavy lines in mm. 2
and 4 as ornaments of a
512
WTC II/18
simpler background. This unveils a backbone that allows both a singing

style and subtle phrasing before the sequence. The corresponding dynamic
entails two soft curves, the second slightly more exuberant than the first.
The principal subjectbecause what has been discussed so far later
turns out to be one of two subjects in this fugueis heard in twelve statements. The subject receives a real answer and suffers no modifications.
1. mm. 1- 5
U
7. mm. 55- 59
L
2. mm. 5- 9
M
8. mm. 97-101
L
3. mm. 13-17
L
9. mm. 103-107
U
4. mm. 19-23
M
10. mm. 111-115
M
5. mm. 33-37
L
11. mm. 125-129
M
6. mm. 45-49
U
12. mm. 135-139
U
From m. 61 onward, Bach introduces a second subject. It begins on the

third eighth-note of the measure with a five-step chromatic descent,
followed by a slightly more straightforward ascent back to the initial pitch
before it concludes with a dosido formula in the traditional rhythmic
pattern. Subject 2 thus contrasts in several ways with subject 1: Its rhythm
is varied, the outstanding rhythmic features being the shortlong, shortlong
of the chromatic segment and the dotted notes with syncopation in the
closing formula. Furthermore, the ornament on the penultimate note adds
faster movement. The pitch pattern is restricted to very small intervals:
there is only one whole-tone step in an overwhelming surrounding of
semitones. And in further contrast to subject 1, the phrase is inseparable,
containing neither sequences nor other features that would invite or even
permit sub-phrasing. All that subjects 1 and 2 share is the four-measure
length and the fact that all entries are separated by episodes.
Despite its entry sixty measures into the fugue, subject 2 appears almost
as frequently as subject 1: in altogether nine entries.
1. mm. 61-65
U
6. mm. 103-107
M
2. mm. 66-70
M
7. mm. 111-115
U
3. mm. 71-75
L
8. mm. 125-128
L
4. mm. 79-83
U
9. mm. 135-139
M
5. mm. 97-101
M
Two of the statements of subject 2 are slightly changed in shape: in m.
83, the final resolution is reached only after a suspension and subsequent
ornamentation of the leading-note, whereas in m. 128, subject 2 already
breaks off on this leading-note.
G minor
513
Both subjects are accompanied by their own companions, but neither

of them proves to be very faithful. CS1 is introduced against the second
entry of S1 (see m. 5 B to m. 91 D). Its structure is strikingly similar to
that of the subject it accompanies: it consists of two subphrases, the second
of which sequences the first. Each subphrase features the dosido formula
Bach also uses in subject 2, together with a chromatic upbeat. The pitch
pattern encompasses only semitones. Dynamic shaping in this countersubject is unproblematic. In each subphrase, the eighth-note upbeat
prepares the climax on the syncopation, which is resolved throughout the
formula. CS1 recurs only once (see mm. 19-23). CS2 is presented together
with S2 (see L: mm. 61-65). Like its leader it is indivisible. Apart from a
syncopation with subsequent written-out inverted mordent, CS2 features
only eighth-notes and a not very distinct melodic pattern. This countersubject recurs twice: In mm. 66-70, where the beginning is completely
changed, it imitates the original only from the last eighth-note of m. 67
onward. In mm. 71-75, by contrast, only the first measure is modified in
such a way that the characteristic syncopation followed by an inverted
mordent now appears twice instead of just once. An intriguing fact about
these two main counter-subjects is that CS1 is distinctly related to subject 2
while, upon close inspection, CS2 is related
to subject 1. Besides the CS1
coincidences in pitch,
CS1 features some of
the shortlong patterns S2
from S2 while
CS2 like S1 is
characterized CS2
by eighth-note
motion. Hence
this is a fugue S1
with two main
thematic ideas.
At the outset of the composition, one dominates and is accompanied by the
other; later, the second idea dominates and is accompanied by a variant of
the first. From m. 97 onward they finally meet in their unique shapes.3
3
Two further S1 counter-subjects recur only once each in considerably altered shape; they
thus exert no real impact on the polyphonic design of the fugue. Both combine dosido
formulas with eighth-note runs, thus mixing elements of both primary components.
514
WTC II/18
The fugue encompasses fifteen subject-free passages.

E1 mm. 9-12
E9 mm. 75-791
E2 mm. 17-18
E10 mm. 83-96
E3 mm. 23-32
E11 mm. 101-102
E4 mm. 37-44
E12 mm. 107-110
E5 mm. 49-54
E13 mm. 115-124
E6 mm. 59-611
E14 mm. 129-134
E7 mm. 65-661
E15 mm. 139-143
E8 mm. 70-711
Among the material used in the episodes are components from the
subjects and counter-subjects,4 several independent motifs, and traditional
closing formulas. The line between motifs derived from primary material
and those invented exclusively for use in episodes is somewhat blurred,
particularly where later variations of originally independent motifs end up
resembling segments of a subject or counter-subject. Moreover, several of
the motifs are related among themselves, and the designation of one as a
modification or as new may seem arbitrary. The list given below,
while aiming at the highest degree of transparency possible in so complex
a piece, tries to mention further links wherever applicable.
Independent motifsmotifs apparently invented specifically for the
use in episodes and not derived from primary materialare introduced in
the first two subject-free passages and determine all others. Oddly, M1 and
M3 anticipate CS2 while M2 and M4 act as forerunners of S2 itself. M5
alone seems at first to be entirely independent of the fugues primary
thematic material, but this too is as if debunked in later variations.5
4
The final 8 eighth-notes of S1 are quoted in U: mm. 37-391 and imitated in M: mm. 39-411.
Both quotes are accompanied by the second subphrase of CS1 (see M: mm. 37-391; U: mm.
39-411). This material builds E4a, which ends with a cadential close in B major at m. 411. In
E14, only the final 6 eighth-notes of S1 appear, but all the more frequently (see L: mm. 128129, M: mm. 129-130, U: 130-131, M: mm. 132-133, and L: 134-135). The ending of CS1
is considerably shortened here and dropped after only one occurrence (see U: mm. 129-130).
S2 material is heard in two versions: as a chromatic descent (taken from the initial six notes
of S2, see M: mm. 115-116 and L: mm. 117-118) and as an ascent of one whole and two half
steps (derived from the five rising notes of S2, see U: mm. 83-84, M: mm. 84-85, and L: mm.
85-86, 86-87, 87-88, 88-89). CS2 appears twice as an extending sequence (see in E7: L mm.
65-66, and in E8 varied: U mm. 70-71). Later on, the entire second half of CS2 (i.e., its final
twelve eighth-notes) is quoted in E13 (see U: mm. 123-125). This recurrence is accompanied
by the shared ending of CS1 and S2 (see L: third eighth-note m. 123 to 1251).
5
For M1 see U: mm. 9-101, for M2 see M: mm. 9-101, for M3 see M: mm. 17-181, and for
M4 see L: mm. 17-181. M5 is introduced in U: mm. 17-191 and continued with a chromatic
variant from U: mm. 107-1091 onward.
G minor
515
Three of the episodes are distinguished from the remaining ones by the
fact that they feature motifs that appear nowhere else in the fugue. As these
three episodes are at the same time the longest subject-free passages, this
observation certainly deserves attention. The components are: in E3, a syncopation followed by five eighth-notes (L: mm. 26-27, M: mm. 27-28, and
L: mm. 28-29), in E10, two consecutive syncopations with upbeat and
subsequent mordent (U: mm. 85-871, M: mm. 86-881, U: mm. 87-891, and
M: mm. 88-90) and the figure characterized by a slide in 16th-notes
(U: mm. 90-91, M: mm. 91-92, and L: mm. 92-93), and in E13, a three-note
ascent with artificial leading-notes (U: mm. 119-120, M: mm. 119-120,
U: mm. 120-121, M: mm. 120-121, and U: mm. 121-122).
The role each of the fifteen episodes plays in the dynamic development
of the fugue can be described as follows: E6 and E15 are nothing but
elaborate cadential closes, extending the preceding subject statements
without any color contrast and bringing forth an immediate and complete
relaxation. E4a, E7, E8, and E14 evolve from the subject entries preceding
them by way of imitation or sequence. They, too, do not bring forth a color
contrast but just a gradual relaxation. The three long episodes (E3, E10,
and E13) describe self-contained curves. They all begin with an increase of
tension followed by an extended decrease and a final relaxation underpinned by a cadential close. The remaining episodes (E1, E2, E4b, E5, E9,
E11, and E12) are short and display independent material. E1, E2, E5, and
E9 are determined by descending sequences, while the others, particularly
E11 and E12, display rising lines. Besides their individual increases or
decreases, all these episodes establish a color contrast to the surrounding
subject statements.
In this fugue, the main subject with its simple rhythm and restricted
melodic expression can be dangerously deceiving with regard to the works
mood, misleading performers into too swift and light-paced a mood. The
complexity of the overall rhythmic pattern and the high content of
chromatic lines and altered notes do in fact indicate a rather calm basic
character. The tempo should express the compound meter in the manner of
a very moderate half-measure pulse.
The relative tempo of the prelude to the fugue sets the larger beats into
proportion, thus providing enough variation on the surface with 16th-notes
in one, triplet eighth-notes in the other piece: a quarter-note (a beat) in the
prelude corresponds with a dotted quarter-note (half a measure) in the fugue.
(Approximate metronome settings: 76 for the quarter-notes in the prelude
and the dotted quarter-notes in the fugue.)
516
WTC II/18
The articulation corresponding with this character demands legato for

all melodic notes. Exceptions occur in cadential-bass patterns and consecutive leaps (see, e.g., L: mm. 28, 40-41, 54-55, 112-113, 114, 136-137, and
142-143). Owing to the high density of relatively short motifs in this composition, the legato touch is frequently interrupted by phrasing, which in
almost all cases comes with both dynamic shaping and a gentle cut in the
sound flow.
The fugue features several ornaments requiring detailed planning.
Among the primary material, S2 is embellished with a cadential trill before
the final note. This trill is approached stepwise and thus begins on the main
note. For the speed of its shakes, 16th-notes seem most reasonable (a
choice which, one should be aware, gives all written-out 16th-note groups
in this piece the value of ornaments). The trill ends in a suffix that touches
the major sixth of the current scale. Bach indicates the trill in m. 64. In m.
69 it was apparently added later. For all we know, the same trill would
have been expected in mm. 74 (L), 100 (M), 106 (M), 114 (U), and 138 (M).
Only in mm. 128 where the subject breaks off early, and in m. 82 where
the resolution of the leading-note is delayed, is a trill redundant. (In m. 82,
a complementary ornamental figure actually appears in the middle voice.)
Two further trills are indicated for notes outside the thematic material. In
mm. 30 and 60, the lower voice carries whole-measure ornaments which
both begin on the main note and end with a suffix.
The structural layout of the G-minor fugue is evident from both the
entering order of the two subjects and the design and material of the episodes. Yet while the facts concerning the number and extension of sections
are quickly explained, it is intriguing to observe Bachs detailed and artful
use of secondary means for creating balance. The fugue encompasses five
sections. The first two are determined by subject 1. Both conclude with a
redundant entry (see section I: U M L M and section II: L U L) followed
by a cadential close (mm. 31-331 and mm. 59-611 respectively). In section
III, subject 2 repeats the entering order of the fugues exposition but ends
with the redundant statement in another voice (U M L U). This section, too,
closes with a perfect cadence (mm. 96-971). Sections IV and V are then
dedicated to the juxtaposition of the two subjects. The cadential close in
mm. 123-1251 separates these sections. Sections II, III, IV, and V all begin
with the three-part ensemble reduced to two parts by one resting voice.
Bachs mastery of balance is apparent particularly in the subtle play of
analogy and symmetry in the episodes. The following summary focuses
only on the structurally relevant details.
G minor
517
E1: M1 + M2
+ extension
E2: M3 + M4 + M5
E3: M3b + M4a + M5
+ extension with E3 motif
E7: M2a + CS2 sequence
E8: M2a + CS2 sequence
E9: M2a + M3a + M5
E10: Ms2
E5: M1 + M2
+ extension
E4b: M3 + M4b + M5
E4a: Ms1 + Mcs1
E6: cadence
E11: M3a
E12: M3a + M4/4a + M5a
E13: Ms2 + M3a + M4a
E14: Ms1, E15: extended cadence
518
WTC II/18
In the absence of any tension-enhancing features such as strettos or

parallel entries, minor/major contrasts, or varying contrapuntal density,
large-scale developments of tension seem less indicated than careful color
contrasts between subject-determined measures and episodes. Within each
section, the dynamic level of the consecutive subject entries increases
slightly. This is due to different means in each case:
In I the increase is engendered by the growing number of voices but
weakened by the receding tendency of the linking episodes E1 and
E2.
In II the less powerful increase of the ensemble is supported by the
use of primary material in E4a (interpreted by some analysts as a
substitute for a fourth entry), but is again counter-balanced by the
decreasing gesture of E5.
In III the entries of the second subjects exposition are strung
together by the two short episodes E7 and E8 that evolve from the
subject statements by way of sequence and thus enable a longer
stretch of a single color, uninterrupted by secondary material. The
increase is, however, discontinued in E9 before the redundant entry.
In IV the exposition of the subject juxtaposition is made similarly
stringent by the two linking episodes with increasing tendency (E11
and E12). The only major-mode statement of subject 1 in the entire
fugue adds further luster here and converts this statement pair into
the climax of the composition.
In V the use of S1-material in the linking episode avoids a color
contrast between the two statement pairs, but the descending
motion of E14 counter-acts any dynamic increase.
WTC II/19 in A major Prelude

This is a composition in three-part texture. All three voices take an
active part in the melodic flow. Although rudimentary imitation is used,
parallel motion and complementary interplay is so frequent that one would
not speak of a strictly polyphonic piece. The preludes predominant feature
is its very even rhythmic organization. The 12/8 time manifests itself in a
constant flow of eighth-notes. All syncopations are complemented by
eighth-notes in another voice, thus losing much of their metric power.
Together with the fact that the flow of eighth-notes is not once interrupted
during the entire composition, this gives the prelude a calm and soothing
atmosphere, classifying it as a metrically determined piece of meditative
intent.
Explicit closing formulas do not stand out. Instead, all cadential closes
melt into the flow of the piece. The home key of A major is first confirmed
in m. 3. Both the suspension in the upper voice and the fact that no subdominant has yet appeared make this cadence not eligible as a structural
caesura. E major is reached in m. 6 with an imperfect cadence, thus still
referring to A major. The secondary key is not truly established until m. 97.
But even here, the simultaneous beginning of a new line in a higher
register does not allow any feeling of decisive melodic closure to arise.
There is one significant structural analogy in this prelude: the opening
development in mm. 1-97 recurs, transposed and varied, in the two upper
voices of mm. 22-307, which are inverted. The remaining harmonic
sections are grouped around these analogous passages as follows:
I
mm. 1-97 tonic to dominant (E major)
II mm. 97-161 modulation to the tonic relative (F minor)
III mm. 16-221 modulation to the subdominant (D major)
IV mm. 22-311 return to the home key (A major)
V mm. 31-33 tonic confirmed (A major)
This metrically determined prelude is of a rather calm basic character.
The effect of tranquility is enhanced by a detail in the pitch pattern that is
important for an adequate understanding: almost all flowing eighth-note
lines consist of a succession of written-out inverted mordents (or, later,
mordents). The A-G-A and C-B-C in the upper voice of m. 1 are prebeat ornaments embellishing the notes of a broken A-major chord. The
initial measures could thus be imagined as in this example:
519
520
WTC II/19
These pre-beat ornaments are eventually complemented by a few onbeat ornaments. These are also spelled out and involve, in contrast to the
ones mentioned above, the only 16th-notes in the prelude. In m. 92, the
upper voice can be read as a dotted quarter-note F precede by a grace-note
appoggiatura and ornamented with an inverted mordent. In m. 21, the
16th-notes hide a downbeat G adorned with a slide. In view of the
ornamental structure of the lines it would be misleading to describe the
pitch pattern as consisting of alternating steps and skips, since these unfold
on different levels of the melodic process: the skips in the background
pattern, the steps on the ornamental surface.
The tempo of this prelude is calm but flowing: calm enough to avoid
any hurried impression (which might be caused particularly owing to the
silent finger shifting necessary for perfect legato) and flowing enough to
convey the feeling of gently swinging compound 4/4 time rather than
creeping eighth-notes. The appropriate articulation is legato in all notes
appearing in the upper and middle voices as well as most notes in the
lower voice, with the exception of octave leaps (see mm. 2, 5, 15, 23, 26,
and 32) and a cadential-bass pattern (m. 30: A-F-D-E-A) which should be
gently detached.
Although it is certainly possible to describe the initial melodic figure
of the prelude (explained above as an ornamented broken chord) and to
state its recurrences throughout the prelude, such a listing does not truly
facilitate an approach to the piece.1 What Bach presents here are general
pitch lines, elaborately decorated but basically simple. They can be
described as follows:
1
The figure recurs in its original version in

U: mm. 1 2 3
20
22 23
M: mm. 1 2
67
11 12
20
22 23 24
L: mm.
3
67
11 12/13 16
and additionally in inversion in
U: mm.
9 10
19
M: mm.
10
L: mm.
9 10
17 19
21
27 28
31 32/33
32
32
27 28
31
30
A major
521
Section I
mm. 1-41 rising
4-61 falling
U: C-D-E-F U: F-E-D-C-B
6-81 rising
M/L
Section II
mm. 97-131
falling
13-151
rising
15-161
falling
Section III
mm. 16-201
falling
20-211
rising
21-221
falling
Section IV
mm. 22-251
rising
25-271
falling
27-291
rising
8-97 falling
U/M/L
29-311
falling
Section V
mm. 31-33
rising
Material and layout make it easy to identify section IV as a recapitulation of section I and section V as the coda. Yet whether it is meaningful,
or even necessary, to declare sections II and III as development, remains
arguable. Structure, just like thematic material, is but a secondary issue in
this piece. The dynamic realization can follow the essential pitch lines
traced above. Thus rising lines would be expressed in gentle crescendo,
falling lines in diminuendo. The basic tone color, however, remains the
same throughout the prelude. There are no contrasts and no moments of
dramatic tension, as implied by the calm and soothing impression this
piece conveys.
WTC II/19 in A major Fugue

The subject spans 1 measures in 4/4 time. Beginning, after an eighthnote rest, on the keynote A, it concludes at m. 23 where C represents the
resolution of the preceding dominant into the tonic. The contour features
several small curves winding around a larger line. Understanding this is
vital for perceiving the phrase as an indivisible unit. Comprehending the
ornamental character of the notes embellishing this curve is also essential
for a correct evaluation of the subjects many chromatic alterations in the
course of the fugue.
522
WTC II/19
The rhythmic pattern, both in the subject itself and in the fugue as a
whole, is simple. The basic features are chains of 16th-notes and dotted-note
groups. These two frequently combine in such a way that the 16th-note
pulse appears as the result two parts in complementary dotted-note groups
(see mm. 5, 7-8, 10, 12, 20, 24, and 28 where the effect of uninterrupted
16th-notes is achieved by the interplay of the accompanying voices with
the subject).
The subjects harmonic background is basically simple, although the
ornamental curves and their syncopations add spots of flavor and tension.
The music example shows the
subject as Bach presents it at the
opening of the fugue and below it
the underlying large-scale curve;
both marked with an interpretation of the main harmonic steps.
The dynamic design follows
the simple phrase structure with
an increase throughout the first measure, a climax on the second syncopation, and a subsequent relaxation. The syncopation on the last eighthnote of m. 1 captures the highest amount of tension: it stands harmonically
for the subdominant, i.e., the most active step in the simple cadence, and
metrically for the only downbeat in this subject, an important fact in a
composition where meter, owing to the characteristics of the rhythmic
pattern, plays a decisive role.
There are ten subject statements in this fugue.
1. mm. 1-23
L
6. mm. 12-133
M
2. mm. 23-41
M
7. mm. 16-173
L
3. mm. 5-63
U
8. mm. 20-213
U
4. mm. 7 -83
L
9. mm. 233-251
M
5. mm. 93-111
U
10. mm. 273-291
U
The subject undergoes only two kinds of modifications in the course of

the fugue: extensions at the phrase beginning and chromatic alterations in
the ornamental 16th-notes. The pair of 16th-notes initially launching the
subject is extended from m. 9 onward to three notes. The additional 16thnote precedes the original beginning either by a leading-note (mm. 9, 20,
and 27) or by a rising fourth interval (mm. 12, 16, and 23). Alterations
within the 16th-notes that do not arise from transposition (as is the case in
A major
523
mm. 12-13) involve the third and the fourth scale degrees (mm. 20-21:
G/G/G/G, mm. 23-24: G/G/G, and m. 28: D/D/D/D). There are no
inversions, strettos, or parallel statements.
Neither does Bach invent a regular companion to the subject that
would display a minimal degree of independence. Instead, the subject
comes accompanied by several kinds of rhythmically varied parallels. The
most frequent pattern moves in thirds below the simplified curve (M: mm.
5-6, 9-10, 20-21, and 27-28) or in sixths above it (U: mm. 23-24). The
quarter-note values are broken into dotted-note groups that split the pitch
into note repetitions (mm. 5-6, 9-10, and 27-28) or add chromatic
semitones (mm. 20-21 and 23-24). Interestingly, such parallels only ever
occur between the upper and middle voices. There is only one lower-voice
entry that incites a parallel. Yet this is even more explicit than the
previously mentioned ones as it involves more fractions of the ornamental
pattern in the subject (see m. 7: L + M).
In four of the previously identified cases (mm. 5-6, 9-10, 20-21, and
27-28), the lower voice adds a further double-third parallel to the subjects
quarter-notes on beats 3, 4, and 1. As a result, this stretch sounds in parallel
triadsan extremely rare phenomenon in a genre renowned for complex
polyphony. This further parallel also comes in a dotted-note rhythm, with
the splitting appearing as a note repetition, an octave displacement, or a
chromatic semitone. The following comparison of one of the excerpts with
its simplified version shows these parallels:
mm. 5-6
The fugue encompasses nine subject-free passages.

E1 mm. 4-51
E6 mm. 173-201
E2 mm. 63-71
E7 mm. 213-233
E3 mm. 83-93
E8 mm. 253-273
E4 mm. 11-121
E9 m. 29
E5 mm. 133-161
The most prominent feature in these passages is a motif taken from the
end of the subject. It will be referred to as Ms. The layout of the episodes
results from the developments of this motif. E1 features descending Ms
524
WTC II/19
sequences above an equally sequenced bass motif. The same combination

recurs, with small modifications, in E4 where the Ms sequences appear as
an imitative play between the upper and middle voices. Owing to their
descending motions, both episodes express decreasing tension. E2 introduces the inversion of Ms. It is accompanied by a three-eighth-note ascent
in the lower voice, which in turn is partially paralleled in the middle voice.
This combination recurs in E5a where the three-eighth-note ascent appears
in the middle voice while the lower voice sustains a short pedal. As both
the Ms inversion and the eighth-note-figure express an active gesture, the
tension in these two episodes rises. E3 combines descending Ms sequences
with a pattern of rhythmically interlocking, descending dotted-note groups.
This pattern begins early, anticipating the episode as an accompaniment to
the ending of the subject. The same combination recurs in E6a. The threepart descent in these episodes creates a considerable dynamic decline. A
complete relaxation is thwarted primarily by the harmonic ending in a
dominant-seventh chord.
E7 introduces the parallel motion of Ms. The pattern merges descending sequences in one voice with an imitative pattern in the other two parts.
A modification of this combination is found in E8 where the descending
sequences intersperse Ms with its inversion, an alternation taken up in the
two imitating voices. The overall descent in these two episodes creates another dynamic decline. Of the three remaining episodes, E5b (mm. 14-161)
displays a certain relationship to E1 or its variant E4. The cadential close
it brings forward is ingeniously wrapped in the previously established
material. Similarly, E6b (mm. 183-201) is remotely related to E7 and E8 as
it displays varied parallel and contrary-motion statements of Ms. Finally,
E9 consists of nothing but a concise cadential formula. Dynamically, E5b
extends the progressive tendency of E5a before it approaches the cadential
relaxation. Conversely, E6b prepares the subsequent subject statement in
an unimpeded upward thrust.
The simplicity of the rhythmic pattern combined with the ornamental
impression of the 16th-notes indicates a rather lively basic character. The
tempo may be fairly swift, and should by all means be fast enough not to
invite listeners to hum along with the 16th-notes.
There are two rather different solutions for the tempo proportion between this fugue and its prelude. Performers who emphasize the flowing
quality of the prelude and thus render the compound beats in a swinging
mood equate the larger metric values. This interpretation emphasizes the
ornamental character of both pieces and stresses the similarities between
them: a dotted quarter-note in the prelude corresponds to a quarter-note in
A major
525
the fugue. By contrast, performers preferring to focus on the calm in the

prelude had better establish a metric rapport between the smaller values of
both compositions. This interpretation enhances the contrast in prelude and
fugue: an eighth-note in the prelude corresponds with an eighth-note in the
fugue. (Approximate metronome setting: fugue beats = 90, dotted quarternotes in the prelude (a) = 90, (b) = 60.)
The appropriate articulation encompasses legato in the 16th-notes and
a very merrily bouncing non legato in the simple and dotted eighth-notes.
The only exception from the non legato occurs in the closing-formula that
accompanies the end of the answer statement (L: mm. 3-4), where E-D-E
must be played legato. The fugue features only one ornament, appearing in
brackets in the final measure. It is a point darrt-trill beginning on the
upper note, shaking in four 32nd-notes the last of which is tied over (or a
32nd-note triplet followed by an eighth-note syncopation) and resolving
(legato) onto the anticipated key note.2
The entering order of the subject statements, together with the fugues
harmonic argument, provides a clear picture of the structural layout.
Section I comprises four entries in the home key, the basic round L M U
and a redundant statement in L, before closing at m. 93 with an imperfect
cadence, a V7 of F minor. The two statements of section II sound in the
minor mode, in the tonic relative and dominant relative respectively. The
analogy of the two episodes E4 and E5a with the first two episodes of
section I creates the impression of a structural correspondence between the
two sections. This section concludes at m. 161 with a perfect cadence in the
home key of A major. Section III comprises again four statements, which in
this case represent the steps of the simple cadence (L U M U = tonic,
subdominant, dominant, tonic). The episodes in this section support the
structure in two ways. The initial E6a relates to E3, the final episode of
section I. The two corresponding episodes thus create a bracket around the
minor-mode middle section. The three longer episodes display a gradual
development of the Ms parallel: from the somewhat unruly patterns in E6b
via the orderly, only slightly varied parallels in E7 to the complex pattern
in E8 with parallels in contrary motion.
2
The fact that this ornament appears in brackets should not induce us to think it is optional.
This kind of cadential trill was so commonplace in Bachs time that performers would
automatically play it even where it was not explicitly stated. Its appearance against running
16th-notes, and the alleged technical problem arising hereof, is no reason for omitting the
ornament either. Anyone who can play this fugue in all its required rhythmical accuracy will
certainly be able to tackle this trill, particularly if two of the final 32nd-notes in the middle
voice (D-C) are taken over by the left hand.
526
WTC II/19
In the absence of tension-enhancing developments in the subject statements, no dramatic climaxes are built up. By the same token, the close
thematic relationship of the episodes with the subject discourages explicit
color contrasts between the primary and the secondary material. Instead,
the only large-scale shading occurs in the minor-mode middle section,
which should sound less brilliant than the surrounding major-mode sections.
This is a playful, virtuoso fugue in which joyfulness rather than dramatic
developments are the goal.
WTC II/20 in A minor Prelude

The prelude in A minor features two halves of equal length and, at first
glance, similar layout. This may remind one of the dance movements in
Baroque suites. The texture is deceptive: while there is consistent two-part
writing with frequent imitation, the voices are only rarely independent of
one another. Two significant motifs dominate almost the entire piece. Each
comes with a steady accompaniment that swaps voices as the respective
motif does, thus impeding any contrapuntal juxtaposition of the two motifs
themselves. The prelude can thus be described as motivically determined,
or more exactly, determined by two-part motifs in the context of a
quasi-independent texture.
Apart from the closure of both halves, all measures are taken up by
motivic material, thus not leaving any room for cadential formulas. In
addition, an extraordinarily high degree of chromaticism in the leading part
of each motif weakens the sense of harmonic progression. Thus the layout
is conveyed only by means of structural analogies created by a similar
order in the presentation of the motifs.
mm. 1
. mm. 8 and
both transposed, voices inverted,
with different bar-to-bar transition
mm. 2-51 . mm. 9-121
mm. 5-81 . mm. 13-161 transposed in inverted voices
mm. 17-19 . mm. 21-23
mm. 1-33 . mm. 25-273 transposed and varied
mm. 4 + 5 . mm. 30 + 31 both transp. / var., voices inverted
. mm. 32
mm. 16
The conclusions to be drawn for a first estimate of the layout are
twofold: The first half of the prelude consists of two sections: mm. 1-8 and
mm. 9-16. The second half is equally divided, with section III subdivided
in itself, while section IV shows traces of a recapitulation. This results in
the picture of a ternary form structured as ||: A A':||: B A":||. At the same
time, the correspondence of m. 32 with m. 16 supports the impression of
binary form, given already by the equal length of the two halves and the
repeat signs.
The prelude displays a large variety of note values (eighth-notes,
16th-notes, 32nd-notes, and frequent syncopated eighth-notes) in combination with a high degree of chromaticism. Almost all of the apparent leaps
in the thematic material represent either different melodic levels or
527
528
WTC II/20
ornamentation (for details see below). The basic character of the piece can
therefore be interpreted as rather calm. The tempo may be fairly slow; the
chromatic 16th-notes should be given time to unfold their full emotional
content, and the 32nd-notes ought not to appear virtuoso.
The corresponding articulation is a continuous legato. This is only
interrupted by phrasing and in the few cases of cadential octave leaps (L:
mm. 16, 28, and 32). Phrasing is a matter that needs to be pondered with
great care. At first glance one would think this to be straightforward, as
one-measure motifs and the imitative texture are obvious. Closer analysis
reveals that there is frequent overlap of a tied notes harmonically required
resolution at the end of one motif with the initial note of the subsequent
motif. Such overlap, in which an interruption of the sound flow would
therefore be wrong, occurs in almost every measure. In all cases, the
progression from the end of one phrase to the beginning of a new one can
only be expressed by exceedingly eloquent dynamics and occasionally by
a change of tone color. The following table gives an overview of the
phrasing at the point after the downbeats in each measure. The caesura ()
suggests a slight interruption of sound flow, the slur ({) indicates that an
interruption is unwarranted.
mm. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
U
{ { { { { { { {
L { { { { { { { {
mm. 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
U { { { { { { { { {
L
{ {
In addition, there is a breath after the middle beats in U: mm. 10, 14, 15,
20, 23, and 24 as well as in L: mm. 3, 6, 7, 12, 19, and 27.
The prelude feature three ornaments, occurring in the final measures
of the two halves respectivelythe two measures that, through the use of
scalar passages and leaping octaves, appear as virtuoso and thus distinctly
different from the measures preceding them. In m. 16, the lower voice is
embellished with what appears in the score as a mordent and a trill. The
mordent, however, is followed by a suffix-like group and thus given away
as another note-filling ornament. Both trills shake in 64th-notes, i.e., twice
as fast as the fastest note values in the piece, the regularly occurring 32ndnotes. The one at m. 161 is approached stepwise and therefore begins on a
slower main note and also ends in slower values, as indicated by Bach. The
other trill begins and ends regularly, launched from the upper note and
concluding with a suffix in the speed of the shakes. As both ornaments
A minor
529
occur in the left hand and against fast notes in the right hand, they intensify
the virtuoso impression that is created in the written-out notes and further
enhanced by the sudden appearance of the additional middle voice. The
inverted mordent in m. 32, by contrast, is simple: it consists of the usual
three notes (A-G-A), which should be played fast enough for the trill to
end before the B that initiates the run in the lower voice.
The preludes main motif (M1) is introduced in mm. 1-21. More
precisely, the lower-voice part of the motif begins on the second eighthnote, after a downbeat note that (like the one in the upper voice of m. 17)
serves as harmonic and metric support but does not belong to the thematic
material. These downbeats should therefore be played in a neutral tone
color and not at all assertively. The three lines of M1 are conceived in
parallel motion, presenting a single gesture in what may be described as
ornamented homophony:
The corresponding dynamic gesture is an overall diminuendo, triggered by a short but fairly strong crescendo in the initial four-note ascent.
While the decrease in tension is gradual and straight in M1b, the leading
M1a should be subtly shaded in such a way that the predominant chromatic
descent stands out against the backdrop of a softer secondary line in the
higher register.
M1 recurs without modifications in mm. 2, 4, 5, 11, 13, 25, andwith
small variationsin mm. 8, 9, 26, 30, and 31. Inversions occur in mm. 17,
18, and 22, modified also in m. 21 and as a partial sequence in m. 20. Very
free adaptations can even be detected in mm. 19 and 23-24. In all this, the
dynamic shaping of the inversion raises a question. Should one retain the
original design with its diminuendo or follow the natural temptation and
ascend in crescendo? This determines whether the B section, which is
where most of the inverted statements occur, emerges as the dramatic
climax of a prelude in ternary form (which results if the rising lines are
rendered in rising tension) or rather as an integrated section in a prelude in
binary form (as happens if the original dynamic shaping is retained also for
the motifs inversion).
530
WTC II/20
M2 is first presented in m. 3. It spans half a measure, beginning after

the downbeat and ending on the middle beat or, in case of a tied note, on
the note immediately after it (see m. 3, L: on D, U: on F). The motif bears
a certain relationship to M1 owing to the combination of 16th-notes +
32nd-notes in the leading voice against eighth-notes in the accompanying
voice and because of the hidden chromatic line with diatonic ending inside
the leading voice. It differs from M1 in that the accompanying voice is
diatonic and moves in contrary motion to the leading voice. The dynamic
profile of M2 is therefore much more complex than that of M1. It begins in
the leading voice with a written-out inverted mordent propelling a sudden
major-sixth leap; this gesture is best expressed in an emphatic crescendo.
The subsequent chromatic descent provides the corresponding relaxation,
interrupted on an even softer level (no accent!) by the weak-beat escape note
in the higher register. Against this highly emotional dynamic curve sounds
the straight yet gentle crescendo of the eighth-note ascent in the accompanying voice. M2 is sequenced in the second half of the measure, with an
adjustment at the beginning that is due to the resolution of the tied note.
The motif further recurs unvaried in mm. 10 and 12 as well as varied and
developed in mm. 27-30. The most interesting modification occurs in mm.
6-7 and 14-15, where the half-measure sequence gives way to half-measure
imitation and the two voices swap some features.
With regard to an overall dynamic shaping, the large-scale developments are of foremost importance: see, e.g., the concave curve of the first
three measures, built from one measure in predominant diminuendo, its
continuation in imitation, and a resurfacing of tension in the rising
sequence of the crescendoing M2 in m. 3, etc.
WTC II/20 in A minor Fugue

This subject spans two measures, with a substantial rest in its middle.
Beginning on the second beat of a measure in 4/4 time with the fifth degree
of the A-minor scale, it concludes at m. 31 with the third degree. Rhythm
and pitch suggest two palindromically related subphrases. With regard to
rhythm, the subjects first half contains four quarter-notes followed by a
quarter-note rest, while the second half viewed in reverse features four
eighth-notes followed by an eighth-note rest. The pitch pattern shows an
exclusive use of interval leaps (no single step within the subject!) in almost
perfect mirror symmetry: the first and last leaps are major thirds, the second
A minor
531
and second-to-last intervals are perfect fourths, and while the third leap
represents a diminished seventh, the third-to-last uses the complementing
minor third. The bridged connection between the two halves of the phrase
is marked by a gap both rhythmicallythe duration of the rest is
longer than any of the note values
major fourth dimin.
min. fourth major
surrounding itand in terms of
tritone
third
seventh
third
third
pitch, with the interval that expresses the highest degree of tension:
the tritone. Owing to this symmetrical correspondence of the two
halves, the phrase must be regarded as an indivisible unit in which the rest,
far from indicating phrasing, represents the moment of highest tension.
This view is supported by the harmonic progression, which displays a
simple cadence. The dominant (in Bachs harmonization often appearing in
the form of a ninth chord) falls on m. 21
and resolves, through the rest and the
leaping eighth-notes, only on the final
note at m. 31. The dynamic design correi iv V9 - - - - - i
sponding with these features consists of
a powerful increase of tension toward the G, followed by a tensionsustaining rest and a gradual release through the eighth-notes.
The subject appears eight times.
1 mm. 1-3
L
5 mm. 13-15
U
2 mm. 3-5
M
6 mm. 17-19
M
3 mm. 6-8
U
7 mm. 21-23
U
4 mm. 94-113
L
8 mm. 254-273
L
Apart from the adjustment of the initial interval in the tonal answer
and the enlarged leap in the major-mode statement (see m. 10: A-B = a
minor seventh instead of a diminished seventh), only small rhythmic modifications occur. In mm. 9, 17, and 25, the subjects initial note is shortened
to an eighth-note.1 Inversions, strettos, or parallels are not used.
1
At first glance, the fourth and the final entries seem to leave it open whether we are dealing
with the original of the subject beginning, with a third in which the gap is filled by a passing
note, or with the answer entering an eighth-note late. Upon closer inspection one discovers
that the metric organization of the melodic lines preceding the entries, together with the
harmonic logic, determine the eighth-note on beat 4 as the end of the preceding development.
532
WTC II/20
Bach invents two companions to the subject. These appear with great
regularity, accompanying all but the initial and final entries. CS1 is
launched after an intermittent partial sequence of the subjects tail from the
sixth eighth-note of m. 3 onward. Its conspicuous feature is a five-note
plunge occurring four times in descending sequences. The first two of
these scale segments are separated by an interval leap and a rhythmic gap.
This interruption is subsequently softened when the gaps in both pitch and
rhythm are fleshed out. After a fourth plunge, an ornamented ascent
concludes CS1 on a relaxed note. This counter-subject is thus independent
from the subject in every possible respect. The dynamic equivalent to the
compositional details consists in a four-fold increase of tension, with the
climax falling each time on the lowest note, followed by a decrease in the
gentle upward motion. CS2 enters almost a whole measure after the subject
and ends undetermined (compare L: mm. 7-8 with M: m. 11, L: mm. 14-15,
U: mm. 18-19, and M: mm. 22-23). Its most characteristic segment contains the first four eighth-notes. These complement the subjects rhythm by
filling the mid-phrase rest in a way that ingeniously combines the three
outstanding features from the subject: The interval pattern in the CS2
segment sounds like a free imitation of the subjects first half, the rhythm
is an anticipation of the subjects second half, and the distinctive interval,
used twice in the CS2 segment, is the tritonethe interval that connects
the two palindromic halves of the subject. At the same time, the CS2
segment forms a parallel in sixths with the central portion of CS1. Rather
than completely independent, CS2 is thus a supporting and enhancing
complement to both the subject and its primary companion. Dynamically,
a crescendo
throughout
the opening
four eighthnotes is answered by a
diminuendo
that parallels
the one in
subject and
CS1.
The fugue features eight subject-free passages. As the two countersubjects enter later than the subject, overlaps of episode material with
subject statements occur regularly. The table below lists only the episode
segments that play a role in the interpretation of the fugue.
A minor
533
E1 m. 3
M: second to fifth eighth-notes
E2 mm. 5-6
M + L: up to the fifth eighth-note
E3 mm. 8-9
M: up to the first or fourth eighth-note in m. 10
E4 mm. 113-13 U: up to m. 131, L: up to the fifth eighth-note
E5 mm. 15-173 L: up to the fifth eighth-note
E6 mm. 19-21 M + L: up to the fifth eighth-note
E7 mm. 23-253 M: up to m. 261
E8 mm. 273-28
The episodes combine motifs derived from subject and counter-subject
with independent motifs. Ms is a free sequence of the subjects second
half. It is first introduced in m. 3, the minute link between the initial
subject entry and the answer. It is later accompanied by a motif derived
from the end of the counter-subject (for Mcs see L: mm. 11-12, M: m. 12,
U: mm. 12-13), which soon spawns developments of its own (see mm.
11-13, all three voices) and ultimately even acquires an active attitude (see
L: mm. 19-213).
M1 is the first independent motif. Neither its pitch nor its rhythm
derives from the primary material. This motif is introduced in E2 (see L:
m. 51-53). In its symmetrical design of ascent and descent, the number of
notes (4 + 4), and the tension-sustaining rest, it shows a certain structural
relationship to the subject.2 M2 appears exclusively in E6 where it is heard
in a pattern with imitation and varied sequence (see U/M: mm. 19-20).
The use of episode material establishes a number of relationships
among the subject-free passages. (Subdivisions in some episodes appear on
the basis of material, not as a result of harmonic closes.)
E1 . E4b
E2 . E3a, E5a, E7b
E3b . E5b
E4a . E8
material
Ms only
Ms + M1
M1a imit.
Ms + Mcs
compare mm.
13 (L)
8-91 ,15-161,24-251
16-17
27-28
with mm.
2 (L)
5-61
9-10
11-131
As was already mentioned, E6 is in a category apart owing to its combination of M2 and the ascending sequences of Mcs. Outstanding for other
reasons is E7, which consists of three segments. E7a is an extension of the
preceding subject entry, E7b follows with a transposition of E2, and E7c
rounds this episode off with scales that are not even abandoned at the
advent of the subsequent subject entry, but replace the counter-subjects.
2
Variants of the motif are the compact version without rest in E3 (see. M, O, M: mm. 9-10)
as well as different extensions that increasingly approach the shape of simple runs.
534
WTC II/20
Both the powerful tension created in subject and counter-subject and

the large variety of rhythmic values used in the thematic material indicate
a rather calm basic character. The tempo should be statelyslow enough
to allow listeners to perceive the syncopations in CS1but not dragging.
The movement from one of the subjects quarter-note to the next must
express dramatic tension of the highest order. Counting eighth-note beats,
desirable as it may appear in light of the rhythmic complexity in the piece,
is detrimental to the subjects character. The tempo proportion between the
prelude and the fugue may be simple and direct, equaling a quarter-note in
the prelude with a quarter-note in the fugue. (Approximate metronome
settings: 60 for all beats in prelude and fugue.)
The articulation of this fugue constitutes an interesting case. While
one generally expects all melodic notes in calm character to be played
legato, the leaping pattern of the subject might cause some headache had
Bach not given a very obvious hint. The hint was obvious at least to the
performers of his time but may require some seemingly crooked thinking
from us: If Bach assumed that performers of his time would interpret the
subject as a pattern of consecutive leaps that must be detached, there would
have been no need to write the wedges on the eighth-notes in mm. 23-32.
He obviously trusted that musicians would be guided, by the high degree
of tension expressed in this particular arrangement of intervals, to choose
legato or at least hardly detached non legato. He marked the eighth-notes
because he wished them to be energetically separated.
As a result of this train of thought, the subjects four quarter-notes are
very long, just slightly detached, while all unmarked shorter notes values
are legato. The eighth-note pattern in the subjects second half and, by
extension, in Ms should be played non legato in a dramatic, not softly
detached way. In CS2, the articulation of the eighth-notes is at the
discretion of each interpreter, although it should be mentioned that playing
them long not only emphasizes their relationship to the subjects beginning
but also makes it easier for the listener to distinguish this second countersubject among the numerous patterns of detached eighth-notes. (Performers who follow this suggestion would thus render the initial four eighthnotes of CS2 in an almost connected, hardly detached style while distinctly
separating the remaining ones where the leaps no longer carry melodic
importance.) Finally, where only one high-tension interval appears (as in
U: m. 20) or where stepwise motion prevails in eighth-notes (as in M: m.
21), unbroken legato is recommended.
The fugue contains one kind of ornament only: the trill at the end of
CS1 and, consequently, in Mcs. This trill begins on the upper neighboring
A minor
535
note. Its shakes are usually taken in 32nd-notes. This is acceptable, though
choosing twice the speed might be preferable for anybody who can manage.
The faster trills not only add drama and brilliance to the work, but also
avoid the problem of fusing ornamental notes with melodically essential
ones (e.g., the 32nd-notes in CS1, M1, and M2). The trill ends with the
suffix in the rhythm notated by Bach.
There are two exceptions in the execution of the trill. In M: m. 27, the
ornament begins on the main note and thus with a note of twofold duration
before launching its shakes.3 In m. 28, a compound ornament beginning
with an inverted mordent is indicated. The trill thus begins, in full speed,
with the lower neighbor note.
As has been shown above, Bachs design of the episodes divides the
fugue into two halves. The first half ends with a cadence in the home key
at m. 131. The second of the slightly uneven halves literally consists of two
sections: section II (mm. 13-211) features two subject statements in the
upper and middle voices, followed by section III (mm. 212-28) with two
statements in the upper and lower voices. The relationship among the
subject entries in this fugue, particularly between U: mm. 13-15 and M:
mm. 3-5 as well as L: mm. 25-27 and L: mm. 9-11, stresses the design in
two halves over that in three sections. The second half is only slightly
more extended, as a table of the corresponding episodes displays:
E1
E4b
E2
E3
E5 (E6, E7a) E7b (E7c)
E4a
E8
The first section begins with a majestic increase from the first to the
second subject statement. The episode that follows, conceived in descending
sequences, reduces the level of tension, which is then picked up all the
more forcefully by the third subject entry in the upper voice. E3 brings
another gradual decrease. Interpreters who regard the redundant lowervoice statement as a false fourth-voice entrya possible viewshould
aim at slightly surpassing the dynamic level of the previous entry. Performers who do not underwrite this concept should render the entry in slightly
lesser intensity, as is fitting for a redundant statement. The final episode of
section I, E4, provides the dynamic closure.
3
Note that in Henles Urtext edition, the counting of the measures is wrong at this spot since
the half measure at the end of the second-last line and its complement at the beg inning of the
last line have accidentally been separated by a bar line and are consequently counted as two
measures. The fugue totals 28 measures.
536
WTC II/20
Section II begins with an upper-voice entry in the highest possible

register. The relaxing episode that follows and the much less exposed
middle-voice statement suggest a gradual decrease through this portion
(mm. 13-191). E6, the final episode of this section, just like that of the
previous one, propels the tension upward rather than preparing a conclusion.
Section III, thus closely linked to the preceding section, begins with a
reiteration of the upper-voice statement in the highest octave available on
eighteenth-century keyboards, thereby repeating if not surpassing the
climax in m. 13. Once again, the subsequent episode together with the
lower-voice entry, not least because of its reminiscence of features from
the redundant entry from section I, present a decrease in tension that is,
this time, completed with a final perfect cadence.
WTC II/21 in B major Prelude

This prelude is composed in polyphonic style. Passages in three-part
texture alternate with phrases in two-part texture where the middle voice
keeps extended silences. Within each section, however, part-writing is kept
consistent. Most of the techniques characteristic of Baroque polyphonic
style are used: strict or free imitation, full and partial sequences of motifs
and models, inversion of shape, and inversion of voices.
The design shows two repeated halves, the second considerably longer
than the first. The compound time signature 12/16 is reminiscent of a gigue.
The uninterrupted flow of 16th-notes and the frequent patterns in gigue
rhythm (pairs of eighth-note + 16th-note, as in the first halves of mm. 2-4
etc.) corroborate the impression that this prelude is composed along the
lines of the Baroque dance genre that traditionally wrapped up a suite.1
The first harmonic progression concludes at m. 31 where the tonic is
confirmed in a perfect cadence. This cadence should not be considered as
structurally relevant, both because it marks the end of the first three-part
combination of the main motif, and because the tied notes in the upper and
lower parts essentially weaken the conclusion on the downbeat. Harmonic
closes indicating structural caesuras occur at m. 171, where the subdominant
key of E major is established, and in m. 32 at the repeat sign, which confirms the dominant key of F major. These two perfect cadences thus divide
the preludes first half into two segments of almost identical length.
An analysis of the material presented in this prelude reveals several
smaller sections that are essential for the works layout without being
harmonically conclusive. It is therefore advisable to describe the design on
the basis of combined observations of harmonic processes and material
presentation.
There are twelve (four + eight) sections in this prelude:
I mm. 1- 812
B major to F major (imperfect cadence)
II mm. 9-171
F major to E major (perfect cadence, subdom.)
III mm. 17-281
E major to F major (imperfect cadence)
IV mm. 28-3212 F major (perfect cadence, dominant)
1
In Book II of the WTC, Bach thus includes a whole suite among his preludes: an allemande
in D minor, a courante in E major, a sarabande in F major, and a gigue in Bmajor.
537
538
WTC II/21
V mm. 33-411
F major to F minor (perfect cadence)
VI mm. 41-4812 to G minor (perfect cadence, tonic relative)
VII mm. 49-5212 B major (imperfect cadence)
VIII mm. 53-5612 to C minor (perfect cadence, subdom. relative)
IX mm. 57-6412 C minor to B major (imperfect cadence)
X mm. 65-761
modulation to F-major7
XI mm. 76-831
continuation of F-major7
XII mm. 83-8712 return to B major (tonic)
These sections encompass three analogous passages:
mm. 812-12 . mm. 5212-56 transposed
mm. 13-171 . mm. 37-411
transposed and varied
mm. 28-32 . mm. 83-87
transposed
Another feature that is vital for a first overview is the sequence. Three
categories can be distinguished: sequencing in all three voices, sequencing
in the leading voice with free contrapuntal work in the other voice(s), and
sequencing involving the inversion of voices.
model
mm. 9-10
mm. 13-14
mm. 19
mm. 221-227
mm. 237-241
mm. 25
mm. 65-66
mm. 72
sequence
mm. 11-12
mm. 15-16
mm. 20
mm. 227-231, 231-237
mm. 241-247, 247-251
mm. 26
mm. 67-68
mm. 73
model
mm. 33-34
mm. 37-38
mm. 41
mm. 457-467
mm. 57
mm. 62
mm. 70
mm. 767-787
sequence
mm. 35-36
mm. 39-40
mm. 42
mm. 467-477
mm. 58, 59, 60
mm. 63
mm. 71
mm. 787-807
The basic character of this prelude is rather lively. This is revealed

above all in the simple rhythm and in the allusion to the gigue. The tempo
should be fast enough to render the exuberance of the gregarious dance,
but not so rushed as to reduce the intricate imitative patterns to a mere
technical show. The score shows several ornaments. As the two compound
symbols in mm. 7 and 26 indicate, Bach expects four trill notes against each
of the 16th-notea good measure for the limit of maximum speed.
m. 7
m. 26
B major
539
In addition to these trills, there are three more ornaments. The mordent
in m. 2 begins on the upper auxiliary and contains four notes, while the
one in m. 8 is approached stepwise, begins on the main note, and needs
only a single three-note shake. The inverted mordent in the principal motif
appears in brackets, which points to its origin not from Bachs manuscript
but from a copy. While the choice whether to include or disregard it is thus
left to each performer, adopting it is, in fact, a good idea as it can be of great
help to listeners in the manifold versions of the main motif. If the inverted
mordent is incorporated in m. 1, it should be transferred to all measures
featuring the version of the motif in which the 16th-note scale segment is
complemented by three dotted eighth-notes in reverse motion, i.e., to mm.
3 (L), 5 (U), 8 (L), 49 (U), 65 (U), and 67 (U). What remains to be decided
is whether an inverted mordent is preferable in all cases or whether the
inverted motif (e.g., L: mm. 3-4) might not benefit from an inversion of the
ornament (i.e., a simple mordent).
The thematic material of this prelude is determined by four motifs. M1
comes in numerous guises, which all share the beginning after a strong
beat and the scale section in 16th-note leading to the next strong beat. With
five regularly appearing variants, this motif dominates a large portion of
the prelude (M1a = U: mm. 1-21, M1b = M: mm. 18-27, M1c = L: mm. 2-31,
M1d = U: mm. 222-8, M1e = L: mm. 31-321). M2, first heard in mm. 9-10,
consists of two contrapuntally dependent parts. It can be heard as a pedal
note with above it an embellished descent in the alternating strong beats of
the two voices (see the octave parallel B-A-G-F, the first three notes with
written-out inverted mordents, interspersed with broken chords). In pitch
pattern, length, and texture M2 is thus distinctly different from M1. The
dynamic shape follows the line formed by the peak notes and thus describes
a two-measure diminuendo. M3 consists of a one-measure model in m. 13
that is sequenced three times in mm. 14-16. The texture is reminiscent of
toccata-style. The right-hand pattern can be described as hidden two-part
structure (see, e.g., m. 13: A-G-G before the harmonic background of a
G-minor broken chord). The left-hand part accompanies with another
two-part pattern consisting of cadential notes in the bass interspersed with
upwards leaping fourths in the high register. M4, a two-measure unit
restricted to the upper voice (mm. 28-29), is similarly designed in hidden
two-part structure: as an ascent above a pedal C ornamented with inverted
mordents. M5 (mm. 58-63, with imitation and inversion) is purely local.
The following table shows these motifs, albeit without specifying
inversions. Cases of voice swapping are marked with *, developments or
extensions with +, modifications with '.
540
WTC II/21
mm.
1-31
3-51
5-77
77-8
9-10
11-12
13-16
17-18
19-20
21-227
22-237
237-281
motifs
M1a, M1b, M1c
M1a, M1b, M1c
M1a, M1b, M1b
M1a, M1b var
M2
M2*
M3
M1b, M1b, M1b, M1b
M1d, M1c, M1d, M1c
M1b, M1b, M1b
M1d, M1c, M1d, M1c
M1+
28-301
30-32
M4, M1b
M1e, M1e, cadence
mm. 33-34
mm. 35-36
mm. 37-40
mm. 41-48
mm.
49-521
52-531
motifs
M1a, M1b, M1b, M1b
M1a, M1c
53-54
55-56
M2
M2*
57-581
58-647
M1d, M1c
M5, M1+
647-671 M1a, M1a, M1b, M1b', M1b'

67-701
M1a, M1b, M1b', M1b', M1b
71-767
M1b'+
767-831 M1+
83-851
M4, M1b
85-87
M1e, M1e, cadence
M4 in indirect parallels
M4 in indirect parallels
M3
M1+
The design of this prelude in the character of a gigue thus reveals

itself as a remote relative of sonata form. This becomes particularly obvious
when one combines structural observations derived from the distribution of
the motifs with those based on the main harmonic steps in the piece:
mm.
exposition section I
1-16
exposition section II
17-32
development section
33-48
recapitulation section I 49-64
recapitulation section II 65-74v83-87
insertion (in lieu of coda) 74-82
motifs
M1, M2, M3
M1+, M4
M4+, M3, M1+
M1, M2, M5
M1+, M4
harmonic function
tonic subdominant
dominant
modulating
tonic
tonic
modulating
WTC II/21 in B major Fugue

Launched after an eighth-note rest and concluding at m. 51, the subject
of this fugue spans exactly four measures. The beginning is particularly
noteworthy as the first note does not form part of the tonic chord. Among
the forty-eight fugues of the Well-Tempered Clavier, there are only two
B major
541
that set off from a note outside the home-key chord: The fugue in F major
(WTC II/13) begins on the leading note, thus approaching the key note
from below, while this fugue in B major begins on the second scale step,
preceding the tonic from above.
Phrasing within this subject is regular. Four segments, each of onemeasure extension, are grouped in the pattern of figure/sequence, figure/
sequence, with the effect that, despite a lack of outstanding rhythmic or
melodic features, the subject is easy to remember even for listeners. The
rhythmic pattern within the subject consists exclusively of eighth-notes.
Throughout the fugue, these are joined by quarter-notes and occasional
longer note values, but the overall effect remains one of simplicity. An
analysis of the pitch pattern reveals that the first subphrase and its varied
sequence are based on broken-chord patterns (see, e.g., B-F-D in mm. 1-2,
with B ornamented by a written-out turn). The third subphrase and its
sequence, by contrast, are composed as three-note descents of iambic note
repetitions. There is thus an
underlying structural skeleton with a distinct rhythm
behind the apparent uniformity of eighth-notes. Highlighting this may present a
challenge for performers:
The harmonic design of the subject is also simple. Although slight
differences occur in the course of the piece, the progression consists
mainly of tonic (mm. 1-31), subdominant (mm. 32-41), dominant-six-four
(m. 42), dominant-seven (m. 43), and tonic (m. 51). The dynamic outline
follows the dance-like design: the first two subphrases increase slightly,
while the next two provide the corresponding relaxation.
The B major fugue encompasses ten subject entries.
4. mm. 21-251 L
7. mm. 47-511 L
1. mm. 1-51 M
5. mm. 32-361 M
8. mm. 54-581 M
2. mm. 5-9 1 U
6. mm. 40-441 U
9. mm. 63-671 U
3. mm. 13-171 L
10. mm. 78-821 U
Apart from an interval adjustment in the answer at m. 53, the subject remains unchanged. Only the final statement features harmonic variation: it
opens and closes in F major but diverts in between to the minor mode.
542
WTC II/21
Bach gives the subject several companions. In the course of the second
to fourth entry one recognizes two recurring patterns, replace by two others
from the fifth subject statement onward. The first two do not meet the
requirements of true counter-subjects, i.e., to be melodically characteristic
and independent from the subject. Their recurrence, however, requires
them to be mentioned. They will be referred to in lower-case letters, as a
reminder of the above reservations: In its most basic form, cs1 consists of
nothing but a protracted dosido formula. As such it appears in M: mm.
14-171 and, with a diverted resolution, in U: mm. 22-25, while mm. 62-91
function as a precursor. Introduced in U: mm. 14-171, cs2 doubles the
subjects second to fourth subphrase in parallels of differing distance.
CS3, the first full-fledged counter-subject, is first heard in U: mm.
33-36. Consisting of three sequential ascents whose syncopations avoid the
subjects phrase-cuts, its dynamic equivalent is a protracted increase of
tension up to the final tie (see, e.g., m. 353), followed by a relaxation that
is either explicit (as in m. 36: G-F) or implicit (as in m. 44). CS3 recurs
faithfully in every further subject statement, although with several modifications including an abridgement at the beginning, a belated resolution,
and a transposition a fifth down. CS4 is paired with CS3 in a manner
similar to cs1/cs2. Introduced simultaneously in the lower voice, it moves
in half-notes and dotted half-notes which sound like a widely spaced
cadential-bass pattern. A dynamic rendering would have to consider the
inherent tension between harmonic steps, with the result that the climax of
CS4 coincides with that of the subject, once again restricting polyphonic
independence. Occasionally shortened like CS3, CS4 also recurs in each of
the remaining entries. Surprisingly given its similarity with a cadential
pattern, it not only wanders through all three voices but also changes its
pitch position with reference to the subject: while in mm. 33-36 it confirms
the subject statements F-major key with a pattern in F major, it ends
rooted in D in entries of quite different harmonic definition, like those in
mm. 40-44 (B major) and 47-51 (G minor).
All entries except for the first are followed by a subject-free passage:
E4 mm. 36-401
E7 mm. 58-631
E1 mm. 9-131
E2 mm. 17-211
E5 mm. 44-471
E8 mm. 67-781
E3 mm. 25-321
E6 mm. 51-541
E9 mm. 82-93
With regard to their material and harmonic formulas we can ascertain
the following subdivisions:
= mm. 58-601-631
E2a/E2b = mm. 17-191-211 E7a/E7b
E3a/E3b = mm. 26-291-321 E9a/E9b/E9c = mm. 82-861-901-93
B major
543
This fugue challenges interpreters more than many others with the task
of distinguishing episodes from thematic phrases since the secondary
passages contain so many fragments of the subject and its counter-subjects.
Thus E2 begins like an extension of the preceding entry, and E5 takes up
the subjects first three measures including the early quasi-contrapuntal
companions. E8, quoting the two distinct subject segments immediately
after the return modulation to the home key, gives the impression of a false
entry. Other episodes establish analogies among themselves. Thus E6 and
E7 are remotely related to E2a (mm. 17-191), E9b (mm. 86-901) is reminiscent of E3a (mm. 26-291), and E9c (mm. 90-93) is an exact transposition of
E3b (mm. 29-321). The role each episode plays in the dynamic design of
the fugue is determined by the rising or falling direction of the sequences
and secondary lines.
The simple rhythmic pattern, the ornamental concept of the subject,
and the comparatively low degree of contrapuntal activity all indicate a
playful fugue of rather lively character. The tempo may be fairly swift.
Feeling a whole-measure pace brings a good result, but care should be
taken that the slurred pairs retain their traditional active-passive or
heavy-light patterns and do not sound hammering. The relative tempo of
the prelude to the fugue may establish a proportion between the larger
beats: half a measure in the prelude corresponds with one measure in the
fugue. (Approximate metronome settings: dotted quarter-notes in the
prelude = 84, quarter-notes in the fugue = 126.)
The articulation includes non legato for the quarter-notes and other
longer note values, a crisp quasi legato for the eighth-notes outside the
slurred patterns, and true legato for the slurred note-pairs in the second half
of the subject. Note, however, that Bach indicates these slurs only once.
According to the conventions of Baroque performance practice, it was the
interpreters responsibility to act accordingly in all corresponding cases. In
the case of this fugue this entails that analogous note pairs must be
articulated analogously. This regards the second half of all subject entries,
their recurrences in episodes, and the parallel segments from cs1.2
2
It pays to be precise. Only seconds that in a slow tempo would be identifiable as

appoggiatura-resolution are slurred. The subjects third segment comprises what in another
mood would be regarded as extended sigh motifs: an iambic note repetition with
resolution, which turns into a new note repetition followed by its resolution, and an eventual
concluding tone. Where the whole extended sigh motif is sequenced (as happens here), the
concluding tone and the new upbeat are not slurred. As a result, the lower voice inserts cuts
at B-E and C-F in mm. 151 and 161 while the upper voice slurs all note pairs.
544
WTC II/21
Outside the confines of thematic quotations, slurring is based on the

harmonic relationship between two notes, and phrasing is dependent on the
components origin.3
The score contains only one ornament, a mordent in the lower voice of
m. 26 which is notated in brackets to indicate its origin from a copy. If
played, it begins on the main note and contains a single three-note shake.
As it bears no melodic or structural relevance and is not even imitated, it
can be omitted without any loss for the composition.
The entering order of the subject statements, the cadential closes within
the episodes, and the introduction of two new contrapuntal companions
from m. 32 onward underline the design of this fugue. There are three
sections. Section I encompasses the two initial statements, a bridging
episode (E1) and the third entry followed by a cadential close in the home
key (E2a). The second half of the same episode then represents a change
of mind in that it undertakes a new start with a modulation and thus
prepares for a redundant statement. Thereafter, a slightly longer episode
with two dynamically decreasing segments (E3) completes this section
with a strong cadential close in the dominant key F major. With regard to
material, this section is unified by the exclusive appearance of cs1 and cs2
as companions to the subject.
The beginning of section II is marked by the simultaneous introduction
of the two true counter-subjects CS3 and CS4, which stay with the
subject thereafter. The section contains three subject statements. They are
linked by short episodes and completed by a passage that takes up the first,
aborted cadential close of the fugue (compare E6 with E2a), concluding
here in E major, the subdominant of the fugues B major. The fact that
this cadence remains unresolved in the middle voice ties the second and
third sections together. This impression is reinforced by the use of the same
contrapuntal material for both sections, in contrast to the first section.
The third section consists of three statements separated by long
episodes. The observation that the third statement is redundant (it repeats
an upper-voice entry), together with the recognition of a false lowervoice entry in the preceding episode (E8), leads to the assumption that
section III is built along the same lines as section I, thus presenting a
disguised recapitulation.
3
Phrasing is easily overlooked between the first two eighth-notes in L: m. 17, M: m. 18,
L: m. 36, L: m. 39, M: m. 52, U: m. 58, L: m. 61, L: mm. 70, 72, 74, 75, and M: mm. 83-84.
Conversely, the first two eighth-notes must be slurred owing to their harmonic relation in
U: m. 19, L: m. 20, M: m. 37, M: m. 39, M: m. 51, and L: m. 59.
B major
Section I
M
U
L
L
545
B major
F major
B major
F major
tonic
dominant
tonic
dominant
dominant
tonic
tonic parallel
main round of entries
Section III
M
E major
subdominant
U
c minor
subd. parallel
(L
false entry in episode
U
F major
dominant
Section II
M
F major
U
B major
L
g minor

redundant entry
redundant entry
This is a playful fugue in which powerful build-ups of tension are not

the issue. The alternation of subject statements and slightly lighter
episodes determines most dynamic processes. Additional developments are
created by the dynamic gesture conveyed in each episode and by the
change of mode in the subject statements.
Within the first section, the tension rises from a graceful but not too
subdued beginning through three subject statements and a bridging
episode. It subsides during the cadential close of E2a, only to be picked up
546
WTC II/21
immediately afterward and brought to a first climax in the redundant

lower-voice entry. Owing to its pitch position a fourth below the previous
one, this statement gives the (deceptive) impression of a fourth-voice entry.
The seven-measure E3 provides a gradual relaxation.
Section II begins, despite its one-measure long rests in the upper and
lower voices, in a more assertive color as the subject is now enhanced by
new and independent counter-subjects. Despite a dynamically diminishing
episode, the second entry maintains the elevated tension, mainly because
of its exposed pitch position in the highest register of the Baroque
keyboard. The next episode with its increase then provides a powerful
preparation toward the first minor-mode statement in this fugue which, due
to the particular character of the subject, is for once much heavier and
more powerful than the original in the major mode. The tension subsides
only in the final episode of this section.
The first entry of section III combines the parallel motion of cs2
(extended here to two measures and even briefly joined in m. 55 by a third
parallel in the lower voice) with the two self-assured counter-subjects. This
statement can thus be interpreted as representing the climax of the fugue.
The inverted curve of E7 leads to another minor-mode statement which,
also reinforced by a cs2-parallel in its first measure, picks up the high
tension of the previous statement with slightly different means. The long
episode E8 with its false subject entry then provides a gradual relaxation.
The redundant statement enters into this considerably lowered tension. It
sounds much softer than its predecessors, not least because its first two
measures are conceived in reduced ensemble. The first segment of E9
thwarts an immediate decline of the tension, but the final eight measures of
the fugue lead to a completely relaxed ending.
WTC II/22 in B minor Prelude

The B minor prelude is written in polyphonic three-part texture. Two
motifs pervade the piece: one substantial, appearing much like a theme,
the other short and versatile, including inversion and multiple sequential
patterns. In view of the fact that the second entry of the principal thematic
component occurs on the dominant, this prelude could be described as
composed along the lines of a three-part fugue.
There are several distinct cadential patterns that indicate one aspect of
the structure in this piece. The following six harmonic units emerge:
I
mm. 1-161 tonic to minor dominant (i-v), B minor to F minor
II mm. 16-241 back to the tonic (v-i), F minor to B minor
III mm. 24-421 on to the dom. relative (i-VII), B minor to A major
IV mm. 42-551 to the subdominant (VII-iv), A major to E minor
V mm. 55-701 back to the tonic (iv-i), E minor to B minor
VI mm. 70-83 tonic confirmed (i-i), B minor
Two substantial analogies can be ascertained:
mm. 1-161 . 55-701 transposed a fifth down, U + M inverted;
mm. 36-402 . 77-812 each voice transposed and slightly varied
Given the alla breve meter, the simple rhythmic pattern dominated by
eighth- and quarter-notes, and the broken-chord pattern characterizing the
subject, the basic character of this piece is rather lively. The tempo may
be very fluent. The half-note beats should be perceived as swinging but
unhurried. Articulation combines a light legato touch in the eighth-notes
with gentle non legato in the quarter-notes. Exceptions occur only in the
dosido formulas, which must be legato (see L: mm. 1-3, M: 8-10, U:
14-16, etc.). No ornament is indicated in the score. It is, however, a good
idea to add one embellishment that would have been understood as implied
by any performer at Bachs time. The dotted note on the middle beat of m.
82 represents, together with the anticipation of the resolving keynote, a
typical closing formula. This was conventionally ornamented with a
mordent. This mordent begins with the upper neighboring note B and
contains two fast shakes, stopping short before the end of the sixth eighthnote of the measure.
The subject of this fugue is introduced in the middle voice. With its
extension of four and a half measures and its interesting phrase structure,
it does not have to fear comparison with officially declared fugal subjects.
547
548
WTC II/22
Its first subphrase [a] begins with a gentle curve in eighth-notes before it
turns to quarter-notes. Reading the figure D-C-B-A-B as D + turn figure
on B, one recognizes that the whole subphrase consists of broken chords
representing a complete harmonic progression: D-B-F, G-E-C, F-D-(B).
The harmonic progression ends at m. 31. That the D ends the phrase
despite marking the center of a broken triad will only become apparent in
two later statements, in which the two subphrases are entrusted to different
voices (see from m. 42-52: L/M, M/U). The overall descent in subphrase [a]
is dynamically best represented in a continuous diminuendo. The second
subphrase [b] begins with the rising octave leap followed by what could be
described as an ornamentation of the simple group B-C-B. This segment
is once again harmonized as a complete progression. The tension grows
toward the last downbeat, to be released in the final ornamented resolution.
(On the level of minute shaping, the crescendo evolves in two staggered
increases, rising twice through the sequenced tripartite note repetitions and
climaxing on the next downbeats appoggiatura before one brief and one
longer relaxation).
ii
VI
iv
There are altogether eight complete statements of the subject. In a

few cases (marked below with an asterisk), the second subphrase seems
extended. Harmonically, however, these extensions do not form part of the
statements, and they are therefore not included in the table below. In two
cases, as mentioned above, the two subphrases are given to different voices.
1. mm. 1- 52
M
5. mm. 42-441/44-462 L/M*
2. mm. 8-122
U
6. mm. 48-501/50-522 M/U
3. mm. 25-292
L*
7. mm. 55-592
U
4. mm. 31-352
U*
8. mm. 62-662
M
In looking for recurring contrapuntal figures one finds that the first
subphrase is accompanied by a dosido formula in the first two and last
two entries of the subject, while the second subphrase is twice surrounded by two shortened, transposed, and varied versions of subphrase [a]
(see the variation of M: mm. 1-31 in L: mm. 3-51 and in U: mm. 32-42; see
also M + L: mm. 57-591 with mm. 55-571).
A second thematic component, serving as an independent motif in this
prelude in the style of a fugue, is introduced in m: 7-81 as a link to the
second statement of the subject. It describes an extensive upbeat to the
B minor
549
subsequent downbeat and is almost omnipresent, filling measures in which

the subject remains silent as well as accompanying some of its entries. This
motif appears in inversion (as in U: mm. 24-251, etc.), with variation of its
opening or closing intervals (as in M: mm. 11-121 or L: mm. 16-171 and
20-211), and repeatedly builds juxtapositions in contrary motion (as in mm.
24-251, 41-421, 54-551, and 77-781). Structurally relevant in this composition are also the two pedals emerging in the final section (see the fourmeasure dominant pedal in L: mm. 73-751/752-771 followed by the nearly
equally extensive tonic pedal in U: mm. 77-801).
When trying to describe the layout of the B-minor prelude with the
vocabulary established for the structure of a fugue, one detects a design in
three sections. Section I in mm. 1-421 encompasses four statements in two
pairs: M, U and L, U. The shorter section II in mm. 42-551 is distinguished
by the two statements whose subphrases change voices. Section III in mm.
55-83 concludes the piece with two further intact entries in U and M and a
passage with secondary material and the above-mentioned pedals.
WTC II/22 in B minor Fugue

Beginning on the downbeat of a measure in 3/2 time, the subject spans
more than four measuresincluding three restsbefore closing at m. 51.
The contour displays predominantly small intervals interspersed with only
two high-tension leaps: the tritones E-A in m. 2 and G-C in m. 4. The
rhythm includes three different note values: half-, quarter-, and eighth-notes.
550
WTC II/22
Of particular impact are the rests. All three are equal in duration and metric
place: a quarter-note each falling on the first half of one of the weaker
beats in a measure (see mm. 13, 23, and 42). These interruptions in the
melodic flow raise a crucial question regarding the structure of the subject:
Do all or only some of the rests mark the end of a subphrase? In the case of
the first rest, this seems easy to answer. This silence cuts into an ascending
line (A: mm. 1-2: B-C-D-E). Complemented by a falling tritone and its
resolution, which returns to the keynote B, the ascent presents the first
half of a perfect curve that exerts a strong unifying force. The rest in m. 1
thus has tension-sustaining function. Conversely, the silence following the
return to the keynote in m. 23 marks a breathing point after the end of the
first subphrase. Finally, the third case is ambiguous. On the one hand, the
figure built by the three notes at the beginning of m. 4 depicts a releasing
gesture and thus allows the consideration of another breath. On the other
hand, mm. 3-4 can be identified as a varied sequence of the first subphrase,
preceded by a quarter-note upbeat and augmented so as to extend into the
quarter-note on the next downbeat:
Bachs subject
its skeleton
tritone + resolution
tritone + resolution
On the basis of this observation, the varied and partially augmented

sequence could also be taken as a larger unit. In this case the rest in m. 4
would be interpreted as a second tension-sustaining rest.
The subjects harmonic background, although somewhat difficult to
discern among the highly chromatic lines, can supports both concepts. As
the final entry in mm. 96-100 shows particularly clearly, the harmonization
of the first subphrase is repeated in the second. Observing that Bach ends
the first somewhat inconclusively with a major-mode tonic (i-ii-i-vii7-I) but
concludes the reiteration of the progression in the minor-mode, one could
argue for an external tail ([vii7] i-ii-i-vii7-i + vii7-V-I), although an internal
extension ([vii7] i-ii-i-vii7 + i-vii7-V + I) may seem more likely.
ii
i vii7 I
vii7 i
ii
vii7 i vii7 V7 i
B minor
551
When we combine these findingsthe layout in two corresponding

subphrases and the relaxing gesture at the beginning of the fourth measure
into a dynamic interpretation, we arrive at the following concept:
Within the first subphrase, the tension increases during the fournote ascent, finds its climax at m. 21 in a vii7 chord, and diminishes
during the return to the keynote. As the note before the tensionsustaining rest is simple and straightforward in its direction toward
the subsequent note, no interruption of the crescendo is desirable.
Within the second subphrase, the tension rises similarly throughout
the (ornamented) rise to the G at m. 41, which is again harmonized
by a vii7 chord. The rest, which occurs here not during the ascent
but after the climax, is preceded by two ornamental notes. These
present a small relaxation (on a secondary level) before the G is
picked up and a proper decline through the falling tritone and its
(ornamented) resolution is launched.
In the competition between the climaxes of the two subphrases, the
second one tops the first: structurally, the G is the peak of the
slightly longer and more complex subphrase and lies in a rising
sequence; melodically, it represents the sixth degree which, in all
minor scales, holds particular tension as a semitone above the fifth.
The entries of the subject in this fugue follow an extremely regular
pattern. There are altogether twenty-four statements:
1. mm. 1- 51 A
13. mm. 67-711 Tinv
2. mm. 5- 91 S
14. mm. 672-712 Sinv
3. mm. 11-151 B
15. mm. 73-771 Ainv
4. mm. 17-211 T
16. mm. 732-772 Binv
5. mm. 27-311 T
17. mm. 80-841 Sinv
6. mm. 272-312 A
18. mm. 802-842 T
7. mm. 33-371 S
19. mm. 89-931 B
8. mm. 332-372 B
20. mm. 892-932 Ainv
9. mm. 42-461 Tinv
21. mm. 96-1001 S
10. mm. 46-501 Ainv
22. mm. 96-1001 A
11. mm. 52-561 Sinv
23. mm. 962-992 Tinv
12. mm. 58-621 Binv
24. mm. 962-992 Binv
The subjects real answer is either identical with the original (as in T:
mm. 17-21) or differs only in its major-mode ending (as in S: mm. 5-9).
Other modifications concern the shape of the phrase ending, extended into
552
WTC II/22
a weak-beat (female) ending in three statements (see S: m. 9, B: m. 15,

and S: m. 56) but shortened in the two final entries. Exactly half of the
statements are inverted, and the focus on a symmetrical design is even
further emphasized with regard to strettos and parallels:
eight times does the subject appear uncoupled: four times in its
original shape, four times in inversion;
eight entries form simple strettos: four of them couple two original
statements each, the other consist exclusively of inversions;
four statements participate in mixed strettos made up of original
and inverted entries;
four statements sound in a combination of stretto and parallel.
Despite this extraordinary emphasis on the subject itself, Bach invents
a regular counter-subject as well as, in lieu of a second counter-subject,
another brief contrapuntal figure. CS1 is characterized by an interrupted
chromatic ascent (mm. 5-71: A-B-B-C, C-D-D-E-E-F), complemented
by an ornamented appoggiatura-resolution pair and, after another rest, its
falling sequence. The descent with its two segments veiling a chromatic
line (F-E, E-D) appears as symmetrical to the chromaticism in the earlier
ascent. Regarding the phrase structure and its dynamic equivalent, there is
thus a distinct contrast between the subject and its most prominent companion. The independence of CS1 from the subject is carried so far that the
counter-subject regularly closes two or three quarter-note beats after the
end of the subject (see, e.g., mm. 9, 15). CS1 faithfully accompanies each
of the single uninverted subject statements. Its inversion, which remains
incomplete, breaking off after the long note, supports each of the inverted
entries (see A: mm. 42-45, T: mm. 46-49, A: mm. 52-54, and S: mm. 59-61).
It makes a further short appearance in connection with the first mixed
stretto (see A: mm. 82-83).
CS2 is a very unusual companion. Far shorter than the subject that it
undertakes to accompany, it only materializes during the build-up to the
subjects second (and main) climax. But as if its very restricted scope was
not strange enough, its pitch pattern is also entirely vague. The only
consistent features are its rhythm and articulation: three quarter-notes,
introduced with wedges and followed each by a quarter-note rest, lead to a
final note of varying duration. More often than not, this note group is
preceded by an upbeat. Nondescript as this may seem, CS2 is nevertheless
a loyal companion. After its first presentation (see S: mm. 12-14) it recurs
eight times, accompanying original and inverted entries, single statements
and strettos alike.
B minor
553
The fugue encompasses thirteen subject-free passages. Owing to the

fact that the endings of subject and counter-subject do not coincide and
that strettos are so uncommonly frequent, there are several cases where
primary and secondary material overlap.
E1 mm. 9-111
E6 mm. 50 -521
E10 mm. 77-801
E2 mm. 15-171
E7 mm. 56 -581
E11 mm. 84-891
E3 mm. 21-271
E8 mm. 62 -671
E12 mm. 93-961
E4 mm. 31-331
E9 mm. 71-731
E13 mm. 100-101
E5 mm. 37-421
The material used in the episodes includes partial sequences of subject
and CS1 as well as independent episode motifs. Apart from E13 at the very
end of the composition, no episode is conceived as a cadential close. Each
of the motifs is limited to a single episode. Two are hybrids, insofar as they
can be traced back to more than one component. E12 is dominated by CS1
(see S/A: mm. 93-95), while E3 features a figure derived from a mixture of
the two counter-subjects (see B: mm. 21-22 from G to B as well as mm.
22-23, 23-24). One independent motif is introduced and imitated in E6
(see A: mm. 50-51 from A to D, T: mm. 51-52 from C to F), a second
one in E8 (see A: mm. 62-63 from E to F, S: mm. 63-64 from B to B;
similarly S/A: mm. 64-661, and S: mm. 66-671).
This fugue is conceived in basically calm character. Although the 3/2
time signature indicates that each of the half-notes represents an undivided
beat, these beats should swing in a stately manner. The tempo proportion
between the prelude and the fugue is a complex one. It can be attained in
either of two translation processes: one can equate two measures in the
prelude with a single measure in the fugue, or convert an assumed triplet
quarter-note in the prelude into an eighth-note in the fugue. (Approximate
metronome settings: half-notes in the prelude = 88, half-notes in the fugue
= 66.)
554
WTC II/22
The basic articulation in this fugue is legato. Several exceptions do,

however, occur: In the subjects first subphrase, Bach has marked the
unornamented ascent with wedges indicating an energetic abbreviation of
each note. Although the markings are not consistently repeated throughout
the composition, the same approach should be used in all further entries of
the subject. CS2 equally carries wedges on three of its few notes. These
seem particularly telling since they precede rests that would have interrupted the sound flow in any case. This shows that what Bach requests is
not so much an interruption but an energetic attack. This articulation of
CS2 should also be regarded as an integral characteristic and transferred to
all further statements. By contrast, it is the interpreters discretion whether
the wedged approach should be transferred to the occasional parallel in
equal rhythm (see A: m. 19, B: m. 44, and T: m. 60) and be retained in the
CS2-derived episode motif. (The latter is less likely since there are several
examples in other fugues where Bach repeats an indication if he desires it
in episode material.) Non legato articulation is appropriate for consecutive
leaps (as in S+T mm. 25-26, B: mm. 39-41 and 62-67) and for unexpected
extensions of the subject (see S: mm. 9, 56, and 84; B: m. 15) and their
sequences or imitations, as well as for all long notes in cadential-bass
patterns, whether occurring in an episode or a subject entry.
The score features only one ornament symbol, the cadential mordent in
m. 100. Approached stepwise, it begins on the main note, may comprise
five notes owing to the reduced tempo in the closing ritardando, and ends
on an eighth-note A preceding the eighth-note B.
The design of this fugue is so transparently laid out that it requires no
sophisticated analysis. There are five sections, each encompassing four
subject statements (one in each voice), one or more intermittent episodes,
and a closing episode. The composition ends with a coda comprising
another set of four subject statements as well as the final episode.
Section I contains four single entries in original shape (A S B T).
The two episodes linking the subsequent entries are both short and
entirely based on sequential prolongations. The closing episode E3,
by contrast, presents in its first four measures material that is more
developed before ending with two non-motivic measures. Harmonically the four entries are all in the home key (alternating tonic and
dominant), and even the final episode does not leave B minor.
Section II begins in three-part texture and only regains the full
ensemble after the intermittent episode. It contains two strettos.
These combine first the inner voices (T+A), then the outer ones
(S+B). Corresponding with the design of section I, the linking
B minor
555
episode (E4) is short and made up of sequencing material while the

closing passage is five measures long and presents more complex
motifs. This section begins in B minor, but the second stretto and
most of the closing episode (until m. 411) are in the relative major
key (D major). The final measure returns to the home key.
Section III introduces the subject inversion with four single statements. As in section I, the first two entries follow one another,
while the other two are each preceded by two-measure episodes
(E6, E7). With regard to material, however, these bridging episodes
appear much more independent than their earlier counterparts. The
closing episode (E8) is again five measures long. Harmonically,
these four statements modulate from the tonic to the subdominant
(E minor). Due to the inversion of the subject and its main counterpart, entries do not necessarily close in the key in which they set off
(compare m. 52 with m. 56 and m. 58 with m. 62). Yet, as in the
preceding section, the closing episode returns again to B minor.
Section IV corresponds with section II insofar as it consists of two
strettos, a short bridge in the center and a closing episode. It is
linked to section III in that it uses exclusively the subject inversion.
The two strettos refer to the tonic and dominant key respectively.
Section V combines in each of its two strettos one straight and one
inverted subject statement. Despite the unexpected length, the inner
episode develops as a sequential prolongation from the preceding
entry, while the closing episode (this time considerably shorter)
displays more independent material. Harmonically, the second of
the two strettos marks the final return to the tonic. The closing
episode with its descending soprano line, its repeatedly sustained
dominant note F in the lower voice, and its final dominant-seventh
chord prepares a confirming cadence in the home key. The expected
resolution at m. 961, however, turns this close into a deceptive
cadence. This provides a truly convincing transition to the fugues
short but extremely powerful coda.
The coda surpasses everything that has been heard so far by combining the mixed strettos of section V into a single four-part block
of two parallel original entries and two parallel inverted statements.
The latter are slightly shortened, giving way to a final cadential
formula that closes this extraordinary fugue.
There is a stunning analogy between the layout of this fugue and that
of the D-minor fugue from the first volume of the Well-Tempered Clavier
(compare the respective chapter on pp. 139-146 in this book).
556
WTC II/22
The development of tension in this fugue presents itself as a mighty increase in consecutive sweeps. After two overall crescendos, I < II, III < IV,
section V begins in greater intensity than any of the preceding sections and
gains even more dynamic momentum in its episodes rising sequences.
Only then does the tension abate somewhat toward the second mixed
stretto. The deceptive cadence should surprise listeners, and the coda may
outshine everything.
WTC II/23 in B major Prelude

This prelude consists almost exclusively of virtuoso patterns. Scale
passages, interplay of hands in toccata style and superimposed peak-note
lines determine the composition, while motivic material remains negligible.
The first relevant cadential close occurs in mm. 11-12 where it confirms
the modulation from the tonic B major to the dominant F major. Only five
and a half measures later, the prelude modulates to G minor, the relative
minor key of B major. As this harmonic close coincides with a change of
texture and pattern, it must be regarded as the second structural caesura.
There are altogether six sections. One of them (see the asterisk) ends in
an imperfect cadence. Given the return to the tonic in the center, the
prelude falls into two equal halves.
I
mm. 1-121
IV
B major to F major
II
mm. 12-173
Vvi F major to G minor
III
mm. 173-233
viI
G minor back to B major
IV
mm. 233-283
I iii* B major to D minor
V
mm. 283-331
iiiV D minor to F major
VI
mm. 33-46
VI
F major back to B major
With twelve and fourteen measures respectively, the two outer sections
are much longer than all others. They also complement each other in their
harmonic development (I: tonic to dominant, VI: dominant to tonic). The
two innermost sections, too, are analogous regarding their harmonic
progression: the modulation from G minor to B major corresponds to that
from D minor to F major. As will be shown, all these harmonic relationships are matched by correlations in material and texture.
The basic character of this piece is rather lively. The constant, never
interrupted 16th-note motion underscores this impression, as do the scales
and ornamented broken chords. With the exception of mm. 12-14 and
23-27, even the eighth-notes are confined to broken-chord figures. The
tempo that best expresses these characteristics is swift; the upper limit is
determined only by the ornaments, which should still sound clear and crisp
and not risk losing their metric precision. The appropriate articulation is a
light touch. An elastically rebounding non legato for the accompanying
eighth-notes and a sparkling, light quasi legato for the 16th-notes are the
predominant features. Broader non legato is used only for the few cases of
melodic eight-notes, and legato is confined to the ornaments.
557
558
WTC II/23
In the lively tempo one is likely to choose for this piece, the number of
ornaments required may pose a problem. It helps considerably to imagine
all trills, slides, and appoggiaturas as fully-written note values. Not only
does that take away some of the fear, it also clarifies their metric position.
This is important since trills in the left hand should have the same speed as
those in the right and slides like the one in m. 23, which is part of a
melodic figure, should be transferred to all recurrences of the same figure,
in this case to the two sequences. (A similar slide in m. 263 is spelled out,
probably because of the required accidental.)
mm. 1-2
The main features providing melodic orientation in this prelude are to

be found in peak-note lines. This is particularly true in sections I, II, and
VI. Section I consists of two distinctive patterns that build different kinds
of curves. The one in mm. 1-2 can be read as two large hooks: in the
right hand, there is an ascent over 1 octaves followed by a descent of
only half the scope. The left hand, beginning like an inverted imitation,
counters with a one-octave descent followed by a half-octave ascent. This
design, obviously geared toward an effect of brilliance, creates corresponding dynamic curves with climaxes in m. 2 on the downbeat (left
B major
559
hand) and the middle beat (right hand) respectively. The following longer
development is strung together by a peak-note curve in the right hand (see
the ascent/descent in mm. 3-6/7-121: C-D-E-F / A-G-F-E-D-C-B-A).
The left hand sets off with a hint of independence but then resigns itself to
mere accompaniment. In this homophonic pattern, the climax thus falls on
m. 71. In section VI Bach uses the same components for a somewhat different pattern, as seen in the right hands descending line in mm. 38-451.
The best dynamic renderinganything but easy given the long breath
neededwould be a continuous diminuendo from m. 38 to the end.
Section II begins with four measures in which the two uppermost
voices of a texture that is momentarily enriched to three parts play melodic
eighth-note groups forming a superimposed line. Section IV, the shortest
of the prelude with only five measures, is the melodically most intense. It
stands out even more than it would elsewhere because the sections surrounding it, III and IV, are composed in toccata style and thus form no
melodic lines, neither directly nor indirectly. By contrast, section IV is
conceived in consistent three-part texture and even presents a small motif.
In its upper voice (mm. 24-25), a richly embellished falling triad leads into
an octave leap before being complemented by relaxing eighth-notes in the
middle voice. The combination is sequenced, at first exactly, then with an
extension. After more sigh motifs and a quasi-cadential trill, an imperfect cadence concludes this section.
The following music example shows the lines in the sections discussed
abovethe preludes skeleton.
mm.
1 1
mm. 12
13
14
15
16
17
10
11
12
560
WTC II/23
mm. 17
18
mm. 23
24
mm. 28
29
mm. 33 34
19
20
21
25
30
35
26
31
36
22
37
32
38
39
23
27
28
41
42 43
33
40
44
45
46
WTC II/23 in B major Fugue

The subject of this fugue is designed with two endings: an accented (or
male) one up to m. 41 and an unaccented (or female) one concluding at
m. 51 . The male ending is determined by the point where the dominant
resolves into the tonic and thus concludes the perfect cadence in B major.
The notes filling the space to the downbeat of m. 5 serve as an extension
that can be interpreted as a metric complement, i.e., as extending the
subject so that it reaches a desired four-measure structure. Harmonically,
however, these notes (A-G-F-E-D) serve no independent function but
only prolong the tonic. It will therefore not come as a surprise that in the
course of the fugue, Bach modifies this female ending liberally, up to a
point where it assimilates the shape of episode material, and that he
eventually drops it altogether.
B major
561
In its main body, the subject consists exclusively of half-notes that

describe a single, indivisible gesture. Except for the final semitone, which
features the resolution of the leading-note into the octave, the pitch pattern
displays only intervals of a
third or larger. The harmonic
backbone is a simple cadential
progression, with one harmonic
step per measure:
Two different dynamic representations of this subject are conceivable.
How the development of tension is felt depends on the performers
susceptibility to harmonic over melodic processes. If the melodic ascent is
regarded as prevalent, then the subject should be taken as a single
crescendo, complemented by a short diminuendo wherever the female
ending retains its original falling line. Conversely, no relaxation occurs in
cases where the male ending is followed directly by episode material.
(Care must be taken not to break the line momentarily in the descent
G-E-C; such rendering would cut the subject in two.) If, however, the
harmonic design is perceived as essential, then the increase of tension
should lead to the subdominant function, the G at m. 21. This climax
would then be followed by a diminuendo through two or three measures.
(In this case, it is vital not to play the C in m. 31 more softly than the
ensuing A so as not to risk fracturing the subject.) As can be imagined,
these two contrasting interpretations of the dynamic gesture have a
far-reaching impact on the entire composition.
The fugue features altogether fourteen subject statements. The table
below indicates the male ending of each statement and shows an asterisk
where a female ending similar to that in the initial measures follows.
1.
mm. 1-41
B*
8. mm. 42-451
S
2.
mm. 5- 81
T*
9. mm. 48-511
B
3.
mm. 10-131
A*
10. mm. 53-561
T
4.
mm. 14-171
S*
11. mm. 60-631
T
5.
mm. 19-221
B(*)
12. mm. 75-781
B
6.
mm. 27-301
T
13. mm. 85-881
T
7.
mm. 35-381
A
14. mm. 93-961
S
The subjects trunk as listed above remains completely unchanged

throughout the fugue. Neither inversions nor stretto and parallel statements
are employed.
562
WTC II/23
The lack of modification in the subject itself leaves ample room for
regular contrapuntal work. Bach has invented three counter-subjects for
this fugue. Two of them appear exclusively in the first round of entries,
while the third accompanies the subject for the remainder of the fugue and
also pervades some of the episodes. CS1 is introduced in B: mm. 5-81 with
a female ending extending until m. 91. Its trunk consists of three measures
with similar patterns, laid out in sequences. The first subphrase (mm. 5-61)
begins with an upbeat to a syncopated half-note and concludes with three
eighth-notes in stepwise descent, thus describing a perfect curve. The
following two subphrases represent ascending sequences, each of them a
fourth higher than the previous one. The entire group should therefore be
rendered as an increase in intensity. Phrasing between the subphrases may
or may not be enhanced by actual cuts in the sound flow. Its expression by
dynamic means (relaxation followed by a new active rise) is, however,
crucial. The tripartite body is finally complemented by a female ending
that prolongs both the final harmony and the melodic descent at the end of
the third subphrase. CS1 recurs three times, once in each of the four voices
in ascending order following the textural rise of the subjects statements
(T: from m. 10, A: from m. 14, S: from m. 19). CS2 is first heard in B:
mm. 10-141. It also begins with an upbeat to a syncopation. But both this
note and its resolution are much more extended than in CS1, so that the
first subphrase spans two measures. A second subphrase follows with an
increase to m. 131 and a decrease to the final note.
These two early counter-subjects are designed to draw the listeners
attention to the subjects female ending. This is particularly obvious in the
case of CS2, which modifies the harmonic resolution on the final note of
the subjects trunk in favor of an interrupted cadence: see, e.g., m. 131
where the expected conclusion in B major is thwarted by the G in the
bass, which converts the chord into a G-minor triad (vi of B major) and
postpones the resolution to the final note of the female ending in m. 141.
or
B major
563
--
---
--
---
--
--
An identical process occurs in mm. 17-18. Interpreters who have opted for
shaping the subject along harmonic lines, i.e., with the climax on the
subdominant and the final tonic as a dynamic resolution, will find Bachs
diversion to the deceptive chord particularly rewarding to play.
CS3 is introduced in S: mm. 28-301 and accompanies almost all of the
subjects further entries. It consists exclusively of eighth-notes winding
downward in a pattern that can be read as an ornamented imitation of the
subjects central segment. This imitative relationship determines the dynamic design in this counter-subject, allowing for the same two solutions
that are possible in the subject:
or
The B-major fugue encompasses twelve subject-free passages. Owing

to the design of the subject with two endings, it must be determined where
exactly one should assume the beginning of secondary material. In entries
accompanied by CS2, the answer is obvious since the diversion to the
interrupted cadence and subsequent resolution do not allow ending with
the subjects trunk. In m. 22, however, CS2 is absent, the female ending of
CS1 is transformed to a tie prolongation, and the ending of the subject is
transposed, thus changing the harmony after the return to the tonic has
already taken place. In such a case, it must be assumed that m. 22 does not
belong to the subject-dominated field. In many cases later in the fugue,
however, no full return to the tonic occurs at all since the final tied note of
CS3 often remains unresolved. Thus mm. 30-35 do not contain a single
downbeat without a suspension in one of the voices.
In all those entries where the subjects original female ending cannot
be ascertained, episodes are therefore regarded as beginning after the
subjects trunk.
E1 mm. 9-101
E7 mm. 51-531
E2 mm. 18-191
E8 mm. 56-601
E3 mm. 22-271
E9 mm. 63-751
E4 mm. 30-351
E10 mm. 78-851
E5 mm. 38-421
E11 mm. 88-931
E6 mm. 45-481
E12 mm. 96-104
564
WTC II/23
The episodes display a variety of material. There are fragments of the

counter-subjects (particularly from CS1 and CS3), small independent
figures, and sequence models involving several voices. E1 and E2 are short
and serve to link subsequent subject statements. Their characteristic features
include a varied imitation of the subjects female ending (B: mm. 9-10, A:
mm. 18-19) and an anticipation of the first subphrase of CS1 (T: mm. 9-10
and S: mm. 18-19). In E3, the severed female ending and its sequence
(S+A+B: mm. 22-231, 24-251) are linked by a measure of neutral material
before this episode combines two quotations of the CS1-subphrase with a
dosido formula in the alto and a cadential-bass pattern. (The bracketed
tie in S: m. 25 is best ignored to ease recognition of the CS1 subphrase.)
E4 presents M1, an ascending tetrachord (S: mm. 30 and 31) in conjunction with two complete quotations of CS3 and neutral passages. The motif
is ubiquitous in E5 and E6, and it even invades the first two measures of
the intermittent subject entry (mm. 42-43). Its complement is a short version
of CS3 (A: mm. 39-40), a dosido formula (in E5), or a measure of
neutral material (in E6). E7 presents the first sequence model: mm. 51-521
recur descending in mm. 52-531. E8 combines CS3-derived material in the
bass with free contrapuntal material in the other voices before it endsthe
first episode to do so after E3in a perfect cadence. E9 is the longest and
most independent episode in the fugue. It consists of three sequence models:
descending in mm. 63-651 . 65-671 and (shorter) 67-681; ascending in
mm. 683-693 . 693-703 and 703-713, descending in mm. 713-731 . 73-751.
E10 begins with several mostly inverted quotations of M1 (mm. 78-801).
These are followed by increasing ascents (S: mm. 80-81, B: mm. 81-82)
and rounded off by an embellished three-part closing formula that will
recur later in the fugue. E11 features another sequence model (A+T: mm.
89-911 . 91-931; B: mm. 883-903 . 903-923). E12 begins with several fournote groups that are remotely related to M1. An incomplete quotation of
CS3 (S: mm. 100-101) leads to a transposition of the embellished threepart closing formula that concludes E10. To sum up: three episodes end in
a perfect cadence with complete resolution of all voices (E3, E8, and E12).
A fourth (E10) must be regarded as structurally corresponding owing to
the relationship of its close with that of the final episode.
Apart from the correspondence between the two embellished closes,
there is another structural analogy. The first two episodes linking subject
entries accompanied by CS1 employ material from CS1 combined with a
line derived from the subjects female ending. Similarly, the first two
episodes connecting subject entries accompanied by CS3 quote CS3, again
combined with a frequently recurring short line in eighth-notes (M1).
B major
565
In view of the rhythmic complexity (the half-notes in the subject are

complemented in other components by quarter-notes, eighth-notes, and all
kinds of syncopated and tied notes; even 16th-notes appear in one of the
sequence models within E9), the basic character of this fugue is rather
calm. The tempo is determined by this character, but also has to take into
account the alla breve time signature. Generously swinging half-note beats
may convey the idea. The relative tempo of the prelude to the fugue is best
chosen in a proportion of 3:2, i.e., three quarter-note beats in the prelude
correspond to two half-note beats in the fugue. (Approximate metronome
settings: prelude beats = 96, fugue beats = 63.)
The articulation requires legato in all components that are essentially
melodic: counter-subjects, episode material, and the subjects female
ending. Cadential-bass patterns (as in B: mm. 25-27, 34-35, 59-60, 84-85,
and 103-104) are, as always, excluded from this rule. So are consecutive
leaps. The latter require special attention in this fugue. As the subjects
trunk consists exclusively of leaps, it should be taken in a slightly detached
style. To obtain the right duration of the articulated notes, together with a
good expressive tone quality (a detail often badly neglected in non-legato
playing), it may help to imagine the first, unaccompanied subject entry
played by a cellist who uses very broad bow strokes. Phrasing, too, is an
extremely important issue. This applies to the sequences in CS1, the partitioning in the middle of CS2, and the episode segments deriving from
primary material, but is equally crucial in the episodes sequence models.
Ornaments do not occur in this fugue.
The structural layout of this composition is clearly determined not only
by the cadential closes observed in four of the episodes, but also by
symmetries in the entering order of the voices and the consecutive build-up
of the ensemble. There are four sections with significant analogies.
Sections I and II, closing on the downbeats of mm. 27 and 60
respectively, contain five subject statements each: I: B T A S B and
II: T A S B T. Both sections encompass a full round of statements
complemented by a redundant entry. In both sections, the voices
enter in ascending order, with a return to the bass after the soprano,
and the redundant entry occurs in the voice that opens the section.
In both sections, the first three episodes are united by the use of
similar material (see E1, E2, and E3, which all use fragments of the
subject and CS1; and E4, E5 and E6, all featuring M1). In section I,
the fourth subject statement does not sound, as could have been
expected, in full four-part texture, since the bass rests from m. 14
onward. The full ensemble is thus postponed for the redundant
566
WTC II/23
entry. Similarly in section II, the completion of the ensemble is

delayed twice (from m. 42 onward, the alto and then the tenor rest)
and also materializes only with the redundant entry.
Sections III and IV feature two subject statements each: III: T B
and IV: T S. Despite this comparatively limited presence of the
principal thematic component, these two sections span 25 and 20
measures respectively. The analogies are: In both sections a tenor
entry in three-part texture (see mm. 60-64 and 85-88) is followed
by a subject statement in a (more exposed) outer voice and in full
four-part ensemble. In both sections, the linking episodes (E9 and
E11 respectively) consist exclusively of sequence models. In both
cases, the latter statement is preceded by a four-measure variation
of CS3 in the lowest voice (mm. 71-74 . 883-923). And in both
sections, the concluding episode ends with the same embellished
closing formula.
Moreover, the fugue as a whole displays a superimposed symmetry
created by harmonic progresses: within the second half of the fugue, the
progression in the harmonic development is exactly reversed:
Section I remains in B major. Not Like the entries of section I,
those of section IV are in the
until the concluding cadence does
home key, on the tonic and
Bach establish the secondary key of
dominant respectively.
F major.
In section II, F major is redefined Just as the beginning of section
II returns to B major, so the
as the dominant of B major. The first
end of section III returns to
three entries sound once again in the
B major.
home key.
The last two entries of section II turn Just like the conclusion of section II (and of the fugues first
toward other tonal anchors, sounding
half), the outset of section III
in G minor (the tonic relative) and
(and of the fugues second half)
E major (the subdominant) respecsounds in the subdominanttively, and conclude the first half of
key of E major.
this fugue in E major.
One can thus state that, while structural processes as created by the use
of material and texture establish very conspicuous correspondences
between each of the two section pairs, I + II and II + IV, thus seeming to
divide the fugue into two halves of very different content and density,
harmonic processes create an almost perfect axis symmetry, thus knitting
the two halves inextricably together.
B major
567
As shown in the detailed observations discussed above, each of the

four sections is conceived as an increase of tension. Within the first two
sections, this increase is very gradual as the full sound of the four-part
ensemble is delayed until the very end. The third and fourth sections, by
contrast, contain a fairly steep build-up of tension between their respective
entry pairs.
Within the first half of the fugue, there is almost no interruption in the
two processes of gradual increase since all intermittent episodes have
bridging function. The second half of the fugue, by contrast, features episodes dominated by independent material as well as by structural processes
of their own. These subject-free passages, which are also much longer than
those bridging consecutive entries in the earlier sections, thus create a
definite color contrast to the surrounding subject statements.
568
WTC II/23
When comparing the level of intensity in the first half of the fugue, the
second section appears somewhat lessened in comparison to the first. This
is due, on the one hand, to the fact that the entering order B T A S B
creates a stronger impact than that of its structural sequence T A S B T. On
the other hand and perhaps more importantly, section I features its five
subject statements surrounded by two independent counter-subjects, while
section II only contains a single regular accompaniment of the subject.
What is more, CS3 is entirely dependent on the subject in terms of both
pitch pattern and dynamic design. Last but perhaps not least, the change of
mode toward the end of section II also softens the increase in this section.
Within the second half of the fugue, the converse process can be
observed: section III contains longer episodes in contrasting color than
section IV, thus suspending the rise of tension between consecutive entries
more. Furthermore, the bass entry in section III is not accompanied by any
counter-subject, thus forfeiting some of its climaxing power. By contrast,
the corresponding soprano statement in section IV comes supported by
CS3 in the tenor and a varied version of CS3 in the alto.
The dynamic rendering of the B-major fugue should thus aim at
depicting the two-fold symmetry as eloquently as possible: the large-scale
structural analogies between sections I/II and III/IV and the harmonic axis
symmetry of the fugue with the correspondences of sections I/IV and II/III.
WTC II/24 in B minor Prelude

The final prelude in the Well-Tempered Clavier is the only one with a
tempo indication. Together with the alla breve meter, the word Allegro
hints at Bachs idea of swift half-measure beats. The mood is supported by
the strikingly regular phrase structure: all patterns take up four measures.
They are either neatly filled by thematic material or feature a 3-measure
motif complemented by a half-measure link. The only interruption of the
dance-like regularity occurs in mm. 57-58. These two measures are not
only extra (see the four-measure phrases in mm. 49-52, 52-56, and 59-62)
but additionally halted by a fermata.
The consistent two-part texture features voice-splitting only in the
final measures. The degree to which the two voices are independent of one
another varies: there are phrases with obvious accompaniment patterns
(e.g., mm. 49-52) and genuine contrapuntal settings (e.g., mm. 33-36). The
principal motif, the only one to recur in immediate imitation, is evenly
distributed between both voices. Other motifs display a strong preference
for the upper voice and do not make use of imitation, thus creating a more
homophonic style. The accompanying figure of each motif can take
various forms but contrapuntal interplay among the motifs themselves does
not occur. The prelude can thus be described as determined by motifs in
two-part setting.
Small-scale harmonic closures occur at the end of most four-measure
phrases, sometimes even within them, and therefore cannot be regarded as
relevant in the overall structural layout. The tonic is thus regained in V-i
cadences on the downbeats of mm. 3, 5, 7, and 9. Only thereafter is the first
modulation launched. In the absence of strongly determining harmonic
processes, structural analogies help to present an overview of the layout.
There are several corresponding phrases:
mm. 1-4 . 17-20 (transposed)
. 41-44 (transposed and varied,
with two measures in inverted voices)
mm. 5-8 . 29-32 (transposed, voices inverted)
mm. 9-131 . 45-491 (transposed)
mm. 13-171 . 49-531 (transposed, lower part varied)
mm. 21-24 . 59-62 (transposed, ending varied)
569
570
WTC II/24
On the basis of these analogies, the prelude can be described as comprising four sections:
I
mm. 1-16 M1 + imitation on the tonic (B minor)
II mm. 17-40 M1 + imitation on the tonic relative (D major)
and M1 + imitation on the subdominant (E minor)
III mm. 41-58 M1 on the dominant (F minor)
IV mm. 59-66 M1 on the tonic (B minor)
Given that the composer has prescribed the character he desired for
this piece, the first questions performers usually face when dealing with
Bachs music are already answered: the character is lively. Almost all
16th-notes are written-out ornamental figures. There are turns in different
metric positions (U: mm. 2 and 21-23, L: mm. 1-3, U: mm. 10 and 12),
inverted turns (mm. 9 and 11), and off-beat trills with diverting suffixes
(mm. 7 and 13-16). Regarding the articulation of the three note values
determining this piecequarter-notes, eighth-notes, and 16th-notesthe
distinction characteristic for compositions in lively character between the
faster and slower note values thus concerns the eighth-notes and quarternotes. The eighth-notes should be played in a light, transparent legato
touch, while the quarter-notes are mostly non legato. There are, however,
a number of exceptions: the light legato of the eighth-notes should be
interrupted where escape notes separate slurred note-pairs (thus in U: m.
26, the F does not form part of the main melodic line). And quasi legato
must give way to dense legato where this is explicitly indicated, as in U:
m. 58. Particularly easy to overlook are interruptions of the eighth-note
flow owing to phrasing. This is particularly crucial in U: mm. 13-16 after
the downbeats and in mm. 23 and 61 before the final eighth-note where the
last of three almost identical sub-phrases must be separated from the
typical upbeat + appoggiatura-resolution group. In the quarter-notes, the
usual non legato touch must be suspended for all note-pairs constituting
appoggiatura-resolution.1 Other quarter-notes to be played legato are those
in the traditional dosido formula (U: mm. 65-66). The quarter-notes that
are marked with wedges (U: mm. 21-22 and 59-60) are not necessarily
shorter than those without but heavier, pointing toward an overall climax
that, after having been twice avoided (in the downbeats after the wedged
notes), is finally reached in the appoggiatura-resolution group after the
third sub-phrase.
1
These may or may not be indicated as slurred: compare U: m. 8 and L: m. 32 with U: m.

24. See also the syncopations in U: m. 37 with sequences and imitations in mm. 38-39 as
well as similar figures in mm. 53-54, and at m. 641 the three-voiced E-F, D-C, B-A.
B minor
571
In addition to the ornamental figures written out in 16th-note values,

the score features a number of embellishments. The mordents on the weak
part of the appoggiatura-resolution pairs in mm. 8, 24, and 32, printed in
brackets and thus obviously deriving not from the manuscript but from an
early copy, should only be played by performers who can trill very softly
without disturbing the relaxation. The same holds true for the mordents
that, in U: mm. 34 and 36, grace notes in the middle of a gradual decrease.
If executed, all these mordents begin on the main note since they are
approached stepwise, and comprise three notes. The inverted mordent in
m. 31 might be interpreted as a means of emphasis; one can easily imagine
the composer-teacher adding it in the score of a student who failed to
properly accent the syncopation in the lower part. The turn in U: m. 32, if
playedit is also given in bracketsbegins on the main note and thus
encompasses five notes: four 32nd-notes and a final 16th-note. The gracenote in m. 57 designates an appoggiatura of eighth-note duration, possibly
with a minimal slow-down for this note. Its resolution, reduced to eighthnote value, is definitely lengthened by the fermata.
The thematic material consists of three motifs. They represent lessening degrees of importance, diminishing polyphonic independence, and
gradually reduced complexity. M1, the principal thematic component, marks
the beginning of all four sections. It consists of two identical subphraseheads (mm. 1-21 . 3-41) that are linked in the latter part of m. 2 and
complemented with a tail in form of a descending broken chord in m. 4. In
both subphrases, the initial measure is composed as a circling motion
around the note D, and the climax falls on the following downbeat C.
Among the two climaxes, the first is more important. The second, a mere
repetition, is slightly weaker. Harmonically, M1 is interesting in its design
as each subphrase ends on the dominant, while the tonic is only transitorily
regained at the beginning of the second subphrase. At the end of the motif,
a link is required to return to both the downbeat position and the tonic. M1
is accompanied by four different contrapuntal companions (countermotifs or CM), two of which are remotely related. CM1a (L: mm. 1-42,
17-202, and shortened, 41-42) also comprises two subphrases of almost
identical design, with the climax falling onto the syncopations. CM1b (U:
mm. 5-8 and 29-32) is differently structured, consisting of three sequences
(the last of them varied) that build a single overall arch and thus do not
share the binary design of M1. CM1c (U: mm. 21-24 and 59-62) resembles
CM1b with regard to the rests on its first three downbeats, its rising peaknote line, and its conclusion in an appoggiatura pair, which serves as a
climax. A fourth companion is heard only once (U: mm. 25-28).
572
WTC II/24
M2 follows the two initial statements of M1 after a link (mm. 82-91).

It consists of a two-measure subphrase and its sequence, with the subphrase itself made up of a one-measure unit and its varied inversion. The
accompaniment in mm. 9-131 and 45-491 moves in simple lines of even
quarter-notes. In mm. 33-371, by contrast, the motif is countered with a
contrapuntal voice. M3, introduced in mm. 13-161 and taken up in mm.
49-531, is conceived as a one-measure element launched from the
syncopated second eighth-note of the measure, sequenced three times in
rising direction, and accompanied by neutral quarter-notes. The preludes
structure can thus be presented as follows:
Section I
Section II
Section III
Section IV
M1 + CM1a
M1 + CM1a
M1 + CM1a var
M1 + CM1c
M1 + CM1b
M1 + CM1c
M2 + CM2a
+ 4 mm.
M2 + CM2a
M1 + CM1d
M3 + CM3a
M3 + CM3
M1 + CM1b
+ 4 + 2 mm.
M2 + CM2b + 4 mm.
Section I is devoted to the introduction of the three motifs. Polyphonic
complexity decreases in the course of the sixteen measures, thus causing a
gradual drop in intensity. Section II is set up as a development. M1 recurs
four times, modulating and taking four kinds of companions. M2 is also
modified as it substitutes its neutral companion with a polyphonically
active one. The final four measures of the section develop the upbeat +
appoggiatura-resolution figure in a syncopated version. Section III is conceived as a recapitulation. While the two M1 statements from the preludes
beginning are here contracted into onesee the inversion of voices halfway through the four-measure unitM2 and M3 follow with their original
accompaniments. Moreover, the harmonic development also corresponds
to what was customary for the relationship between exposition and
recapitulation in Bachs time: the tonic of the opening measures recurs as
a subdominant at the outset of section III. Four measures with a falling
bass line (see L: mm. 53-56 B-A-G-F-E-D-C-B-A) and a free continuation of the upper voice lead into a dominant-seventh chord with fermata
and general pause before the two last measures of this section carry out the
return to the tonic. With regard to dynamics, the slight decrease in M1 and
M2 is counteracted here by a stronger increase owed to the modification of
M3, whose sequences no longer fall but rise in fourths. It is left to mm.
53-57 to complement this intensification with a relaxation. But even this
decrease is not yet complete by the time it is interrupted. The resolution
follows all the more stringently in the ensuing cadential close. Section IV
B minor
573
appears as a coda. It consists of a single statement of M1 followed by four

cadential measures in hardly disguised homophonic texture. Interrupted by
three further general pauses on the middle beats in mm. 62, 63, and 64,
these measures pick up the appoggiatura-resolution prominently featured
in the preludes middle section and lead to a final climax at m. 641.
This B-minor prelude, determined by thematic material made up of
two-part motifs, is thus laid out as a Baroque sonata movement with
exposition (mm. 1-16), development section (mm. 17-40), recapitulation
(mm. 41-58), and coda (mm. 59-64).
WTC II/24 in B minor Fugue

The subject of this fugue spans five and a half measures, from the
upbeat to m. 1 to the downbeat of m. 6. The combination of simple upbeat,
downbeat ending and 3/8 time evokes a lightly swinging character. There
are two subphrases: the initial falling tonic chord complemented by the
leading-note and a return to the keynote, and a four-measure embellishment of the descent G-F-E-D. The rhythmic pattern in the subject itself as
well as in the entire fugue is simple, consisting predominantly of eighthnotes and 16th-notes. The pitch
pattern contains not only broken
chords but also many melodic
octave leaps. The subjects harmonic design is a simple cadential progression that Bach lavishly
ornaments in later statements.
The dynamic design traces these
features. In the first subphrase,
the upbeat energizes m. 11 while
m. 21, representing the return to a
note already heard, is completely
relaxed. In the second subphrase, the peak-note line coincides with the harmonic design: the climax falls on m. 31, the representative of the subdominant and the beginning of the descent. The second of these two climaxes is
stronger than the first. In terms of phrase structure, the subject thus
presents itself as consisting of a smaller preliminary gesture followed by
a larger principal gesture.
574
WTC II/24
The fugue comprises ten subject statements, one of them incomplete:
1. mm. 0-61
M
6. mm. 44-501
L
2. mm. 6-121
U
7. mm. 54-601
M
3. mm. 15-211
L
8. mm. 70-761
L
4. mm. 26-321
M
9. mm. 81-871
U
5. mm. 35-411
U
10. mm. 96-981
M incomplete
Apart from the usual adjustments of the first interval in the answer, the
only modification the subject undergoes is the varied and extended upbeat
in m. 70. Neither inversions nor grouping into strettos or parallels occur.
Moreover, the polyphonic density never surpasses that of two thematically
shaped voices. While there are two counter-subjects, they never coincide.
CS1 only accompanies the subject in mm. 6-12, 15-21 and, partially, in
mm. 26-32. As soon as CS2 emerges, CS1 breaks off and disappears. (This
retreat of CS1 at the arrival of CS2 is repeated in m. 73). CS1 begins with
a varied imitation of the subjects first subphrase, continues in a trill with
anticipated main note and ascending resolution, and ends after a linking
scalar descent with the sequence of the same trill. Dynamically, the
imitating segment follows the design in the subjects first subphrase and
the trill prolongs the relaxation. The descending scale then serves as a new
upbeat, complemented by the second trill with a second decrease. CS2 is
first presented in L: mm. 29-321. Its features are most unusual for a contrapuntal figure in Bachs polyphonic style. Moving in even 16th-notes that
zigzag in large leaps, it resembles accompaniment patterns in homophonic
textures. Its hidden two-part design shows a bass line BAGC-F
below an ornamented descent A-G-F. The dynamic layout with a single
decrease presents no polyphonic contrast whatsoever to the subject.
B minor
575
The ten subject statements are interspersed with nine subject-free

passages. Some of them are extremely short, others span up to fourteen
measures. There are, however, good reasons to take even the very short
links seriouslynot only on paper but, more importantly, in performance.
This is required both by the particular design of this fugue with its
alternation of statements and episodes and by the beginning of CS1, which
imitates the subjects initial subphrase and thus enters late.
E1 mm. 6-7 (only M)
E6 mm. 50-551 (54-55 only L)
E2 mm. 12-16 (U + M)
E7 mm. 59-691 (59 only L)
E3 mm. 21-27 (U + L)
E8 mm. 76-821 (82 only L)
E4 mm. 32-361 (35-36 only L) E9 mm. 87-100
E5 mm. 41-441
Material occurring in the episodes includes motifs derived from the
counter-subjects (Mcs1 and Mcs2), as well as a number of independent
motifs. In addition, there are false entries (subject: see M: mm. 96-981;
CS1 see L: mm. 973-991). Interestingly, the motifs stemming from primary
material occur almost exclusively in the initial third of the fugue, while
those in the central third introduce more and more independent features as
secondary material, and those in the final third review material introduced
earlier. The following chart gives an overview of the material presented in
the episodes of this fugue, and the relationship among the episodes:
E1
link
E2
E3
E4
E6
E7
Ms/Mcs1
Mcs1a
M1/M1/Mcs2
M1/M1a/M1b M2/M2a
contrapuntal imitative imitative +
contrapuntal
imitative
accompaniment
E5
link + M1
E8
Msb/Mcs2a
contrapuntal
E9
E4 + link + E6 + S/CS1 incomplete + cadence
imitative
contrapuntal
accompaniment
In other words: E8 is related to E2 (both are extensions of preceding

statements), and E9 reviews E4 and E6. The entries surrounding the two
final episodes may thus be considered as a kind of structural recapitulation.
The basic simplicity of the rhythmic pattern together with the thematic
broken chords and octave leaps indicate a rather lively character. The 3/8
time signature hints at fairly swift whole-measure beats. Probably the best
tempo proportion between the prelude and the fugue is achieved by
relating the larger beats of the two pieces: a half-note / half-measure in the
576
WTC II/24
prelude corresponds with a dotted quarter-note / a measure in the fugue.

The appropriate articulation comprises quasi legato for the 16th-notes and
non legato for the eighth-notes. The most significant exceptionand one
that is bound to cause errors for all but the very careful performersoccurs
in the pairs of longer note values that constitute appoggiatura-resolution.
Several of them are set against eighth-notes in another voice that are not
harmonically coupled and thus to be played non legato, a challenge for
judicious fingering. Moreover, appoggiaturas often develop out of tied
notes, where they are even more easily overlooked.2
Ornaments in this fugue include the trills in CS1 and in the episode
motifs Mcs1 and Mcs1a as well as the grace-notes in the final measure. All
trills begin on the upper neighbor note, shake in 32nd-notes (i.e., twice as
fast as the faster regular note values in the piece, the 16th-notes), and end
without a change of speed in the suffixes Bach spells out. Each trill thus
comprises eight notes. The grace-notes in m. 100 represent appoggiaturas.
As their notation in brackets indicates, they derive not from the manuscript
but from a copy. Ending the fugue without these appoggiaturas gives it a
very straightforward close, while playing them results in a more complete
ending and recalls its emotional qualities. The duration of the appoggiaturas is that common in connection with dotted notes: one third of the
ternary value. This gives each appoggiatura an eighth-note and leaves a
quarter-note for its resolution.
One word about the alternative readings given in the Urtext for mm.
16 and 21: The beginning of CS1 in M: m. 16 would sound meaningful in
both versions, while closing the Ms sequence in U: m. 161 without a link to
the tied F seems structurally more logical. In m. 21, the alternative reading
is strictly logical in the terms of the episode it opens, but may be perceived
as interrupting the particularly close link between the preceding subject
entry and the episode.
The structural layout of the fugue in B minor is supported by two
groups of factors that complement one another to provide a clear pattern.
2
The appoggiatura-resolution pairs in this fugue, in the order of their appearance:

mm:
15 18 20 33 34 35 43 50 51 52 53 54
voice:
M
U
U
M
U
M
U
M
M
M
M
M
notes:
B-A F-E D-C C-B F-E B-A E-D D-C C-B E-D D-C -E
mm:
voice:
notes:
77 79 87 88 89 90 92
93 94 95 96
U
U
M
O
M
O
M
M
M
M
M
G-A F-E C-B F-E B-A E-D G-F F-E A-G G-F B-A
B minor
577
As mentioned above, the first three subject entries are accompanied exclusively by CS1, the fourth statement and all further full entries by CS2. This
change defines the boundary between the first and the second sections. The
entry in L: mm. 70-76 is singular since its beginning is not only varied but
anticipated in another voice (see M: mm. 69-711). The sequence of this
anticipation makes one believe to hear CS1 revived after a long absence.
But the remainder of the subject is accompanied by CS2, the countersubject on duty in this portion of the fugue. This exchange is reminiscent
of that which in M: mm. 26-32 opens the second section.
The fugue thus consists of three sections: Section I covers twenty-six

measures and contains a subject statement in each of the three voices. It
ends, in the upper and lower voices, at m. 271, while the middle voice
overlaps with the next subject entry beginning in m. 26. Section II is
forty-two measures long and encompasses four subject statements. It concludes at m. 691. Section III begins in reduced ensemble after the downbeat
of m. 69. It covers thirty-two measures with two complete entries. While it
structurally retraces much of the layout presented in section II, it also
features traits from section I, particularly regarding E8 and the recurrence
of CS1 at the beginning and end of the section.
The characteristics that distinguish the three sections structurally by
means of a change in the counter-subjects and the episode material also
influence the varying intensity.
Section I is characterized by a relatively high degree of polyphonic
sophistication. The imitation of the subjects initial subphrase in
CS1 evokes the momentary impression of a stretto, while the close
578
WTC II/24
relationship of the episode material in E2 and E3 allows for no real
color contrast. The tension thus rises gradually, subdued only
briefly in E1 and in the dynamically decreasing E2 with its
descending sequences.
Section II begins, overlapping with the end of section II and
growing out of the ascending sequences of E3, on a fairly elevated
level of tension. As soon as CS2 emerges and CS1 is abandoned
(which means: at the moment when a polyphonic companion is substituted by a virtuoso accompaniment pattern), the earlier intensity
gives way to extroverted playfulness. This is supported by the
episodes, which invite contrasts in color owing to their unrelated
material, and contrasts in mood, owing to their graceful suspensionresolution figures. Moreover, the tonal definition of the subject
statements enhances this extroverted appeal insofar as two of them
are in the major mode (see U: mm. 35-41 and L: mm. 44-50).
Section III brings a return to the more intense character of the first
section. Beginning in a setting of reduced voices and without any
overlap with the previous material, it is launched as a genuine new
beginning. In the further course of this section, the return of the
counter-subject posing as a homophonic accompaniment pattern
(CS2) and of the episodes in contrasting colorboth characteristics
of section IIsuggests a return to the more outgoing mood. E9
combines a diminuendo in the descending sequences of the original
E4 with a crescendo in the ascending lines of the original E6. This
inverted dynamic curve is then topped by the incomplete subject
statement that retrieves, for the last time, the initial phrase of CS1.
The fugue thus ends in a very confident mood.
Further Reading
Benary, P., J. S. Bachs Wohltemperiertes Klavier: Text, Analyse,
Wiedergabe, Aarau 2005.
Bergner, C., Studien zur Form des Wohltemperierten Klaviers von
J. S. Bach, Neuhausen-Stuttgart 1986.
Bodky, E., The Interpretation of Bachs Keyboard Works, Cambridge,
MA. 1960.
Borris, S., Das Bearbeitungsverfahren bei den 11 kleinen Prludien im
Friedemann-Bach-Buch, in Die Musikforschung 5 (1952), pp. 50-52.
Brandt Buys, H., Het Wohltemperirte Clavier van J. S. Bach, Arnhem
1955.
Buchmayer, R., Cembalo oder Pianoforte, in Bach-Jahrbuch 5 (1908),
pp. 64-93.
Constantini, F.-P., Zur Typusgeschichte von J. S. Bachs Wohltemperiertem Klavier, in Bach-Jahrbuch 55 (1969), pp. 31-45.
Czaczkes, L., Analyse des Wohltemperierten Klaviers, Wien 1956/1965.
Dadelsen, G., Zur Situation der Auffhrungspraxis Bachscher Werke,
in R. Brinkmann (ed.), Bachforschung und Bachinterpretation heute.
Wissenschaftler und Praktiker im Dialog, Kassel 1981.
Dahlhaus, C., Bemerkungen zu einigen Fugen des Wohltemperierten
Klaviers, in Bach-Jahrbuch 41 (1954), pp. 40-45.
Dannreuther, E., Die Verzierungen in den Werken von J. S. Bach,
in Bach-Jahrbuch 6 (1909), pp. 41-101.
David, J. N., Das Wohltemperierte Klavier: Der Versuch einer Synopsis,
Gttingen 1962.
Dreyfus, L., Matters of Kind: Genre and Subgenre in Bachs Welltempered Clavier, Book I, in P. Brainard u. R. Robinson (ed.),
Festschrift fr W.H. Scheide, Kassel/Chapel Hill 1993, pp. 101-119.
Drr, A., Zur Frhgeschichte des Wohltemperierten Klaviers I von J. S.
Bach, Gttingen 1984.
Drr, A., J. S. Bach Das Wohltemperierte Klavier, Kassel 1998.
Emery, W., Bachs Ornaments, London 1953.
Engels, M. W., Bachs Well-tempered Clavier: An Exploration of the 48
Preludes and Fugues, Jefferson, N.C. 2006.
579
580
Further Reading
Forkel, J. N., ber Johann Sebastian Bachs Leben, Kunst und

Kunstwerke, Berlin 2000 (Nachdruck der Ausgabe Leipzig 1802).
Goldberg., L., Johann Sebastian Bach, The Well-tempered Clavier:
Open Score, San Francisco 1988.
Gray, C., The Forty-Eight Preludes and Fugues of J. S. Bach,
London 1938.
Groocock J./Tomita Y., Fugal composition: A Guide to the Study of
Bachs 48, Westport, CT 2003.
Grss, H., Tempofragen der Bachzeit, in Bach-Studien V (1975),
pp. 73-81.
Kalisch, V. (Ed), Bachs Wohltemperiertes Klavier in Perspektiven,
Essen 2002.
Keller, H., Das Wohltemperierte Klavier von J. S. Bach. Werk und
Wiedergabe, Kassel 1965.
Kirkpatrick, R., Interpreting Bachs Well-Tempered Clavier: A
Performers Discourse of Method, New Haven/London 1984.
Kunze, St., Gattungen der Fuge in Bachs Wohltemperiertem Klavier,
in M. Geck (ed.), Festschrift fr W. Blankenburg, Gttingen 1969,
pp. 74-93.
Kurth, E., Zur Motivbildung Bachs. Ein Beitrag zur Stilpsychologie,
in Bach-Jahrbuch 14 (1917), pp. 80-136.
Ledbetter, D., Bachs Well-tempered Clavier: The 48 Preludes and
Fugues, New Haven/London 2002.
Neumann, F., Ornamentation in Baroque and Post-Baroque Music,
with Special Emphasis on J. S. Bach, Princeton 1978.
Riedel, H. H., Recognition and Re-cognition. Bach and the WellTempered Clavier I, Berkeley 1969.
Rosenthal, K. A., ber Sonatenvorformen in den Instrumentalwerken
Joh. Seb. Bachs, in Bach-Jahrbuch 23 (1926), pp. 68-89.
Schulenberg, D., The Keyboard Music of J. S. Bach, New York 2006.
Siegele, U., Zu Bachs Fugenkomposition, in Bach-Studien 9 (1986),
pp. 19-24.
Tomita, Y., J. S. Bachs Das Wohltemperierte Clavier II: A Critical
Commentary, 3 Bnde, Leeds 1993-2005.
Werker, W., Studien ber die Symmetrie im Bau der Fugen des Wohltemperierten Klaviers von J. S. Bach, Leipzig 1922.
Zenck, H. J. S. Bachs Wohltemperiertes Klavier, in W. Gerstenberg
(ed.), Numerus und Affectus. Studien zur Musikgeschichte, Kassel
1959, pp. 67-85.

Siglund Bruhn - Bach Analysis

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Siglund Bruhn - Bach Analysis

Enviado por

Direitos autorais:

Formatos disponíveis

J. S.

Considerations and Model Answers for Book I

How to Use This Book

Before you approach the catalogue of questions, decide which prelude

How to Use This Book

This book seeks to list alternative solutions and views wherever

voices in four-part polyphony: soprano, alto, tenor, bass

subject: the principal thematic component in a fugue

the downbeat of measure 5

right hand (righ-hand part), left hand (left-hand part)

Questions Regarding a Prelude

Questions Regarding a Prelude

What is happening in your harmonically determined prelude?

What is happening in your motivically determined prelude?

What is happening in your fugue- or invention-type prelude?

What is happening in your rhythmically determined prelude?

What is happening in your metrically determined prelude?

Questions Regarding a Fugue

Questions Regarding a Fugue

2.4 What happens in the episodes of this fugue?

Of course that is true. Yet, while rhythm is never actually absent, it

If we wish to investigate further, how can we distinguish between

1.2 What factors determine the overall design of a prelude?

1.3 How do you express the character of the prelude?

1.4a How do you deal with preludes

1.4b How do you deal with preludes

repeatedly succeeding the motif may be no more than an extensionand

1.4c How do you deal with preludes

1.4d How do you deal with preludes

determined by two factors: by its metric position (i.e., it may begin on a

1.4e How do you deal with preludes

2.1 What exactly is a subject?

b) What is a phrase, and what are sub-phrases?

e) What is most important in a harmonic progression?

f) Which are the features likely to increase tension within a phrase?

2.2 What is the importance of the subject in the fugue?

c) What are stretto and parallel?

2.3 What exactly are counter-subjects?

b) What phrase structure can you expect to find?

2.4 How do you find the episodes in a fugue?

a) How many episodes can you expect?

c) In what way can one episode be related to another?

2.5 Performance practice in Baroque polyphonic style

2.6 What determines the design of a fugue?

2.7 Which factors determine the development of tension?

WTC I/1 in C major Prelude

All aspects of performance in this prelude reveal themselves by means

WTC I/1 in C major Fugue

(on either D or G; the dominants point of arrival varies throughout the

Three of these subject statements are varied: The bass entry in m. 17

WTC I/2 in C minor Prelude

As in the preceding prelude, this harmonically-determined piece does

in mm. 1-27 = in mm. 28-33

This option results in a well-balanced overall structure within the portion

Another option assumes a 2:3 proportion between the outer sections on

On two occasions, the harmonic patterns that determine the message

virtuoso, toccata-style texture. This new surface pattern prepares the

WTC I/2 in C minor Fugue

The subjects rhythm shows a fairly regular combination of 16th-notes

The sequential structure is a determining factor in the subjects dynamic

There are six subject-free passages in this fugue:

only in descending but also in ascending direction. The most consistent

If conceived as a crescendo, E3 appears as a link between the

The last two-and-a-half measures appear as a coda: the final cadence