The Principle of Composition

Author(s): Kenneth Burke

Source: Poetry, Vol. 99, No. 1 (Oct., 1961), pp. 46-53
Published by: Poetry Foundation
Stable URL: http://www.jstor.org/stable/20588546
Accessed: 04/10/2010 10:03
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=poetry.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Poetry Foundation is collaborating with JSTOR to digitize, preserve and extend access to Poetry.

http://www.jstor.org

POETRY
year measures exactly the fiftyyears since that firstissueof October
I9I2. I cannot thinkofmany thingsthatwould make me happy not to
be in Italy, but thework of planning thiscelebration, for the sake of
both I9I2 and I962, is theperfectwelcome home.
HENRY

"THE PRINCIPLE

RAGO

OF COMPOSITION"

ThePhilosophy
Forourpointofdeparture
here,let'susePoe'sessay,
oj

where he gives his suspectaccount of theway inwhich he

Composition,
wrote The Raven. According to Poe (and few ifany readershave seemed

of thepoem "is
tobelievehim)no pointin thecomposition
inclined
todemon
Andhisessay
toaccdentor intuition".
was designed
referable
workproceeded,
withthe
strate
"thatthe
stepbystep,to itscompletion
of amathematical
precision
andrigidconsequence
problem".
I.A. Richards
has referred
toPoe's explanation
as "an ostentatious
of selected
meansto fullyfore
paradeof allegedly
perfect
adjustment
seen ends". (Cf. "Poetic Process and LiteraryAnalysis", inStyle andLan

guage,
edited
byThomasA. Sebeok.)Further:
"Poe,soeager-inHarry
no
Levin'sphrase-'toconvincetheworldof his self-mastery',
spares
Andheconcludes:
"However'TheRaven'may
clear."
painstomakethis
in facthave been written,we know thatmost poems are not composed
so; theauthor'smanuscripts,where firstdraftsare available, at leastshow

us that."
It'snotunlikely
that
Poe couldat least
havecomeclosertothelogical
lydeductive
he retailsthan
be thecase.
procedures
would ordinarily
no
Afterall,hewas theauthor
ofworkslikeTheCold-Bug.
And there's
a kindof
reasontodenythatsuch"stories
of ratiocination"
do require
planning
notunliketheseries
of logical
bywhich,according
deductions
toPoe,he decideduponthelength
of thepoem,itssubject-matter,
its
contrasts,
itstone,itsrefrain,
etc.,etc.
stories
of thissort,
When planning
"It isonlywith thedlnouement
constantlyinview thatwe can give a plot

airof consequence,
itsindispensable
or causation,
bymakingtheinci
andespecially
of
dents,
thetoneat allpoints,tendto thedevelopment
Authorandreader
herereverse
with
theintention."
things.
Beginning

46

KENNETH

that
would re
out thekindof situation
theauthorfigures
thesolution,
intheother
direction,
But thereader
by
proceeds
quiresucha solution.
andgradually
towards
thesolu
progressing
beginning
withtheproblem
insuchstories,
at
is involved
planning
tion.So a kindof "deductive"
toboth"thebird"and"the
AndPoe himself
(ina letter
referring
least.
bug" in the same breath) spontaneously indicates that The Raven and

inthesamebin.
wereforhimclassifiable
TheGold-Bug

There are many kinds of story,and many kinds of poem, which

pillartopost,
from
bewritten
handtomouth,from
mightconceivably
wherehe isgoinguntilhe getsthere.
notsureexactly
writer
himself
the
of ratiocination
likeTheGold
thata story
But it'salmostinconceivable
ofaHoudini.
Houdini
Here theartislikethat
thus.
Bugcouldbewritten
from
whichhe shouldescape.
didn'tletthepublicsettheconditions
himtoescape,
wouldpermit
he
having
hituponthedevicethat
Rather,
whichwouldmake
ofconfinement
outtheexactconditions
nextfigured
such a means of escape possible. In brief,fromhis ideaof thed6nouement

his ideas
of thepriorcomplications.
hededuced
my concern
witha possible
similarity
ofmotivesbetween
However,
thepoem of thebird and the storyof thebug does not requireme to be

ourpresent
ofhowhewrotethepoem.As regards
lievePoe'saccount
truth
Poewas telling
theliteral
matter
whether
itdoesn't
speculations,

or a bare-faced lie. In fact,to get thepoint I am after,let's assume that

fromthewholecloth-andthatitwas
was a fabrication
Poe'saccount
or (asHarryLevin'se?fective
below
doneeither
bywayofshowmanship
to supplya formal,
publicdenialof the
seemsto suggest)
thebeltism
For thepresent,
we'll passup
notorious
weaknesses.
author's
personal
the personalmotives thatmay have figured in the rationalizingof his

project
when,havingdecidedthattheidealtopicof hispoemmust
forth
hiscrucialfor
he brought
combine
deathandbeauty,
somehow
the
women is,unquestionably,
of a beautiful
mula: "Thedeath,then,
on
maybeourspeculations
Whatever
world."
mostpoeticaltopicinthe
tendencies
(as
necrophile
thispointwith regardto Poe's poetically
who losthis license
hewereprovedtohavebeenanundertaker
though
neednotdetainushere.Our question
for
morbidreasons)
thequestion
is:What could be said in favorof Poe's procedure in his essay on The
Raven, even if (tomake our case as clear as possible)we flatlyassume that

false?
ofhowhewrotethepoemiscompletely
hisaccount

47

BURKE

POETRY

First,let'sdisposeof themost likelyalternative:

the"historical",
or temporally
account
of a work'sdevelop
"biographical",
"genetic"
ment.Herewouldbe thehit-and-run,
catch-as-catch-can
record
of the
ofplanning
poeminitsvarious
stages
andrevision.
Theprimary
evidence
forsucha study
of poeticcomposition
isusuallytakentobe various
draftsof thegivenwork, arranged as faras possible in theorder of their

production.
Oftentheinspection
of such
material
doesgiveusnew in

sight into theauthor both as a personalityand as a craftsman.However,

we must remember that:Even ifwe had a recordof every such single
step involved in theactualwriting of a poem, of theexact order inwhich

weremade,of theauthor's
andpersonal
revisions
dreams
dur
quandaries
ingthe
ofhisborrowings
writing,
from
otherauthors
or from
situations
inlifeitself
that
(unrelated
tooccurat thetime
incidents
happened
when

thework was being produced and that the author foundways of trans
forming for the particular purposes of his poem)-even ifwe had a
mountain of such data,we should have but a fractionof the information

neededtochart
the
ifa poemisworth
fully
work'sgenesis.
Forobviously,
the troubleof a second look, thegerm of itsbeginningshad been planted

longbeforethestageof actualwriting.In thissense,thepoetbegan

"planning"
hispoem (consciously
or unconsciously)
manyyearsbefore
he gotaroundto thewriting
of it.An inspection
of successive
drafts,
theauthor's
notebooks,
habitsingeneral,
literary
etc.,helps-butinthe
lastanalysis
thepoem'suniverse
ofdiscourse
dissolves
intothemystery
of theuniverse
itself.
However,though
studies
of thissortcan'tpossibly
go deepenough,
theydo provide us with a high-class kind of gossip that is oftenworth

theeffort.
My pointisnot thatsuchpursuits
shouldbe neglected,
but

simply that theydo not replace the "principle" involved in Poe's essay.
And our problem is to seewhat can be said for thatprinciple.

We comecloser
a secondkindofderivation,
whenwe consider
the

kind thathas to do with thepoem as a finishedpublic product, an "art

object",theformal
commodity
forwhichyoupayyourgoodmoney.
Regardless
ofwherethepoetstarted,
ofhowmanyrevisions
hemade,of
whathe addedor leftout,etc.,etc.,hereisa self-consistent
symbol
system,a structurewith beginning,middle, and end, a whole with in

And thecritic's
related
ternally
parts.
job istoappreciate
this
production.
The problem
of derivation
hereprimarily
a closestep-by
involves

step analysisof the particular text,with the attempt to show how the

48

KENNETH

BURKE

various elements in thework require one another in thecourse of shap

ingandguiding
theexpectations
andexploiting
of thereader.

Quite as Iwould not sacrificethehigh-classgossipof geneticcriticism,

so Iwould be all themore loath to abandon thisrealmof formaland
appreciative criticism,ranging frommere "news" of a work (as in a

as
bookreview)tosuchconcerns
withtheprinciples
ofa literary
species
areembodiedinAristotle's
treatise
onGreektragedy.
But thisreference
to"theprinciples
ofa literary
us to
species"
brings
notes,
whichconcerns
thecruxofthese
thefollowing
paradox:
Onceyou

begin asking about theprinciples involved in theproduction of a given

work, you set the conditions for a surprisingkind of reversal.Strictly
speaking, it isnot the same kind of reversalaswe mentionedwith regard
to theway inwhich a Poe "storyof ratiocination"might be planned by
beginning at the end. But the twomethods are sufficiently
alike to be
come confused,unless there is a specificattempt to draw thedistinction
andmaintain it.Myjob here is firsttomake thedistinctionclear,and then
to show, on the basis of it,how significantPoe's essay becomes, as a

guideforcritics.
Thereisa third
kindof "derivation",
thus:

The critic,letus say,beginswith thework as "art object", or "formal

thefinished
publicproduct.
Inexamining
it,he seesthat
commodity",
a great
init.That is,regardless
manyprinciples
areimplicit
ofjudgment
of whether the author of thework explicitly asked himselfwhy he
formedthework as he did, thework embodies a seriesof decisionswhich
implyanswers to such questions. For instance,if thework is a playwith
a blood-and-thunderending, implicitin itssheernature there is, firstof
all, a principle thatamounts to saying: "Resolved: That thiskind ofwork
should be a playwith a blood-and-thunderending." Similarly, if theau

in
thoradoptscertain
forleadingto thisending,implicit
procedures
these
areprinciples
thatamounttosaying:
procedures
there
"Resolved:
That such-and-such
insuch-and-such
kindsof characters
kindsof situa
tions
aretheproper
andundergoing
transformations
pro
such-and-such
cedures for a work of thissort." Etc. In brief, insofaras a work isde
veloped in accordancewith theauthor's senseof propriety (insofaras he
constructsit inways that"feel right" tohim), thenno matter how spon
taneousand purely "intuitive" his approach to hismaterialmay be, im

setof "principles".
pliedinallhischoicesthereisa corresponding
How

49

POETRY

ever, in his capacity purely as thewriter of the originalwork, he need

neverstatethese
principles.

As a matter of fact,usually authorsdo aim to state some of theprin

ciplesbywhichtheyareguided,or thinkthemselves
That is,
guided.
to such"statements
make fragmentary
ofpolicy"as
they'll
approaches
orWordsworth's
inSidney's
aretobe found
grandly
Apologiefor
Poetrie,
or Shelley's
preface
to theLyrical
Ballads,
Defence
But usually
ofPoetry.
their
pronouncements
alongtheselinesaremuch lessthorough-going

than that,being confined to rulesof thumb (aswith notions thatthekind

of work theywant should be statuesque rather than conversational,or

or shouldavoidrhyme,
or should
rather
thanstatuesque,
conversational
or shouldavoidinversions,
neverbewithoutrhyme,
etc.).
But forourpresent
purposes,
thepointtobe stressed
is:Whetheror

not authors do formulate the principles of choice by which they are

inthechoicesthey
arenecessarily
guided,suchprinciples
implicit
make.
ofprinciples
can thework
onlyby suchan internal
consistency
Indeed,
possesstheconsistency
anddevelopment
itself
neededtogiveitintegrity
as an artisticform.

Even ifthepoetdoesformulate
suchprinciples,
however,
he doesso

not as poet but as another self,as critic.For it is thecritic'sjob to attempt

theprinciples
of composition
thathe finds(or
systematically
specifying
thinkshe finds) embodied in the given poem. And insofaras the poet

himself
makessuchpronouncements,
abouteitherhisownwork or

otherpeople's (as for instancein theprefacesof Racine), he ishere speak

ing as a critic.Often, on thisscore,he does a much betterjob of it than
the critics themselves-but the fact remains that, at these times,he is

notpoetry
butcriticism.
Andwhether
ornotthepoetwantsto
writing
concern
withsuch
himself
todemand
matters,
he iscertainly
entitled
that
thecriticdo so.He isentitled to demandthatthecriticaim at a conceptual

to translate
architectonic
whichwill somehow
contrive
thepoet'sintui
tionsintotheterms
of their
corresponding
critical
principles.
But insofar
as thecritic
proveshimself
equalto thistask,thewhole
issuecannowbe turned
For "principles"
around.
are"firsts".
As such,
were"therefromtheverystart".
In thesense
they
ofpurelylogical
pri
of composition"
ority,"principles
in a givenliterary
implicit
species
(such as a lyric like The Raven) "were thereeven before"theybecame

embodiedin theparticular
work thatexemplifies
them.
They "were
I say,inthesenseof a purelylogical
there",
in
especially
priority-but,
SO

KENNETH

BURKE

ves
century
(and itstwentieth-century
nineteenth
thehistory-ridden
to interpret
all suchpurely
invitation
hasbeena constant
tiges),there
priority.
of temporal
priority
interms
logical
ofPoe's
forshowing
theessential
rightness
We nowhavethe
material
however
badlyhegotside
ofcomposition,
with"theprinciple"
concern
of poetic
theory
todevelopa truly"principled"
in theeffort
tracked
he found
whatprinciples
asa critic,
He really
didaskhimself,
derivation.
of TheRaven).
inhisactaspoet(author
he found)implicit
(orthought
a theory
principles
(including
theaesthetic
Ineffect,
he thusformulated
whichseemedtohim theconceptual
effects)
of beautyandof lyrical
had implicitly
guidedhiminthewrit
that
of theprinciples
equivalents
ingof thepoem. So far,so good.

interms
of a
suchprocedures
himself
intoexplaining
Thenhe tricked
kindof
series,
aswith thefirst
temporal)
purely"genetic"(narrative,
him
Andhereby
heopened
inthese
notes.
we considered
critical
analysis
self to thedistrustthathis essayhas aroused since theday of itspublica

tion.
approach
ofhispeculiar
that,
because
it'spossible
Asnotedpreviously,
to suchmatters,he did comemuch closer to such away ofworking than

purposes.
mainpoint,forourpresent
notthe
But that's
domostwriters.

The main point is thathe hit upon the ideal formforan "architectonic"
critic to aim at.

by the
ofhow anyworkarose(as tested
Inother
words,regardless
of
thegenetic
process
studying
gossipavailabletouswhennarratively
the
in time),thecriticshouldaim to formulate
thepoem'semergence
in it.Thenhe shouldtestthepower
of composition
implicit
principles
theprocess.
Thus,"prophesy
byreversing
andscopeofhisformulations
how,ifhisformula
byshowing
hewouldproceed
theevent",
ingafter
fromthe
deducible"
thepoemshouldbe "logically
tionsareadequate,
hehasformulated.
principles
in
anysuchtendencies
distrusts
Ironically
enough,thepoetusually
matter.
He should
on this
hisattitude
But thepoetshouldreverse
critics.
an
ofpoetry
byperforming
worthy
provethemselves
thatcritics
demand
equally creative taskof theirown. For if thepoet can prod the criticsto

will helprevealtheessentially
of thissort,thencriticism
performances
of theorigi
theproduction
underlying
nature
of thejudgments
principled
nalworks.
Si

POETRY

Thus,instead
of stopping
with theobviousfaults
inPoe'sessay,let's
recognize
whatanadmirably
soundcritical
procedure
was struggling
for
expression
The changes
instyle
ofpresentation
there.
wouldbe surpris
ingly slight (though theywould amount to thekind of deflectionat the
center thatshows up as quite a deflectionat thecircumference).Essenti
ally, the shiftwould amount to this:
Poe need simplyhave said: "Implicit inmy composition there are

certain
towritethiscomposi
principles.
ofhow I happened
Regardless
tion, itnecessarilyembodies theseprinciples.As a critic,I have sought to
formulate the principles.Then (just as people check multiplication by

I'llcheck
or addition
division,
by subtraction)
my critical
formulations
by reversing
and instead
ofderiving
things;
thecritical
principles
from
the examination of thework, I'll tryderiving thework from theprin

ciples."

Then hemight have added: "Insofar as theprinciples I have formulat

ed do not seem to account for asmuch as Iwould have themdo, other
criticsshould trytheirown hand at the same game, to seewhether they
can formulate the principles theythink are needed, to account for the
nature of thework."
In sum: To write thepoem at all, thepoet necessarilywrites a certain
kindof poem. Insofaras thepoem iseffective,itwill necessarilyproduce
a certainkindof effect.And thepoem isnecessarilycomposed of theele
ments by which it produced the particularkind of effect"proper" to
thatparticularkind of poem.
Turn this situationaround, and the particularmethods and subject
matter of thepoem can be, as itwere, "deduced" or "derived" from the
definitionof thepoem as a kind. Thus, Poe triedto persuade his readers
thathe deduced the topic and treatmentof The Raven not just "in prin
ciple" but actually (genetically)fromhis definitionofwhat he considered
to be the ideal lyric.
However, while arguing for Poe's derivations "in principle", I still
feel thathis particular formulasmust be considerablymodified, before
we shallhave the best "principles" for "prophesying afterthe event" in
Poe's case. Possibly,we need more thanhis defimitionof what he takes
to be the ideal lyric,or any such definition.Possibly, fora complete job
of such "derivation inprinciple",we should also need a formulaforPoe's
particular personality (at least,his particular personality as a poet, re

52

KENNETH

gardlessof what he may have been as citizen and taxpayer). Such con
siderationswould also involve the belowthebeltism to which we re

inconnection
withHarryLevin'sobservation.
ferred

For instancewith regard to Poe's preferencesfordead women as a

"beautiful" topic for a poem, we might recall that, in La Vita Nuova,
Dante tellshow he dreamed of Beatrice as dead while shewas stillalive.
motives for such a fantasy,and whether or
Whatever thepsychological
not in some cases it is "necrophile", there is the purely logologicalfact
thatdeath is a speciesofperfection
(that is,"finishedness").And however
Dante and Poe may have conceived of poetry, both were
differently
as a poeticmotive.
concernedwith perfection
Many concernsof thissortwould need treatment,fora fullyrounded
discussionof Poe's essay.But though I have a small bbl. of notes on the
subject, thesewould call for another, quite differentpresentation.
KENNETH

BURKE

relevantpassage in TheMeaning oj
Along theselines,thereis a beautifully
ourWorldofToday,byJ.W.Mackail.Here theauthorpointsup thenature
Virgilfor
of theAeneidby sayingnotwhat thepoem isbutwhat itoughttobe (as thoughthe
Yet he is"putting
poemwere stilltobewritten,inaccordwith hisspecifications).
inan order"forexactlythekindof poem thattheAeneidactuallyis.For instance,
themethod:
mostlyquicklyillustrate
topick a fewdetailsthat
The work must be a nationalpoem.... Itmust establishand vindicate
ofRome with Italy.... Itmust linkup Rome and
thevital interconnection
thenew nationto theGreek civilization.... Itmustbringwell intothefore
betweenRome andCarthage.... It
groundof thepicturethehistoricconflict
fortheromantic
mustcelebratethefeatsofheroes.... Itmust findexpression
spirit,in itstwoprincipalfieldsof loveand adventure.... Itmust exalt the
new regime,etc.
But a stillbetterexample is in
There are twelvesetsof such"specifications".
thesecondparagraphofWordsworth'sPrefaceto thesecondeditionof Lyrical
Ballads:
The principalobject, then,which I proposedtomyselfin thesePoems
and situations
throughout,
as faraswas possible,in a
was to chuse incidents
selectionof languagereallyused bymen, and, at thesame time,to throw
should
wherebyordinarythings
of imagination,
over thema certaincolouring
and above
be presentedto themind in an unusualway; and, furthermore,
by tracingin them,
and situationsinteresting
all, tomake theseincidents
theprimarylawsof our nature:chieflyas
trulythoughnot ostentatiously,
far as regards themanner

inwhich we associate

53

ideas in a state of excitement.

BURKE