Spectromorphology: explainingsound-shapes

DENI S SMAL L EY
Department of Music, City University, Northampton Square, London EC1V 0HB, UK
1. INTRODUCTION to beableto separatethemfor discursivepurposes
wecannot inthesamebreath describewhat isshaped
Theart of music is no longer limited to thesounding
and the shapes themselves. The termmay be rather
models of instruments and voices. Electroacoustic
jargonistic and it is perhaps an ungainly word, but
music opens access to all sounds, a bewilderingsonic
I have not managed to invent an alternative which
array ranging from the real to the surreal and
encapsulates the interactive components so acc-
beyond. For listenersthetraditional linkswith physi-
urately. Eachcomponent of thetermbelongsto other
cal sound-making are frequently ruptured: electro-
disciplines (visual, linguistic, biological, geological),
acousticsound-shapesandqualitiesfrequentlydo not
which is tting since musical experience radiates
indicate known sources and causes. Gone are the
across disciplines. But thecombination is unique: in
familiar articulations of instruments and vocal utter-
music weoften need words which areinvented spec-
ance; goneis thestability of noteand interval; gone
ially for deningsonic phenomena.
too is the reference of beat and metre. Composers
A spectromorphological approachsetsout spectral
also haveproblems: howto cut an aesthetic path and
and morphological models and processes, and pro-
discover a stability in a wide-open sound world, how
vides a framework for understanding structural
to develop appropriatesound-making methods, how
relations and behaviours as experienced in the tem-
to select technologies and software.
poral ux of themusic.
How are we to explain and understand electro-
acoustic music?Music isnot created fromnothing. I f
a group of listeners nds a piece of electroacoustic 2.2. Therelationshiptocompositional method
music rewarding it is because there is some shared
Spectromorphology is not a compositional theory or
experiential basis both insideand behind that music.
method, but a descriptive tool based on aural per-
Weneed to beableto discuss musical experiences, to
ception. I t is intended to aid listening, and seeks to
describe the features we hear and explain how they
help explain what can be apprehended in over four
work in thecontext of themusic.
decadesof electroacoustic repertory. Howcomposers
conceivemusical content and formtheir aims, mod-
els, systems, techniques, and structural plansis not
2. DEFINITIONS AND CONTEXTS
the same as what listeners perceive in that same
2.1. Denition
music. What thecomposer has to say (in programme
notes, talks, sleeve notes) is not unimportant, and it
I have developed the concepts and terminology of
undoubtedly inuences (both helping and impeding)
spectromorphology
1
as tools for describing and
the listeners appreciation of music and musical
analysing listening experience. The two parts of the
ideas, but it isnot alwaysperceptually informativeor
term refer to the interaction between sound spectra
relevant.
(spectro-) and the ways they change and are shaped
Although spectromorphology is not a composi-
through time (-morphology). The spectro- cannot
tional theory, it can inuencecompositional methods
exist without the-morphology and vice versa: some-
since once the composer becomes conscious of con-
thinghas to beshaped, and a shapemust havesonic
ceptsandwordstodiagnoseanddescribe, thencompo-
content. Although spectral content and temporal
sitional thinking can be inuenced, as I amsure my
shapingareindissolubly linked, weneedconceptually
owncomposinghasbeen. I ntheconfusing, wide-open
sound-world, composers need criteria for selecting 1
This essay was rst published in French in 1995, and has since
appearedin I talian(Smalley1995, 1996). I t isan extensiverewrit- sound materials and understanding structural
ing of my Spectromorphology and structuring processes in The
relationships. So descriptive and conceptual tools
Languageof Electroacoustic Music (Smalley 1986). Thedevelop-
which classify and relate sounds and structures can
ment of spectromorphological thinking owes most to Pierre
Schaeffers Traite

des objets musicaux(Schaeffer 1966). bevaluablecompositional aids. Spectromorphological

OrganisedSound2(2): 10726 1997CambridgeUniversity Press. Printed in theUnited Kingdom.
108 Denis Smalley
awareness may thereforehelp thecomposingprocess of the degree of detail is a problemsince recordings
(CDs) allow one to listen repeatedly to the briefest but it will not necessarily make a better composer.
Although spectromorphological ideas can help per- passages in a work, discovering much more detail
than it is possible to hear in the course of normal ceivers to focus on the imagination and craft of the
composer, they will not help thecomposer to become musical ow. Howmuch istoo much, andhowmuch
is not enough? There is no objective method of moreimaginativeor to improvecraft.
achieving a visual spectromorphological represen-
tation, and the analyst hopefully becomes only too
2.3. About scoresandsonograms
aware of subjective decision-making and alternative
readings. This is as it should be. I nelectroacousticmusictherearethreetypesof score
which might contain perceptually relevant infor- I t is undoubtedly truethat thetask of preparinga
full, relevant, graphic spectromorphological tran- mation. The rst type is intended to be used by a
performer (in mixedworks, liveelectronic music) and scription is vexatious both because of transcription
problems and thetediumof thetask. Even thesono- may contain graphic transcriptions of acousmatic
material. The second, rarer type, is the realisation gram presents difcult, time-consuming decisions.
Considering thededication required, it is no wonder score, usually technical to a degree, intended as a
record of how a work was produced. I t represents that exhaustive analyses are few. The problems of
representation combined with thelack of consistent, content andforminsomeway, althoughthismaynot
beperceptually very relevant. Wecould also include thorough and fairly universally applicable analytical
tools have undoubtedly inhibited electroacoustic composers (private?) form-plans and similar rep-
resentationsinthiscategory. Thethirdtypeisthedif- musicsacceptanceinmoreintellectual, musicological
circles. But we must be cautious about putting too fusion score of an acousmatic work, often a free,
sketchy, graphic representation of thesounding con- much faith in written representations becausewriting
freezes theexperienceof temporal ux. I t is a device text produced primarily as a timing and memory aid
for theperson diffusingawork in concert. This third for counteracting theeeting and selectivenatureof
wayward aural attentiveness and memory during the type, andalso anytranscriptionof acousmaticmater-
ial in therst typeof score, isusually concerned with sounding ow of music. Electroacoustic musics
strengthisthat unliketraditional Westernart musicit spectromorphological information: events and tex-
tures aregiven shapes whosevertical dimension rep- cannot so easily suffer by beingreducedto anotation
system, thereby courting thedanger that music writ- resents spectral space, while the horizontal plane
showschangeover time. Thesetranscriptivescoresat ing might be regarded as a separate entity, a substi-
tute for perceptual experience. I n this regard, thevery least usually givea senseof thebroad struc-
tural outlines, and they can be aurally interesting electroacoustic music has more in common with
much orally transmitted music than with theheritage sincethetranscriber chooses to represent certain cri-
teria and not others. On theother hand this may not of Western art music.
I s a scorean obligatory aid to spectromorpholog- bea freeand considered choicesincetranscription is
often done quite quickly and spontaneously, and is ical description and analysis? Certainly for the ana-
lyst somekindof representationisnecessarybut what limited by what can beeasily represented on paper in
two dimensions. I t is a score with a particular pur- this should be is determined by the purpose of the
analysis: whether something exhaustive and detailed pose and only needs to be accurate enough for its
taskalist of timingsmight do just aswell. Thetran- is intended or whether only a particular aspect of the
music is to be examined. Anything from a diagram scriptivescore, or a morethoughtful, developed ver-
sion of it, is sometimes used as a listeningaid. (representing some of the criteria presented in this
chapter) to a graphic representation might beappro- Thesonogram, agraphicspectral analysisby com-
puter, has been regarded as a solution to the visual priate, accompanied by sound examples. For other
types of listener, however, perhaps a bundle of con- representation of electroacoustic music. I regard it as
a very useful aid rather than a solution. A sonogram ceptual tools will be enough to enable useful
discussion. is a typeof literal spectral analysis at a chosen visual
resolution: at too high a resolution detail becomes
lost in a blur; at too low a resolution there is insuf-
2.4. Ignoringtechnology
cient detail. But a sonogramis not a representation
of themusic asperceived by a human ear in asense I n spectromorphological thinking we must try to
ignore the electroacoustic and computer technology it is too objective. I ts shapes therefore have to be
interpreted and reduced to perceptual essentials. I n used in themusics making. Surrenderingthenatural
desire to uncover the mysteries of electroacoustic other words, someone has to decide what to retain
and discard fromtherepresentation, and moreparti- sound-making is a difcult but necessary and logical
sacrice. Thedesireto knowisnatural because, tradi- cularly, try and determine how much detail is perti-
nent to thealert listener. For theanalyst thisquestion tionally, all cultures have considerable knowledge
Spectromorphology 109
about howsounds aremadeas a result of continuing cutting across national boundaries and individual
styles. However, it isintended to account for typesof visual observation and listening. Once we can grasp
the relationship between the sounding body and the electroacoustic music which aremoreconcerned with
spectral qualities than actual notes, more concerned causeof thesound wefeel wehavecaptured acertain
understanding: intuitive knowledge of the human with varieties of motion and exible uctuations in
time rather than metrical time, more concerned to physical gesture involved is inextricably bound up
with our knowledgeof music as an activity. account for sounds whosesources and causes arerel-
atively mysterious or ambiguous rather than blat- Electroacoustic music composition in its acous-
maticformsisnot thesamekindof activity. A sound- antly obvious. I f a particular styleof electroacoustic
music is more traditionally note-based or metrically textureor event initsnishedguiseisrarelytheresult
of a single, quasi-instrumental, real-time, physical organised, then a spectromorphological approach
will not be very helpful. Of course notes and inter- gesture. There may be no real sounding body
involved nor any aurally identiable causal action vallic relationships are often very important in elec-
troacoustic music, though they may be covertly supposedly responsiblefor making thesound. I nfor-
mation about the synthesis method, computer pro- hidden in the fabric. I n such cases a knowledge of
tonal traditionandmeaningin(Western?) culturewill gram, treatment device, etc., is not a substitute for
also beneeded as a descriptiveaid. knowledgeof sourcecauseinteraction: theworking
Spectromorphological thinking is primarily con- gestures of theacousmatic compositional process do
cerned with music which is partly or wholly acous- not carry perceptual information equivalent to an
matic, that is, music where(in liveperformance) the intuitive knowledge of the physical gestures of
sources and causes of the sounds are invisiblea traditional sound-making. Therefore, while in
music for loudspeakers alone, or music which mixes traditional music, sound-making and the perception
live performance with an acousmatic, loudspeaker of sound are interwoven, in electroacoustic music
element.
2
However, what is and is not acousmatic is they are often not connected. Not that gesture,
not clear-cut, sinceeven music whereliveperformers sourcesandcausesareunimportant inelectroacoustic
are involved can become acousmatic when the list-
music. On the contrary they are very important, as
ener cannot connect the sounds heard with the
weshall discover.
observedphysical activitywhichsupposedlyproduces
The composer, or other listeners conversant with
them. This can happen in live electronic music per-
technology and techniques, cannot easily brush aside
formance, and isacategory of music I call liveacous-
a particular listening modewhich I call technological
matic music. The distinction between what is and is
listening. Technological listening occurs when a list-
not acousmaticbecomesblurredevenfurther inaCD
ener perceives the technology or technique behind
recordingwhereeverythingbecomes invisible.
themusicrather thanthemusicitself, perhapstosuch
Some contemporary instrumental music can also
anextent that truemusical meaningisblocked. Many
beapproachedspectromorphologicallyfor example,
methodsanddeviceseasily imposetheir own spectro-
themusic of Xenakis and of younger composers like
morphological character and cliche

s on the music.
Grisey, Saariaho, Murail, Dillon and many others
I deally the technology should be transparent, or at
concerned with spectral and textural complexity. I n
least themusic needs to becomposed in such a way
thismusicthereisoftenalossof instrumental identity
that the qualities of its invention override any tend-
astheorchestraisresynthesised into akind of spec-
ency to listen primarily in a technological manner. I t
tromorphological hyper-instrument. While we may
is difcult for thecomposer to adopt a purer spec-
sometimes be conscious of instrumental identity, we
tromorphological ear untainted by technological list-
can equally be persuaded to forget individual note-
ening when there are so many technical
gesturesastheseindividuals aresubsumed in streams
preoccupations which interfere with the creative
and collective motions. Even though this music is
stream, cloudingperceptual judgement.
represented and achieved through musical writing,
Spectromorphological thinkingisbasedoncriteria
thescoreitself is a very inadequaterepresentation of
whichcanpotentially beapprehendedby all listeners.
perceptual qualities. An aural approach which treats
I n electroacoustic music, theseparation between the
recordings of such works in themanner of an acous-
act of sound-making and perception, combined with
matic tapework is often much morefruitful.
thespecialised nature, proliferation and transienceof
A spectromorphological approach cannot deal
methods and devices, indicate that technological
adequately with electroacoustic music which is very
knowledgecannot bepart of any methodfounded on
strongly anecdotal or programmatic, that is, music
perceptual consensus.
wherea very widepaletteof sonic references may be
employedrecordings of cultural events and behav-
2.5. Whichmusics?
iour, musical quotation and pastiche, and so on. I n
I intend spectromorphological thinking to be appli- 2
For further reading refer to Vous avez dit acousmatique? (see
Smalley 1991a). cable to a wide variety of electroacoustic musics,
110 Denis Smalley
this typeof electroacoustic music, meaning is closely Extrinsicintrinsic threads of the as if variety
need not refer only to sounding experience. Non- allied to recognising the sources, identifying with
them, knowing which context they havebeen drawn sounding extrinsic links are also possible, whether
based on human physical movement (seethesection from, and reinterpreting their meaning in their new
musical context. Such music is thereforetranscontex- on gesture) or environmental experience. For
example, spectromorphology is concerned with tual or intertextual. Much acousmatic music straysin
and out of transcontextual suggestionsthe sounds motion and growth processes, which are not exclu-
sively or even primarily sonic phenomena: sonic of nature, theelements and environment areparticu-
larly common. More often than not a mixture of motion can suggest real or imagined motions of
shapes in free space. Spatial experience itself can spectromorphological and transcontextual insights is
required. Spectromorphological qualities can often involve sounds or not. Energy, which is inherent in
spectral motion, is part of both sounding and non- help qualify thepower of a transcontextual message.
sounding experience, linked not only to motion in
general but to human gesture as well the energetic
2.6. Intrinsicextrinsicthreads
impact of an implement hittinga soundingbody, for
example, has spectromorphological consequences.
Spectromorphology concentrates on intrinsic fea-
I have invented the term source bonding to
tures. That is, it is an aid to describing sound events
represent the intrinsic-to-extrinsic link, from inside
and their relationships as they exist within a piece
the work to the sounding world outside. I dene
of music. However, a pieceof music is not a closed,
sourcebondingas:
autonomous artefact: it does not refer only to itself
but reliesonrelatingto arangeof experiencesoutside
the natural tendency to relate sounds to supposed
thecontext of thework. Musicisacultural construct,
sources and causes, and to relate sounds to each
and an extrinsic foundation in cultureis necessary so
other because they appear to have shared or
that theintrinsiccanhavemeaning. Theintrinsicand
associated origins.
extrinsic areinteractive.
3
Thewordbonding seemsparticularlyappropriate
I n transcontextual music, intrinsic qualities and
sinceit evokes a binding, inescapableengagement or
relations as created by the composer determine the
kinship between listener and musical context. The
impact of extrinsic messages. But there is a sort of
bondings involve all types of sounding matter and
transcontextuality throughout electroacoustic music,
sound-making, whether in nature or in culture,
though not of the intentionally encoded varieties
whether they arise as a result of human agency or
which in authentic trans-contexts are so strongly
not. Sourcebondings may beactual or imaginedin
reliant on sourceidentication. Thewide-open sonic
other words they can be constructs created by the
world of electroacoustic music encourages imagina-
listener; different listeners may sharebondings when
tive and imagined extrinsic connections because of
they listen to the same music, but they may equally
thevariety and ambiguity of its materials, becauseof
havedifferent, individual, personalised bondings; the
its relianceon themotion of colourful spectral ener-
bondingsmaynever havebeenenvisagedbythecom-
gies, its emphasis on the acousmatic, and not least
poser and can occur in what might beconsidered the
through its exploration of spatial perspective. There
most abstract of works; wide-ranging bondings are
is quite a difference in identication level between a
inevitableinmusicswhich arenot primarily weighted
statement whichsaysof atexture, I t isstonesfalling,
towards xed pitches and intervals. Bonding play is
a second which says, I t sounds like stones falling,
an inherent perceptual activity.
and a third which says, I t sounds as if its behaving
likefalling stones. All threestatements areextrinsic
connections but in increasing stages of uncertainty
2.7. Instrumental andvocal sourcebonding
and remotenessfromreality. I f alistener, elaborating
Onemight think that in moreabstract instrumental
oneither statementstwo or three, commentsonqual-
music, source bondings do not exist, but they are
ities and features of the texture as heard within the
there in force, revealed through gesture and other
musical context, then attention turns away fromthe
physical activity involved in sound-making. The
primarily extrinsic towards special intrinsic features
bonding of instrumental activity to human gestureis
and thereforemoves moredeeply into theparticular
somewhat ignored not only because it is invariably
musical experience. I t is thus that this listener starts
expected in music, but also possibly because much
to engagein spectromorphology.
music study hastended to concentrateon music writ-
ing(notation), theoryandanalysis, whichtendto dis-
3
The intrinsic and extrinsic aspects of music are discussed fairly
fully in Music and Discourse: Towards a Semiology of Music
tance a work from the gestural activity of its
(Nattiez 1990: 10229). (The French edition, which is not the
performance. Vocal presence, whether revealed
sameastheEnglish edition, waspublished by Christian Bourgois
in 1987, under thetitleMusicologiege

ne

raleet se

miologie.) throughstylisedsingingor direct utterance, hasdirect

Spectromorphology 111
human, physical, and therefore psychological links. encouraging deeper exploration, they also causeper-
ceptual distortions. My experience of teaching com- I n electroacoustic music the instrumental and vocal
are included, but unlike traditional music where posers has often revealed to methat such distortions
arefrequent. instrumental and vocal presence are assumed and
known in advance, in acousmatic electroacoustic As far as the spectromorphological approach to
electroacoustic music is concerned, reduced listening musictheyarebothto beencountered(unexpectedly)
and inferred. Electroacoustic music, then, subsumes mechanisms liebehind thedevelopment of concepts.
They areanecessary aid to any full analysis, particu- instrumental and vocal experience: instrumental and
vocal resourcesareasubset of thewide-open sound larly at the lowest levels of structure, and they are
necessary to any investigation which seeks to reveal world which is electroacoustic musics territory.
those intrinsic features which invest the extrinsic I dentifyingintrinsicextrinsic threads is onething.
threads with their psychological power. The CD I nterpretingtheir meanings, expressivity, andpsycho-
player now, for therst time, through itsexiblerep- logical signicance is a more distant but ultimate
etition processes, offers listeners other than thecom- goal, a study of which is beyond the scope of this
poser of the work the opportunity for investigating essay. Someof theconceptsand languageor spectro-
what were previously but eeting moments. Retain- morphology have formalist tendencies, but ideally,
ingarealisticperceptual focusis, nally, aprecarious intrinsic spectromorphological description, although
balancingact. not directly interpretative, should becapableof help-
ing a listener to pinpoint those musical qualities
which arecarriers of meaning.
3. GESTURE AND ITS SURROGATES
Until the electroacoustic medium arrived, all music
was created either through forms of vocal utterance
5
2.8. Thecomposersear
or through instrumental gesture. Sound-making ges-
Attention has already been drawn to technological tureisconcernedwithhuman, physical activitywhich
listening as a listening mode where the composers has spectromorphological consequences: a chain of
perceptions can bedifferent fromthoseof other list- activitylinksacauseto asource. A humanagent pro-
eners. But thecomposer also has to combat another ducesspectromorphologiesviathemotionof gesture,
specialised listening modethat of reduced listening using the sense of touch or an implement to apply
(e

coute re

duite).
4
Reduced listening comes about energy to a sounding body. A gesture is therefore an
through concentrated, repeated listening to a sound energymotion trajectory which excites the sounding
event, a common activity in theelectroacoustic com- body, creating spectromorphological life. From the
posingprocess. I t is an investigativeprocess whereby viewpoint of both agent and watching listener, the
musical gesture-process is tactile and visual as well detailed spectromorphological attributes and
as aural. Moreover, it is proprioceptive: that is, it is relationships are uncovered. This requires that the
concernedwiththetensionandrelaxationof muscles, distractions of sourcebonding and intrinsicextrin-
with effort and resistance. I n this way sound-making sic threads beblocked out in order to concentrateon
is linked to more comprehensive sensorimotor and reningspectromorphological detail and sound qual-
psychological experience. ity. Reduced listening is therefore an abstract, rela-
Weshould not think of thegestureprocessonly in tively objective process, a microscopic, intrinsic
the one direction of causesourcespectromorphol- listening.
ogy, but also inreversespectromorphologysource Outsidethecreativeprocess outlined above, many
cause. When wehear spectromorphologies wedetect composers regard reduced listening as an ultimate
the humanity behind them by deducing gestural
mode of perceptual contemplation. But it is as dan-
activity, referring back through gesture to proprio-
gerous as it is useful for two reasons. Firstly, once
ceptive and psychological experience in general.
one has discovered an aural interest in the more
Everyoneuses this spectromorphological referral pro-
detailed spectromorphological features, it becomes
cess when listening to recordings of instrumental
very difcult to restore the extrinsic threads to their
music. Not only do we listen to the music, but we
rightful place. Secondly, microscopicperceptual scan-
also decode the human activity behind the spectro-
ning tends to highlight less pertinent, low-level,
morphologiesthrough which weautomatically gain a
intrinsic detail such that the composerlistener can
wealth of psycho-physical information.
easily focus too much on background at theexpense
of foreground. Therefore, while the focal changes
5
For reasonsof spaceit is not possibleto enter into a comprehen-
permitted by repetition have the advantage of
sive discussion of the vocal sound as a spectromorphological
model. Theideaof gesturecan, of course, beapplied to thevoice,
wheretheenergy aswell asthesound isprojected frominsidethe
4
Reduced listeningis a Schaefferian concept. SeeSchaeffer (1966)
and Chion (1983) for a full discussion. human body.
112 Denis Smalley
The listeners experience of listening to instru- Third-order surrogacyiswhereagestureisinferred
or imagined in themusic. Thenatureof thespectro- ments is a cultural conditioning process based on
morphology makes us unsure about the reality of yearsof (unconscious) audiovisual training. A knowl-
either thesourceor thecause, or both. Wemay not edge of sounding gesture is therefore culturally very
besureabout howthesound was madeto behaveas strongly imbedded. This cannot be ignored and
it does, what thesoundingmaterial might be, or per-
denied when we come to electroacoustic music. I t is
haps about the energymotion trajectory involved.
particularly important for acousmatic music where
For example, a resonant spectromorphology could
the sources and causes of sound-making become
sound as if excited by gestural impact of some kind
remoteor detachedfromknown, directly experienced
(inferred cause) even though wedo not knowexactly
physical gesture and sounding sources. The process
what thesourcemight bebecauseitssound-quality is
of increasing remoteness I refer to as gestural
unfamiliar, or because the resonance behaves in an
surrogacy.
6
unexpected way (uncertainunknown source).
Original, primal gesture, on which sounding ges-
Remote surrogacy is concerned with gestural
tureis based, occurs outsidemusic in all propriocep-
vestiges. Source and cause become unknown and
tive perception and its allied psychology. First-order
unknowable as any human action behind the sound
surrogacyprojects theprimal level into sound, and is
disappears. The listener may instead be concerned
concerned with sonic object use in work and play
with non-sounding extrinsic links, always, of course,
prior to any instrumentalisation or incorporation
based on perceived spectromorphological attributes.
into a musical activity or structure. I t is here that
But some vestiges of gesture might still remain. To
musical potential begins to be recognised and
nd them we must refer to tensile, proprioceptive
explored. Traditionally in art music this rst level
properties, to those characteristics of effort and
does not become music in itself: it develops into
resistanceperceivedinthetrajectoryof gesture. Thus,
instrumental, second-order surrogacy. But for elec-
remote surrogacy, while distanced from the basic,
troacoustic music it becomes an important level in
musical rst order, can yet remain linked to the
its own right since many unique sound-gestures are
psychology of primal gesture. But in order for such a
transplanted directly into music from this level, for
gestureto be felt, therehas to besufcient directed,
example gestural play with materials like wood or
propagating or reinjected energy in the
metal. First-order surrogacy includes recordings of
spectromorphology.
sound-making not intended for musical use. On the
Acousmatic music, therefore, can stay close to
other hand, rst-order surrogacy may involve more
traditional, gestural causesourcerelations, but at its
developed gestural play purposely used as compo-
most adventurousextends into third-order ambiguity
sitional material, a sort-of personalised, nascent
and beyond to a music which, although remotefrom
instrument which never achieves, or can never
traditional sound-making activity, can nevertheless
achievefull cultural, instrumental status. But wecan
maintain a humanity. Theaccess to a morebasic but
only award such sounds rst-order status if we can
veryrichrst-order surrogacyhasonlybecomeviable
recognise source (the type of material) and type of
becauseof recordingtechnology. Theimagination of
gestural cause. I f in the compositional process the
thethird and remoteorders is oneof thegreat offer-
source is transformed and either gesture or cause
ingsof compositionsalliancewith technology. I ven-
becomesdubious, thenthird-order or perhapsremote
tureto suggest that an electroacoustic music which is
surrogacy will beinvoked.
conned to thesecond order does not really explore
Second-order surrogacy is traditional instrumental
the potential of the medium, while a music which
gesture, a stage removed fromthe rst order, where
doesnot takesomeaccount of thecultural imbedding
recognisable performance skill has been used to
of gesture will appear to most listeners a very cold,
develop an extensive registral articulatory play. An
difcult, even sterilemusic.
acousmatic music which, for example, uses only
recordings of identiable instruments remains in the
4. SPECTROMORPHOLOGICAL
second order. Much music which uses simulation of
EXPECTATION
instrumental sounds can also be regarded as second
order since, although theinstrument may not bereal, The note is the basic gesture-unit of instrumental
music. Every note-gesture, however short, hasaspec- it is perceived as theequivalent of thereal. Commer-
tral historythe energymotion trajectory of its cial synthesizer usage is of this type when werecog-
spectromorphology. Every note must start in some nise both the gesture involved and the instrumental
way; somemay besustained or prolonged for a time
sourcesimulated.
and some may not; every note stops. These three
linked temporal phases I refer to as onset, continuant 6
Thenumber of levels has been expanded beyond thosediscussed
in Smalley (1992). andtermination. Theyarenot distinctly separable: we
Spectromorphology 113
cannot tell the very moment when an onset passes (3) By increasing the spectral energy towards
termination, leading towards, and creating the into a continuant phase, nor when a continuant
passes into the terminal phase. Nor do all three expectancy of, a new note-gesture. The continu-
ant phase is lengthened and termination is avo- phases haveto bepresent in the note-gesture. There
arethreeimportant spectromorphological archetypes ided, or rather theonset of thenew note-gesture
is also thetermination of theprevious note. foundininstrumental usage. I nall threemodelsspec-
tral richness is assumed to be congruent with the
What the archetypes and their variants demon-
dynamic shape of the morphologythe louder, the
strate is that thenote trains us in spectromorpholog-
more spectral energy, the brighter andor richer the
ical expectation. We have a very wide experience of
sound. Thesearchetypes are:
the circumstances in which spectral changes occur,
(1) Theattack alone. This is a momentary energetic
not just in singlenote-gestures but in thearticulation
impulse. Two temporal phases are merged into
of chains of note-gestures within the larger gestures
onethereisasuddenonset whichisalso theter-
of phrase-motion. Our acquired knowledge of the
mination. Wedo not havesufcient timeto hear
contextsof spectral changeprovidesan almost natu-
any appreciable changein spectral energy as the
ral reference-basenot only for developingthewider,
sound moves rapidly to its termination. Thereis
more imaginative spectromorphological repertory
no continuant phase. Awarenessisfocusedonthe
into the third-order surrogacy of electroacoustic
attack-energy. A dry percussiveattack or a stac-
music, but for decoding patterns of expectation in
cato sound (without resonance) areexamples.
musical form. We predict or try to predict the
(2) Theattack-decay. Theattack isextendedbyares-
expected tendenciesof spectral change. Electroacous-
onance. The onset and terminatory phases are
tic music, even whendeprived of known instrumental
present, andtheremaybeahint of thecontinuant
spectromorphologies and tonal harmonic language,
if wefeel that thesound is being prolonged at a
still reliesonculturallyacquiredexpectationpatterns.
consistent level beforeit decays. I n thisarchetype
an initial gesture is enough to set a spectromor-
phology in motion, after which thereis no gestu-
5. GESTURE AND TEXTURE AS FORMING
ral intervention as the sound continues towards
PRINCIPLES
termination. A string pizzicato or a bell are
The basic gesture of traditional instrumental music
examples.
produces thenote. I n tonal music notes forma con-
(3) The graduated continuant. I n this archetype all
sistent low-level unit, andaregroupedintohigher lev-
threephases arepresent. Theonset starts gradu-
elled gestural contours, into phraseological styles,
ally asif faded in, and thenoteterminatesgradu-
which traditionally have been based on breath-
ally as if faded out. I n between, the note is
groups. Singers and wind-players, after all, have to
sustained for a time. Examples are sustained
breathe. I n electroacoustic music thescaleof gestural
stringor windsounds, wherecontinuingenergetic
impetus is also variable, from the smallest attack-
input (breath or gestural) is needed to sustain or
morphology to the broad sweep of a much longer
prolongthesound. This is themost intervention-
gesture, continuous in its motion and exible in its
ist of thearchetypes becauseweareawareof the
pacing. Thenotion of gestureas a forming principle
continuing imposition of gesture. I t is therefore
is concerned with propelling time forwards, with
themost open to thedevelopment of variants.
movingaway fromonegoal towards thenext goal in
Wecan say that instrumental music is madeup of
the structurethe energy of motion expressed
strings of variants of the above archetypes, and (in
throughspectral andmorphological change. Gestural
ensemblemusic) superpositionsandmergingsaswell.
music, then, is governed by a sense of forward
Variants are created by manipulating the durations
motion, of linearity, of narrativity. The energy
and spectral energy of thethreephases. For example,
motion trajectory of gestureis thereforenot only the
the graduated continuant archetype (which is rather
history of an individual event, but can also be an
idealised) could bevaried in a number of ways:
approach to thepsychology of time.
I f gestures areweak, if they becometoo stretched (1) Bycompressingtheonset so that it isswelled(less
out in time, or if they become too slowly evolving, linear). More energy is thereby injected into the
we lose the human physicality. We seem to cross a onset, giving a more pressured, pushing effect.
blurred border between events on a human scaleand Wend thisonset variant quiteoften in Baroque
events on a more worldly, environmental scale. At performancepractice.
the same time there is a change of listening focus (2) By lengthening the continuant phase, drawing
the slower the directed, gestural impetus, the more attention away fromtheonset towardsinterest in
theear seeks to concentrateon inner details (insofar thesounds continuity, which could bestable, or
moreor less varied, prior to termination. as they exist). A music which is primarily textural,
114 Denis Smalley
then, concentrates on internal activity at theexpense precise enough that a given class of sounds (which
have been verbally described) will invariably and of forward impetus.
But most musics are texturegesture mixtures, unambiguously correspond to a given symbol.
There is no permanent type of hierarchical either in that focus shifts between them, or because
they exist in somekind of collaborativeequilibrium. organisation for all electroacoustic music, or even
within a single work.
9
Undoubtedly there are struc- Whereoneor theother dominates in a work or part
of a work, we can refer to the context as gesture- tural levels, but theydo not needto remainconsistent
innumber throughout awork, and asinglelevel does carried or texture-carried. I ndividual gestures can
havetexturedinteriors, inwhichcasegestural motion not need to run permanently through thewholespan
of a work. For example, one might detect three or frames thetextureweareconscious of both gesture
and texture, although the gestural contour domi- four levels in onepart of a work and fewer or more
inanother part; onesectionof awork might comprise nates, an example of gesture-framing. On the other
hand, texture-carried structures are not always a neat hierarchy of small, unit-groupings, while
another section might be a much larger, indivisible, environments with democratic interiors where every
(micro-) event is equal and individuals aresubsumed higher-level whole.
I t is fair to say that much electroacoustic music in collectiveactivity. Gestures can stand out in fore-
ground relief fromthetexture. This is an exampleof does not offer sufcient hierarchical variety. This
occurs where the types of sounds and the structural texture-settingtexture provides a basic framework
within which individual gestures act. continuity direct oneto listen continuously in a glo-
bal, high-level mode. With textured structures it is
very easy for thecomposer, listening too hard to the
textural material, to be deceived into thinking that 6. STRUCTURAL LEVELS
there is lower-level interest within the texture when
I t is wrong to seek in electroacoustic music thesame
thereis not. And music which deals mainly in broad,
kinds of structural hierarchies as tonal music. I n
high-level, gestural sweepswith littleinternal interest,
tonal music thenoteis regarded as thelowest struc-
may not offer much to thelistener on repeated hear-
tural level, and all tonal music is made up of note-
ings. A rewarding balanceof perceptual interest at a
groupingsof increasingdimensionsasonemovesout-
variety of structural levels is unfortunately morerare
wards through the formfrom note to motive to
than it should be.
phrase, and so on. I n addition, metrical structure
I n my spectromorphological approach, the con-
gives the lowest-level note a pulse which denes the
ceptsof gestureand texture, motion and growth pro-
minimumpossibledensity of movement.
cesses, behaviour, structural functions, spectral space
The discussion of gesture and texture as forming
and density, and space and spatiomorphology may
principlesimplied that gesturesand textures could be
beapplied to smaller or larger time-spanswhich may
small scale or larger scale. Electroacoustic gestures
beat lower or higher levels of structure. Finding the
and textures cannot be reduced either to note or
right levels or temporal dimensions to apply the
pulse;
7
themusic is not necessarily composed of dis-
attributes of these concepts must remain the per-
creteelements; nor canwendthat (consistent) meas-
ceivers decision.
ure of minimum movement density. Therefore it
cannot beconveniently segmented, and indeed often
7. STRUCTURAL FUNCTIONS
resists segmentation. At one moment in a work one
may befollowingdiscrete, short units, andat another
Structural functions are concerned with expectation.
a large-scale structure whose continuity and coher-
Like other musics, electroacoustic music has its
ence refuse to be dissected and demand to be con-
expectation patterns, and I have already suggested
sidered more as a whole than as the sumof minute
that these are based on our wide cultural acquaint-
parts. A piececould besuchthat theideaof anything
ance with the perceived spectral changes of a wide
called a unit lacks relevance. There can be no
varietyof sounds. Duringlisteningweattempt to pre-
. . . explicit criteria for segmentation and denomi-
dict thedirectionality implied in spectral change. We
nation of objects. . . as called for by Nattiez
might ask ourselves, for example, where a gesture
(1990: 80),
8
let alone the requirements expressed in
might beleading, whether a textureis going to con-
the continuation of this sentence, . . . and notation
tinue behaving in the same way, whether change is
9
Thecriteria for hierarchical organisation set out by Lerdahl may
7
I n much contemporary instrumental music, notation can be a
very misleading guide to the lowest level of structure, when the apply to tonal music, but they do not suit electroacoustic music.
That, of course, does not mean that electroacoustic music is not written noteisnot heard asadiscreteunit but aspart of acollec-
tivegestureor texture. Notation may show us howgestures and hierarchical (see Lerdahl 1987: 1378; the French version, Les
hie

rarchies de timbres, appeared in Le timbre, me

taphore pour textures aremadeup, but a readingof individual notes will not
tell us what should beheard in them. la composition, J .-B. Barrie
`
re, ed., Christian BourgoisI RCAM,
1991).
8
Seenote3for thereference.
Spectromorphology 115
likelyor not, whether changeislikelyto beconcerned function until weknow what comes afterwards.)
Since we are always busy concentrating on the with gradual mergingor sudden interruption, and so
on. The ideas of onset (how something starts), con- present, our attentions may be diverted before
any nal attribution can becompletedcontexts tinuant (how it continues) and termination (how it
ends) can be expanded into a list of terms, some of slip away. And which levels of structure are we
following anyway? How many levels can our themtechnical, some more metaphorical, which can
be used to interpret the function-signicance of an attention and memory spans takein? Arewefol-
lowing each impending moment or broader event or context. These functions can be applied at
both higher and lower levels of musical structure, outlines?
(3) Function attribution may bedoubleor ambiguous. referring, for example, to anote, an object, agesture,
a texture, or a type of motion or growth process, A context may havedifferent, simultaneousfunc-
tions. This is particularly so when events are dependingon our focus of attention.
Theonset terms reect varying degrees of abrupt- overlapped or motion is continuous. For
example, a contour which seems to resolve a ness. What they have in common is that they are
moving out fromor away froma starting-point, cre- motion could also formpart of the anacrusis to
a following peakin this case the termination ating expectant tensions. Thecontinuant terms vary.
Some look forward, expressing betweenness (tran- function is also an onset function on the same
level. sition, passage); others arelinked backwards to the
onset function (prolongation, maintenance), while (4) Thereisnoclear temporal border betweenthethree
function-types. For example, the point at which statement signals a moredenitive, almost indepen-
dent status. The terminations vary in their feelings wedecidethat anevent hasfully emergedandthe
point at which we decide that it has entered a of completion. Disappearance isaweak termination
without much purpose, while resolution and transitional phasemay not bea point but a pro-
cess of evolvingrealisation. release havestrongrelaxant functions. Arrival and
plane express structural goals achieved.
Theattribution of a function to a particular event
or context is not a simplecognitiveprocess: 8. MOTION AND GROWTH PROCESSES
The metaphors of motion and growth are appropri- (1) Function attribution is not normally a conscious
thought-process. We are not continually asking ateways of consideringa time-based art likeelectro-
acoustic music. Traditional concepts of rhythm are ourselves a reasoned series of function-questions
resulting in function-decisions. Rather, function inadequate to describe the often dramatic contours
of electroacoustic gestureand theinternal motion of attribution is part of theintuitiveexpectations of
psychological time. texture which are expressed through a great variety
of spectromorphologies. Quite often listeners are (2) Function attribution is a continuing, incomplete
process, subject to revision. We make interim reminded of motion and growth processes outside
music and the terms selected are intended to evoke function attributions, and may changeour minds
in the continuing course of a context and after thesekinds of connections. Sincemotion and growth
havespectral contours, they areset in spectral space. it has passed by. (We cannot nally attribute a
116 Denis Smalley
Figure1. Motion and growth processes.
Thereforetheir occupancy of spectral spaceand their Generally in music, centric motion is expressed by
spectral density will be important additional quali- spectromorphological recycling, givingan impression
ers (see gures 6 and 7). I n addition, motion and of motion related to a central point. This can be
growthcanbeenhancedbyactual spatial distribution achieved through spectromorphological variation
(see gures 8 and 9). The basic reference types are alone, but isfrequentlyaidedbyspatial motion. Con-
set out in gure1. Theunidirectional, reciprocal and tinuing recycling, like other forms of repetition, can
cycliccentric groups should be regarded more as give an impression of structural stasis, but centric
motions, whilethebimultidirectional group is more motions can also be strongly directional vortical
concerned with growth processes. and spiral motions havethis possibility, for example.
Motion and growth have directional tendencies Centric motions can also beassociated with growth.
which lead us to expect possibleoutcomes, and they
For example, I can think of rotating motions which
arehelpful guides in attributing structural functions.
gather textural materials to them as they expand
Unidirectional motion provides a simple example. I f
spectrallya combination of rotation and exogenous
we encounter a slow, ascending contour, we could
or endogenous growth. The spin, spiral and vortex
expect avariety of outcomesbut not any outcome. I t
arerotational variations. Centrifugal (ung out) and
couldascendandfadeasit goesout of sight; it could
pericentral (merely moving around a centre) arealso
increase in richness leading to an impact point; it
a related group.
could be joined and absorbed by other events; it
Bimultidirectional motions create expectations,
could change direction, turning into a parabola; it
and most havea senseof directed motion. They can
could reach a stable ceiling. Whatever it eventually
be regarded as having both gestural and textural
does may surpriseus (if therearesudden changes) or
tendencies, and could be large structures in them-
it may do what we expect particularly if its rate of
selves. Agglomeration(accumulatinginto amass) and
changegives us clues to its future. This hypothetical
dissipation (dispersing or disintegrating) are textural
description could imply a termination function (dis-
processes. Dilation (becoming wider or larger) and
appearingupwards), an anacrusis (increasingin rich-
contraction (becoming smaller) are concerned with
ness leading to another event), and so on. Finally,
changing dimensions and could be regarded as a
this example emphasises the gestural nature of its
different aspect of agglomerationdissipation.
contour (no internal, textural interest has been
Divergence and convergence are strongly directional
referred to) even if it is stretched out in time.
and could begestures or texturegrowths, or a simul-
I n reciprocal motion, movement in one direction
taneous linear descentascent. Exogeny (growth by
is balanced by a return movement. Oscillation and
addingto theexterior) could beallied to dilation and
undulation, which are contour variations, could
agglomeration, whileendogeny(growingfrominside)
apply to internal, textural motions, as well as being
implies somekind of framewhich becomes lled, or
descriptions of external contour. Parabolas areoften
texturewhich becomes thickened.
moregestural, aclassof curved trajectories. They are
I n gure 2 I have grouped together seven charac-
common in electroacoustic music, probably because
teristic motions in order to sketch out their
of thedramatic possibilities of varying theduration,
implications.
velocity and spectral energy of the outward and
return journeys. A parabolic trajectory can move (1) Motion rootedness. Some are more likely to be
earthbound (push, drag) while others are not continuously through a variety of spatial locations.
Spectromorphology 117
Figure2. Seven characteristic motions.
rootedto asolid plane. Many spectromorpholog- loosebut collectivemotion of micro- or small object
elements whoseactivity and changes in density need ies areinherently non-rooted becausethereis no
bassanchor (fundamental note) to securethetex- to beconsidered as a whole, as if moving in a ock.
One can imagine ocking motion passing through a ture. Thusanalogieswithight, drift and oating
can becommon. Motion towardsaroot could be variety of multidirectional growth processes. The
interior of a stream could be ocked, and a ock impliedinaspectral descent towardstermination,
but a root may not be achieved if the motion could be part of a streamed texture. Convolution
(coilingor twisting) and turbulence(irregulation uc- fades in theair.
(2) Motion launching. The launching of motion var- tuation, possibly stormy) involve confused spectro-
morphological entwining, but nevertheless tend to ies. Some may be considered as self-contained
eventswithgesture-based, pressuredonsets(drag, concur in their chaos.
Texture motion may vary in internal consistency. throw, ing) whileothers could bethought of as
emerging as if they had always existed (ow, Continuous motion is sustained while discontinuous
motion may be more or less fragmented. The conti- oat, drift).
(3) Contour energy and inectionthe direction and nuitydiscontinuity continuum runs from sustained
motion at one extreme to iterative motion at the energy of motion through spectral space. Some
areinherently slowor evolving(ow, drift, oat), other. I f iterative repetitions become too widely
spaced then separateobjectswill beheard. Thistend- needingtimeto establish themselves, whileothers
imply rapid energymotion trajectories. ency is possible with some of the multidirectional
growth processes if the internal texture becomes (4) Internal texturingcontinuity and discontinuity.
Most of these motions imply a certain internal sparser during fragmentation in the growth process.
Granularity occupies an ambiguous mid-point since textural consistency in order that the type of
motion may remain coherent (seebelow). it could be considered either (roughly) sustained or
iterativedependingon howclosely packed thegrains
are. Both continuity and discontinuity can movein a
9. TEXTURE MOTION
moreor less periodicaperiodicerratic manner, with
internal uctuations in tempi. Continuousdiscon- Most of thebimultidirectional motions imply inter-
nal textural change, but theother motion-groupscan tinuous texturemotion may need to beconsidered as
a totality, or may follow grouping patterns if con- also be textured. Moreover, in terms of occupancy
of spectral spacethey could vary in dimensions, and tours, uctuations or discontinuities are subject to
repetitions, cycles or pauses which imply higher-level consist of more than one layer. As far as structural
level is concerned, it is themotion as a wholewhich groupings.
is the most signicant perceptual unit. Texture
motion does not necessarily imply that it is possible
10. BEHAVIOUR
to segment the texture into lowest-level units,
although it might bepossibleto separateout oneor The metaphor of behaviour is used to elaborate
relationships among the varied spectromorphologies morespectromorphological types.
Figure 3 sets out several qualiers of texture acting within a musical context. I believe that list-
eners can intuitively diagnose behavioural relation- motion. Theleft column lists four ways in which the
internal textural components may collaborate in ships (or a lack of them) in electroacoustic music
contexts and that this diagnosis affects the listeners motion. That implies a type of behaviour (see
below). Streamingrefers to a combination of moving interpretation of and reactions to the music. I n this
respect, behaviour is archetypal. We may be aware layers, and implies some way of differentiating
between the layers, either through gaps in spectral of behaviour in tonal instrumental and vocal music,
although it is fairly circumscribed because of the space or because each layer does not have the same
spectromorphological content. Flockingdescribes the nature of the harmonic system which brings with it
118 Denis Smalley
Figure3. Texturemotion.
constraintsof polyphonyandcounterpoint. I nacous- rates, and the type of motion are strong determi-
nants. Causality, whereoneevent seems to causethe maticmusictheinvisiblefreedomof spectromorphol-
ogical content and motion creates a much wider and onset of a successor, or alter a concurrent event in
some way, is an important feature of acousmatic more variable pool of extrinsic, behavioural refer-
ences. We hardly need reminding that in an acous- behaviour. The arrow leading from the voluntary
pressured continuum shows that causality can be matic music behavioural relationships arecarried by
spectromorphology alone, and that in mixed work weaker or stronger, depending on the inuence or
impact one spectromorphology has on another. On the perceived behavioural relationships between the
visible, gesture-bearing performer and thesurround- the whole, a strongly causal music relies on gesture,
and impact-coordination points which will be ing acousmatic context will be crucial to the works
understanding. regarded as goals in thestructure.
The upper part of gure 4 is intended as a guide As with other concepts, behaviour may beapplied
at a variety of levels, for example, discrete events, to interpreting behavioural relationships and is
dependent on two oppositional pairs of concepts low-level texturemotion, or themuch higher level of
relations between groups of textures or growth pro- dominancesubordination and conictcoexistence
which form the basis for a group of relationship cesses. Behaviour (gure4) has two interactive, tem-
poral dimensions (shown linked by the double- modes listed in thecentre.
headed arrow), onevertical theother horizontal. The
vertical dimension is concerned with motion coordi-
11. SPECTRA
nation (concurrence or simultaneity), while the hori-
zontal dimension is concerned with motion passage
I must nowturn in moredetail to thespectro- aspect
(passingbetween successivecontexts).
of spectromorphology. I adopt the more general
Theloosetight continuumrepresents thedegreeof
terms spectra or spectral space to represent the
coordination freedom. Whereas synchronicity has
wide variety of sound-qualities, timbres and pitches
been the rule in tonal art music, it is no longer the
perceived over the spectrum of audible frequencies.
case. Today there is an extreme distance between a
No doubt thereader has already been triggered into
very tight, perhaps rigidly controlled, punctual,
imaginingtypesof soundsin association with motion
homorhythmic, minimal music, and thevery relaxed,
and growth, texturemotion, or behaviour, but when
malleableassociations found in someelectroacoustic
it comes to describing more precisely what these
music. I ndeed, spectromorphologies could be so sounds might bewerun into difculties. Composers
different in their character and motion that theonly and researchers are used to thinking of spectra in
relationshiptheyseemto haveisthat they exist inthe termsof their spectral componentswhichcan beana-
same space at the same time. That, of course, is in lysed, extracted, reconguredandtransformedto cre-
itself a behavioural relationship. atenew spectromorphologies and transformexisting
I n motion passage, the voluntarypressured con- ones. This detailed, analytical approach to spectra
tinuumexpresses how one context or event yields to can only beused in spectromorphological thinkingif
thenext. Wearereminded that such questions as the it is perceptually based and relevant to the musical
degree of motion coordination, the energymotion context. Eventhen, what terminologydo weuse, par-
ticularly when individually perceived pitches are not trajectory of gesture, thepressuringurgency of onset
Spectromorphology 119
Figure4. Behaviour.
involved? Source-bondingin itself draws attention to among the components of the harmonic series. So a
notecan bea notealonewhen viewed externally, or spectral attributes associated with thetypeof source:
namingareal or imaginedsourceisanact of differen- morethan a singlenotewhen theinterior is entered.
Opportunities for internal spectral focus arecommon tiating spectral qualities. No doubt we can describe
spectral attributes of the source in terms of whether in electroacoustic spectra in general, not just with
notes. thesound isbright, dull, hollow, thin, intense, and so
on. We need to expand this terminology of qualita-
tive description in order to deal more comprehen-
sively with aspects of spectral space.
11.2. Notecollectives relativeandintervallicpitch
Whether pitches are heard or not is of prime
When external notes movetoo quickly for theear to
importanceinspectral discrimination. I must rst dis-
hear precisepitch-intervals, or arepiled up in densi-
cussthenatureof thenoteboth asamanifestation of
ties, the note ceases to be perceptually signicant as
pitch and as a typeof spectrum. That will then allow
a pitch. The external note may be theoretically pre-
meto discuss spectra without audiblenotecontent.
sent (we actually knowit is in complex instrumental
textures) but perceptually absent. Wemust therefore
11.1. Thenote internal andexternal spectral focus
differentiatethepoles on a continuumbetween inter-
vallic pitch and relative pitch. I n intervallic pitch we
Electroacousticmusicallowstwo note-views. Therst
can hear pitch-intervals, and thereforetheir relation-
isthetraditional viewwherethepitch of anoteisthe
ship to cultural, tonal usage will become important.
most important factor and the spectral components
I n relativepitchcontexts wehear with much less pre-
areregardedascolouringthenote. Thespectral com-
cision thedistancebetween pitchesand canno longer
ponents of the note remain largely unheard, or at
hear exact pitches or intervals in spectral space.
least ignored because it is the note itself which mat-
I nsteadweareinclinedto followhigher-level gestures ters most to thecontext thenoteand its colour are
and motionsnote collectives. Much electroacoustic perceptually fused. An alternativeview(aspermitted,
music merges intervallic and relative pitch for example, by close recording of an instrumental
approaches. J udging which approach predominates note), takes the ear inside the note so that spectral
at a given moment is a problem for the composer components can be heard. Such a change of focus,
(whomweassumehasa good ear for pitch). For list- created by actually moving insidethesound, is com-
eners the borderline between the intervallic and the mon in spectromorphological music. The compo-
relative is equally important. Firstly it is not a well- nents are then heard to have a pitch relationship,
dened border sinceit isvery much inuenced by the which for themost part insidetheinstrumental note
will be a harmonic one, with varying weightings listeners perceptual skills. Secondly it is important
120 Denis Smalley
becauseintervallic pitch, if obviously present, will be to describe its riches. On the other hand, noise can
occur in narrower or wider bands, and become theprimefocus of attention for most listeners.
Theuseof thewordintervallic in intervallicpitch colouredandresonant so that pitch(either relativeor
absolute) becomes blended in. Therefore, whileinter- draws attention to the fact that more than a single
pitch is needed for pitch to becomeperceptually rel- vallic pitch and noisearein onesenseextremes, noise
can takeon apitch identity, just aspitch can takeon evant. Therearemany circumstancesinelectroacous-
tic music wherepitch is a recognisableattributeof a noise content. Although noise-based spectromor-
phologies have been extensively explored in electro- spectromorphology, but if there is no intervallic
relationship with another pitch, the pitch-attribute acoustic music, their qualities have yet to be more
widely appreciated, so dominant is the conditioning becomes a background factor, and there may be
other qualities of the particular spectromorphology of the intervallic pitch heritage. Electroacoustic
music, in enabling noise exploration, seeks to which arecontextually moresignicant. Even with a
pitch-drone, theactual pitch (as opposed to thefact embrace the full spectral potential of the wide-open
sound world. that it is high, low, or a stablereference-point, or its
motion is textured) is unimportant if intervallic pitch
relationships areabsent.
11.4. Harmonicityandinharmonicity
While harmonic spectra have a specic intervallic
11.3. Fromnotetonoise
organisation based on the vibrational properties of
The difference between the note (as an intervallic
strings and columns of air, inharmonic spectra do
pitch) and noise is an extremely important spectral
not. The bell and metallic resonances are the usual
distinction, yet this is another border which is dif-
examples of inharmonicity, and they suitably repre-
cult to dene. For my purposes, noise, as an extreme
sent the inharmonic dilemma because inharmonic
compared with thenote, can bedened in two ways.
spectra can be ambiguous in that they can include
Therst denitionisqualitativenon-pitchedrough-
some intervallic pitches. To be regarded as properly
ness, granularity or grit. Extrinsically we associate
authentic, an inharmonic spectrum cannot be
granular noise with sea, water textures, wind, static
resolved as a single note, and its pitch-components
interference, granular friction between rubbed and
need to be considered relative, not intervallic. As a
scraped materials, fracturing materials (e.g. stone),
result, continuous inharmonic spectra have a tend-
unvoiced vocal consonants, and certain types of
ency to disperseinto streams.
breathing and uid congestion. Therefore a wide
I nharmonic ambiguity allows spectral change in
range of source bonding is involved, ranging from
two directions. Firstly one can move into intervallic
humanutterance, to playwithobjects, totheenviron-
andharmonic(tonal) spectra. Secondly, likethespec-
ment. This is a rich reservoir for electroacoustic
tral compression mentioned in discussing the note,
exploration. Granular noiseistextured impulses, and
inharmonic saturationthe adding of spectral
need not be dominant in a spectromorphology. For
componentscan be a means of moving towards
example, it can beadecorativeor subordinatestrand
noise. I nharmonicity can therefore occupy a useful
or trace, or a pocket of added intensity.
Thesecond denition is not distinct fromtherst,
and is concerned with densitya saturated spectral
statewhichcannot beresolvedinto intervallicor rela-
tive pitch. Saturate noise can be looked upon as
another aspect of some of the sources mentioned
above(e.g. sea), but it can also comeabout through
spectral compression, wherean areaof spectral space
isclosely packed suchthat pitchawarenessisimposs-
ible. I n addition it can occur when spectral space
becomes lled by the active contours of convoluted
and turbulent motions. Thus there are certain
accumulativeprocesseswhichtendtowardsnoiseand
can beused to createnoise.
Noise is relative rather than absoluteit exists
becausewehavea concept of pitch. I ntervallic pitch
is an absolutewe can perceive and name intervals
preciselywhereas noise is a generality and has to
beconsidered spectromorphologically in terms of its
motion, texture, and behaviour if we are to be able Figure5. Fromnoteto noise.
Spectromorphology 121
middleground which allows movement towards har-
monicity and intervallic pitch on the one hand, and
noiseon theother.
I n gure 5, the possibilities of transformation
among note, inharmonicity and noise, are set out,
along with the possibility for the note (intervallic
pitch) to move in the direction of both harmonicity
and inharmonicity.
11.5. Spectral spaceanddensity
Spectral space covers a distance between the lowest
andhighest audiblesounds. I ndiscussingmotionand Figure7. Spectral density.
growth processes it was assumed that spectro-
morphologies move through spectral space as they
changeover time, and in discussingtheseven charac-
(1) Emptinessplenitudewhether thespaceisexten-
teristic motions, attention was drawn to the root-
sively covered and lled, or whether spectro-
edness of certain motions and thenon-rootedness of
morphologiesoccupysmaller areas, creatinglarge
others. I n other words, certain motions need to
gaps, givingan impression of emptiness and per-
occupy spectral spacein different ways.
haps spectral isolation.
I n instrumental music (if weknowtheinstrumen-
(2) Diffusenessconcentrationwhether sound is
tation) and vocal music, wehaveprior knowledgeof
spread or dispersed throughout spectral spaceor
thepotential spectral spacenot only of theensemble
whether it is concentrated or fused in regions.
but of individual instruments and voices as well, and
(3) Streamsintersticesthe layering of spectral
wehaveexpectationsof theuseof spectral spacerela-
spaceinto narrowor broad streams separated by
tive to the musical style. I n electroacoustic music,
interveningspaces.
spectral spaceboundaries arenot known in advance
(4) Overlapcrossover how streams or spectro-
but dened in the course of a work. The occupancy
morphologies encroach on each others spectral
of spectral space, the impression of the space-
space, or move around or across each other to
breadth, howit unfolds, wherethehighsandlowsare
another region. This is directly related to motion
locatedandhowthey arereached, aredirectly related
and growth processes.
to thelisteners interpretations of extrinsic factors as
well as beingstrongformal determinants. Weneed a
Figure7denes qualities of spectral densitywhich
descriptivevocabularyto helpdenetheoccupancyof
can beimagined as a fog, curtain or wall of broader
spectral space. Figure 6 shows three basic reference-
or narrower spread which allows soundsto penetrate
points. Canopies and roots can be regarded as
or not. I n other words, is thedensity such that other
boundary markers which may have functions. For
spectromorphologies, if superimposed, can or cannot
example, texturescan behungfromcanopiesand use
stand out in relief? Thus a packed or compressed
themas goals or departure points, while we already
spectral spaceis compacted so that is suffocates and
know that the drone can act as a root-reference.
blots out other spectromorphologies. A transparent
Together they framespectral space, although they do
spectral space lets spectromorphologies through,
not haveto beheardsimultaneously to do so. Weare
whilesomethinginbetween(translucent, opaque) has
used to a central region in music, and even if this is
a masking effect. Spectral density is related to dis-
not clearly dened in an electroacoustic work, weare
tance perspective and needs to be considered along
nevertheless awareof whereit is. A frame, of course,
withspaceingeneral. Thisperspectiveaspect isrepre-
implies a vacant central region.
sentedbythedistantclosecontinuumingure7. For
The following four qualiers help further to
example, a packed density of full spectral range, set
describetheoccupancy of spectral space, whether in
in close foreground focus, will prevent other
part of a work or in thework as a whole:
spectromorphologies fromgettingthrough becauseit
creates a solid wall very close to the listener. The
same density set further back will free space closer
to the listener so that it can be occupied by other
spectromorphologies. Of course a density need not
have a xed perspective. There are perceptual limits
to how much spectromorphological information can
occupy spectral and stereo space: high density is the
enemy of low-level detail. Figure6. Occupancy of spectral space.
122 Denis Smalley
12. SPACE AND SPATIOMORPHOLOGY thedouble-headed arrow shows theinterdependence
of thecomposed space(thespaceas composed on to
The concept of spectral space is analogical: higher
recorded media), and thelisteningspace(thespacein
pitches can be thought of as spatially higher, and
which thecomposed spaceisheard). Listeningspaces
lower pitches as lower, but in terms of actual spatial
themselves comprise two categories which are dis-
locationtheyarenot normallyphysically higher and
tinguished by the different position of the listener
lower. Onereason for equatingspectral heightdepth
relative to the radiating sound sources (loud-
with actual heightdepth in electroacoustic music
speakers). I n personal listeningthelistener is closeto
could be to do with the fact that high pitches are
the sources of a frontal image, while in public con-
regarded as physically smaller and therefore not
texts thelistener could bein any oneof a variety of
rooted. Another reason could bethat high degreesof
distant or off-centre positions relative to a frontal
spectral mobility areconcerned with higher registers
reference-image. Furthermore, diffused space(where
(analogy with ight?) which can often be clearly
a multi-speaker system is used) permits a radical,
localised compared with the vague, more spread
expansive, multidirectional reorientationof the(com-
localisation of bass sounds. Anyway, spatial percep-
posed stereo) space. Boththepersonal listeningspace
tion combines thesetwo aspects, one analogical, the
and diffused listening space are open to widespread
other actual.
abuse which undermines spatial perception. I n per-
The discussion in this section is concerned with
sonal spaces, loudspeakers and listener are often
dening a grammar of localisation. Spectromorphol-
casually positioned without thought for stereo imag-
ogical changesin spectral space and density, in
ing needs, whilesuccessfully diffused spaceis a com-
dynamic level, in motion and growth processesin
plex combination of loudspeaker types, placements,
themselves imply spatial settings and motions, and
and theart of diffusion itself.
theseimplicationscanbedevelopedby thecomposer.
I haveidentied ve, basic, personaldiffusedspace
As well as enhancing the character and impact of
variantswhich can bedrastically affected, or changed
spectromorphologies, changes in spatial perspective
for better or worse, accordingto listeningspacecon-
are a means of delineating musical structure. Real
ditionsanddiffusionstyle. Listenerscanonlybecome
spaces can besimulated faithfully, as is thecaseif a
really aware of the variants if they have had an
music is set in a simulated roomor hall acoustic. But
opportunity to compare perceptions of the same
electroacoustic music is not limited to spatial reality,
work under different listening conditions, not a
and the composer can, for example, juxtapose and
possibility which is commonly available.
rupture spaces, an impossible experience in real life.
Thecomposed spaceitself can bedivided into two
Electroacoustic music can encapsulate a wide range
categories. Internal space occurs when a spectro-
of spatial experience, perhaps even a life-longexperi-
morphology itself seems to enclose a space. Reson-
enceof intimateand immensespaces, both of which
ances internal to objects (hollow wooden resonance,
can becompressed into therelatively short time-span
metallic resonance, stringed instrument pizzicato res-
of a musical work. This makes electroacoustic music
onance, etc.) can give the impression that their
a uniqueart.
vibrations areenclosed by somekind of solid mater-
Spatial perception is inextricably bound up with
ial. I nternal spaceis thereforesourcebonded in that
spectromorphological content, and most listeners
oneneeds this senseof an actual or imagined sound-
cannot easily appreciate space as an experience in
ingbody.
itself. Spatial appreciation can be acquired by con-
External space (gure9) is much moresignicant
sciously listening to the spaces in works as distinct
than internal space because there can be no music
from regarding space only as spectromorphological
without it, whereasasenseof internal spaceneed not
enhancement. I use the term spatiomorphology to
exist at all in a work. External space, which is made
highlight this special concentration on exploringspa-
apparent to us through reections, exists outsideand
tial propertiesandspatial change, suchthat theycon-
around spectromorphologiesoutdooropen spaces
stitute a different, even separate category of sonic
or indoorenclosed spaces which providesettings for
experience. I n this case spectromorphology becomes
spectromorphological activity. External spaces are
themediumthroughwhichspacecanbeexploredand
perspectival, and in the case of frontal stereo, the
experienced. Space, heard through spectromorphol-
analogy withlinear perspectivevisioncanbestriking:
ogy, becomes a newtypeof source bonding.
in looking through the stereo window between the
Both the grammatical details and the psychologi-
loudspeakers, thelistener canapprehendspacesmuch
cal messages of spatial apprehension are unstable broader than the real, space-breadth between the
becausethey depend not only on spaceas composed, speakers, spaceswhich stretch beyond theconnesof
but on therelationship between thecomposed space the listening spaces actual depth. A sense of spatial
and the space(s) in which listening takes place. This intimacy occurs when spectromorphologies seem to
act closeto thelistener asif inhabitingthesamespace is therst major distinction shown in gure8, where
Spectromorphology 123
Figure8. Composed and listeningspaces.
Figure9. External space.
124 Denis Smalley
asthelistener. (Suchintimacyiseasilylost indiffused Their velocity will vary accordingto thedesiredener-
getic proleof thetrajectory relativeto thedistance listeningspace.)
Image denition can be elaborated in two sets of to be covered, and will be very closely linked to
changes in spectral energy. Five basic paths repre- variants. Opacityandtransparencyhavealreadybeen
mentioned in discussing spectral density. Here they senting the simple breadthdepth directionalities are
listedapproach, departure, crossing, rotation, and arerepeatedasattributesof spatial focus, but arealso
extended by the idea of spatially blurred or clear wandering. A trajectory is not necessarily a concen-
trated point-source. As thehead or bulk of a gesture imageswhether the occupancy of space is clearly
dened or not. Weshould notethat a distant image moves through space it can leave residues behind.
Trajectoriescanthereforeleavetrails, canbesmeared could beblurredor clear, ascould acloseimage. The
second variant related to imagedenition is concen- across space or be spread in a more egalitarian way
through space. I t may be that the establishing of a trateddiffusefocuswhether thelisteners attention
focuses on a small (central) area of space or on residue is part of a transformation of a gesture into
a spread settingthespread setting is introduced by diverse activity spread through space. This, in turn,
is linked to the idea of spatial ll how sparsely or a trajectory.
Finally, herearesomeguidelinesto help denethe densely populated thespaceis.
Spatial texture is concerned with how the spatial global spatial stylein a work as a whole:
perspective is revealed through time. This is a ques-
(1) Single spatial setting. The single setting has two
tionof contiguity. Spaceiscontiguouswhenrevealed,
aspects. A work can be set in a single type of
for example, in continuous motion through space
spaceof which thelistener is awareat theoutset.
(such as in a leftright gestural sweep), or when a
On the other hand, different aspects of a space
spectromorphology occupies a spread setting (with-
can be revealed over time. Spatial awareness is
out spatial gaps). Non-contiguous space is revealed
cumulative, and the listener eventually realises
when spectromorphologies arepresented in different
that thereis a global spatial topology into which
spatial locations such that two successive events are
thewholework ts. For example, theextremesof
not considerednear neighbours: thereisno sensethat
proximity and distanceareunlikely to beknown
a spectromorphology occupies or moves through
until thework has advanced somewhat.
adjoining sectors of space. The distinction between
(2) Multiple spatial settings. Throughout the work,
contiguity andnon-contiguity isnot necessarilyclear-
the listener is aware of different types of space
cut. I magine a very active, ongoing, rough texture,
which cannot beresolved into a singlesetting.
presented across thebreadth of stereo spacein many
(3) Spatial simultaneity. I magine a very present
scratchy point-sources of varying intensity. These
granular texture directly in front of you as if
point-sources arepresented non-contiguously so that
actually within your listening space, whilein the
attention may bedrawn to any part of thehorizontal
distance a door closes in a large reverberant
spacewherea particular scratch stands out in relief.
space. You areawareof simultaneous spaces.
Thus the attentive ear (eye) darts among non-
(4) Implied spatial simultaneity. I mplied simultaneity
contiguous positions. Yet, taken over a sufciently
occurs when the listener remains aware of the
long duration, the individual non-contiguous points
existenceof aspaceinitsabsence. Thiscanoccur,
of thistexture, perceived asawhole, cover acontigu-
for example, when contrastingspacesareintercut
ousspace. Thusthisspaceisnon-contiguously erratic
and alternated (spatial interpolation), giving the
at a low level but contiguous at a higher level, a
impression of simultaneity even though the
characteristic spatial behaviour for active textures.
spaces are presented successively. This is related
Contiguity therefore differs according to the struc-
to lm, wherein spiteof thecuttingbetween suc-
tural level considered, and is determined by thetype
cessiveevents, they areconsidered concurrent.
of spectromorphology and the degree of continuity
(5) Spatial passage. Passagebetween spaces can be
over time. The distribution style of non-contiguous
sudden (interrupted passage), repeatedly intercut
space can vary between the presentation of isolated
(interpolated passage), or moregradually merged
spatial points, and scatterings over smaller or larger
(graduatedpassage).
areas. Distribution patterns may emerge as a result
(6) Spatial equilibrium. What is the relative balance
of recurrences of the same spatial positions or rep-
between types of perspective and spatial texture
etitions of sets of locations. For example, leftright
in the work? I s one type of space emphasised
exchangepatterns arevery common.
more than another? Are there alternations or
Contiguous spacecan berepresented by a spectro-
reciprocal exchanges between spaces?
morphology which is spread over an area, but gestu-
ral trajectories are particularly characteristic. They
13. FINAL REMARKS
can follow direct (linear) or indirect paths (direction
changes, or varieties of curvilinear trajectory), and Spectromorphology is concerned with perceivingand
thinking in terms of spectral energies and shapes in even subdivide and disperse in multiple directions.
Spectromorphology 125
Cogan, R. 1984. NewImagesof Musical Sound. Cambridge:
space, their behaviour, their motion and growth pro-
Harvard University Press.
cesses, and their relative functions in a musical con-
Delalande, F. (no year). Analyse musicale et conduites de
text. Although the detail of spectromorphological
re

ception: Sommeil de Pierre Henry. Paris: I NA-

descriptionmaysometimesnot beeasytofollow, par-
GRM.
ticularly without anextensiveexperienceof electroac-
Duchez, M.-E. 1991. Le

volution scientique de la notion

ousticmusicrepertory, it isfar frombeinganesoteric
de mate

riau musical. I n J .-B. Barrie

`
re (ed.) Le timbre,
activity. Spectromorphological thinking is basic and
me

taphore pour la composition, pp. 4781. Paris:

easily understood in principle because it is founded
I RCAMChristian Bourgois.
on experience of sounding and non-sounding
Emmerson, S. 1986. Therelation of languageto materials.
phenomena outside music, a knowledge everyone
I n S. Emmerson (ed.) The Language of Electroacoustic
hasthereis a strong extrinsicintrinsic link. I n this
Music, pp. 1739. Basingstoke: Macmillan Press.
sense spectromorphology derives from a common,
Erickson, R. 1975. Sound Structure in Music. Berkeley:
shared, natural basewhich provides a framework for University of California Press.
Lerdahl, F. 1987. Timbral hierarchies. I nS. McAdams(ed.) the individual, cultural works of electroacoustic
Music andPsychology: A Mutual Regard. Contemporary
music. Discovering and dening the natural link is
Music Review2(1): 13560.
important for composerlistener communication
McAdams, S., and Saariaho, K. 1985. The qualities and
because new musical languages (if such a thing is
functions of musical timbre. I n Proc. Int. Computer
really possible) or signicant shifts in language are
MusicConf. 1985, pp. 36774. SanFrancisco: Computer
not created in a vacuum, but must havesomeshared
Music Association.
natural-cultural basisif theyareto makesenseto list-
McAdams, S. 1993. Recognition of sound sources and
eners. No rationale or theory has to be worked out
events. I n S. McAdams and E. Bigand (eds.) Thinking
in advance of language shifts, but ultimately the
inSound: TheCognitivePsychologyof HumanAudition,
workings of any new language must be explicable.
pp. 14698. Oxford: Clarendon Press.
As a result, not only can we help explain how and
Manoury, P. 1991. Leslimites dela notion detimbre. I n
why electroacoustic music is as it is, but wecan also
J .-B. Barrie
`
re(ed.) Letimbre, me

taphorepour la compo-
havea meansof articulatingproblemswhen wereact
sition, pp. 293300. Paris: I RCAMChristian Bourgois.
badly to a particular work (and there are plenty of Meyer, L. B. 1973. Explaining Music: Essays and Explo-
rations. Berkeley and Los Angeles: University of electroacoustic works which are unrewarding). This
California Press.
is all the more important for a music which is so
Nattiez, J .-J . 1990. Music andDiscourse: TowardsaSemiol-
closely allied to a means of productioncomputers
ogyof Music. Princeton: Princeton University Press.
and technologywhose role is mysterious and
Risset, J .-C. 1978. Hauteur et timbredes sons. I n IRCAM
unknowabletomost listeners, particularlysincetradi-
Reports 11/78.
tional, instrumental and vocal gesture are often
Risset, J .-C. 1991. Timbre et synthe
`
se des sons. I n J .-B.
absent or not immediately apparent.
Barrie
`
re(ed.) Letimbre, me

taphorepour la composition,
Spectromorphology, therefore, in starting from a
pp. 23960. Paris: I RCAMChristian Bourgois.
decodingof perception, isan attempt to makecollec-
Roy, S. 1994. Analysefonctionelleet implicativedOmbres
tivesenseof a widerangeof individual electroacous-
Blanches. I n A. VandeGorne(ed.) Francois Bayle: par-
tic musics created since the birth of the medium in
cours dun compositeur, Lien, revue desthe

tique musi-
the1950s.
cale, pp. 1349. Ohain: Musiques et Recherches.
Schaeffer, P. 1966. Traite

des objets musicaux. Paris: Seuil.

Smalley, D. 1986. Spectromorphology and structuringpro- REFERENCES
cesses. I n S. Emmerson (ed.) The Language of Electro-
Agawu, V. K. 1991. Playing with Signs: A Semiotic Inter-
acoustic Music, pp. 6193. Basingstoke: Macmillan
pretationof Classic Music. Princeton: Princeton Univer-
Press.
sity Press.
Smalley, D. 1991a. Acousmatic musicdoes it exist? I n
Bachelard, G. 1994 edn. The Poetics of Space. Translated
A. VandeGorne(ed.) Vousavezdit acousmatique?Lien,
by Maris J olas. Boston: Beacon Press.
revuedesthe

tiquemusicale, pp. 212. Ohain: Musiques

Barrie
`
re, J .-B. 1991. I ntroduction. I n J .-B. Barrie
`
re(ed.) Le
et Recherches.
timbre, me

taphorepour la composition, pp. 1113. Paris:

Smalley, D. 1991b. Spatial experience in electro-acoustic
I RCAMChristian Bourgois.
music. I n F. Dhomont (ed.) LEspace du son 2, Lien,
Bayle, F. 1993. Musiqueacousmatique: propositions. . . pos-
revuedesthe

tiquemusicale, pp. 1214. Ohain: Musiques

itions. Paris: I NA-GRMBuchetChastel.
et Recherches.
Cadoz, C. 1991. Timbreet causalite

. I n J .-B. Barrie

re(ed.)
Smalley, D. 1992. Thelisteningimagination: listeninginthe
Le timbre, me

taphore pour la composition, pp. 1746.

electroacoustic era. I n J . Paynter, T. Howell, R. Orton
Paris: I RCAMChristian Bourgois.
and P. Seymour (eds.) Companion to Contemporary
Chion, M. 1983. Guide des objets sonores. Paris: Buchet
Musical Thought, Vol. 1, pp. 51454. London:
ChastelI NA-GRM.
Routledge. Chion, M. 1991. Lart des sons xe

s ou la musique con-
Smalley, D. 1993a. Canelectro-acousticmusicbeanalysed? cre
`
tement. Fontaine: Editions MetamkineNota-Bene
Sono-Concept. I n R. Delmonte and M. Baroni (eds.) Atti del Secondo
126 Denis Smalley
ConvegnoEuropeodi Analisi Musicale. Trento: Universi- 87110. Translated by Alessandro Cipriani and
ta di Trento. Giuseppi EmanueleRapisarda.
Smalley, D. 1993b. Dening transformations. Interface Ten Hoopen, C. 1994. I ssues in timbre and perception.
22(4): 279300. ContemporaryMusic Review10(2): 6171.
Smalley, D. 1994. Dening timbrerening timbre. Con-
Thompson, dArcy. 1961 edn. On Growth and Form.
temporaryMusic Review10(2): 3548.
Cambridge: CambridgeUniversity Press.
Smalley, D. 1995. La spectromorphologie: uneexplication
Truax, B. 1984. Acoustic Communication. Norwood, NJ :
des formes du son. I n Louise Poissant (ed.) Esthe

tique
Ablex PublishingCorporation.
des arts me

diatiques, To

me2, pp. 12564. Translated by

Wishart, T. 1996. On Sonic Art, S. Emmerson (ed.).
SuzanneLeblancandLouisePoissant, revisedbyDaniel
Amsterdam: Harwood Academic Publishers.
Charles. Sainte-Foy: Presses delUniversite

du Que

bec.
Wishart, T. 1986. Sound symbols and landscapes. I n
Smalley, D. 1996. La spettromorfologia: una spiegazione
S. Emmerson (ed.) The Language of Electroacoustic
delleformedel suono. I n MusicaRealita
`
50/51(divided
Music, pp. 4160. Basingstoke: Macmillan Press.
in two parts), J uly 1996November 1996, pp. 12137