Order 2336844: Timbre/Sound Color: What is it?
ErnstBloch-MagicRattle_Press1986dragged Fales_ParadoxofTimbre
- Type of paperEssay (Any Type)
- SubjectMusic
- Number of pages3
- Format of citationMLA
- Number of cited resources0
- Type of serviceWriting
For this assignment you′re going to analyze the two readings by Bloch and Fales and write how it relates to timbre and figuring out what exactly is timbre (as best as you can describe it). Please don’t include lengthy desсrіptions or historical information about the readings, its composer, etc. The point of this essay is to analyze the readings, not describe it. The max amount of words used should be 750 and in PDF Form.
II
Magic rattle, human harp
The vibrating note travels. It dóes not remain in its place, as colour
does. True, colour is likewise emitted to catch the attention, but then
it stays put. For a white to detach itself from a garment, or a wall, is
unthinkable. In contrast, the whole of the surrounding air can be full
o f a sound.
There was a time when the musical note did not appear such a free
agent. It was linked quite specifically with the instrument producing
it. When first consciously produced, it was wholly attached to its
instrument and had no other association. Thus the original rattle
rattled as the thing it was; the rattling sound is merely its verb, as it
were. The thunder-stick whirred and the drum beat itself: that was
the main thing. Thus here the sound is an attribute of the instrument,
to which it is linked in a purely material sense. Its sonorousness is
used for magical purposes, for healing the sick, driving away evil
spirits and summoning good ones. But it is not primarily the sound
which performs the spell, but the actual magic drum being used. The
crucial element is the sacred instrument, as ancient as possible,
specially painted or with special incisions. At the most it is through
the power o f rhythm, in the dance, that music as such becomes more
advanced, itself exerting an immediate influence. Yet, even now, the
drum will fascinate by being hollow, subterranean, the cymbal by
being metallic.
The more the musical note developed, the more it parted from this
foundation. It surmounted its instrument, so to speak, and now used
it as merely a means o f assistance. Now the ringing and tinkling
broke loose from the ringing brass and the tinkling bell; musicians no
longer just ‘attended on’ their instruments but availed themselves of
them. This change is marked by the invention of the pan-pipe: the
first instrument which did not emit fearsome or muffled sounds in
isolation but gave out a well-ordered series o f notes. It is at this point
that music is born, music as a shout of sorrow or pleasure and not as
material magically sounding. No longer do we hear the speech o f the
wood or metal, but neither do we discern the ‘soul’ of a flute or horn.
140
MAGIC RATTLE, HUMAN HARP
Instead the miracle of music is now itself triumphant in every quar-
ter. Previously the ‘tricking out’ o f the rattle with bird feathers
‘charged with spirits’, and o f drums with magical incisions, was at
least as important as the frightening or alluring sound that they
made; but now, this faith in the material thing disappears completely
from the musical viewpoint. On a miniature scale, the rattles, drums
and thunder-sticks had even been worn as lucky charms, but not the
pan-pipe, which was unconsecrated. The rattling field o f the instru
ment yields to the opening up of the auditory field, in which music
had not until now, after its deliverance from the magic o f the material
thing, made its home. T o be sure, the compass, technique and
expression o f the instruments are and remain of importance, for
it is only with amateurs that the instrumentation is a secondary con-
sideration, or even interchangeable. But even where,. desper ate
attempts were made to bind the exodus o f music tcfthe knownFealm^
a magicking of the instruments or f i j i F u r á ^ in
this oíd style was absent. It is not the tinkling bell but^Ke tinkle itself\ JjA
that now emerges from(materíal^leprived o f itsímagic} And in the
process, the material can Tounder completely in the covered
orchestra^The musical note evinced much vigour in turning from the
attribute o f a thing intojthe yery thing that matters, i n a developed
State; from an adjective into a substantive; from ag^YtulToug)
excrescence of objects that were rubbed, struck or blown ínfo^mnH
versal, t h o u g h ) ^ ^ ^ l!k t tis t ic ^ é a ln ^ with melodic and above all
human relations o f its own!/%usih which to start with only denoted
a piece o f wood that men shook or sacred implements for their hear-
ing, furthermore became, over and beyond its physical character, a
colossal realm whgse.,qbjective nature was still largely unknowru/If
one were bent on^demonstfatíhg an object\for it in the known realm,
one no longer referred to sacred instruments, even in the context o f
magic objects. Rather, notes and their scales were related to whole
constellations, to the harmony of the spheres, to the ‘Lyra Apollinis’
that was supposed to stretch across the seven known planets. But
even here, the actual melodies were still floating through space, not
located or indeed incarnated in some existing object. Even so-called
tone painting ‘describing’ rain or thunderstorms, where it possesses
any musical merit, registers only the psychomorphous action of rain
falling, a storm erupting, daybreak or nightfall, without having
localised such events in a concrete sense. And as for the ‘Lyra
Apollinis’, the harmony of the spheres as echoed in the scales and
ffe ie o i l m
> i íA
Q Ut f/
3 {/■(
A
h
t í (■ >
141
MAGIC RATTLE, HUMAN HARP
‘ ̂ harmony of music (% mperstitior^ which persisted from the time of
\ the Pythagoreans up to Kepler and indeed Buxtehude), music’s real
development has confounded all such ‘natural’, indeed astronomi-
cal, object-definitions. W hy, even in the much more genuine realm of
‘human’ object-relations, music still travels under top-secret orders.
And assuredly these contain no references, directly or primarily, to
an origin linked with material things and the magic o f things.
There are nonetheless, amid the fluctuations, some noteworthy
remains. Sound’s ability to be everywhere and nowhere has asserted
its claim, yet it is not necessarily lacking in the instrumental ‘where’
which bound it to start with. In the first place this is still conveyed by
the fact that trumpets are associated with kings and trombones are
associated with priests. Trumpets which ‘proclaim that kings are
approaching’ – what we have here is not just convention or external
historical association. On the contrary, the trumpet and the trom-
bone are themselves standing for monarchic or sacerdotal might,
alm ost irrespective o f w hat they are playing. Granted, the trumpet
has ceased to be a sonorous fetish, yet neither is it just the attribute of
a king, since it resounds as the actual r o ja l splendour which this king
is projecting. And secondly, instruments portrayed as having their
own magic still occur on the stage, in magic operas; here they are por
trayed with great forcefulness, even. W ith their diabolical clattering
in the orchestra, Dr M iracle’s tinkling phials in T he Tales o f
H offm ann belong to this category. They make a light, glassy sound
compared to the D o ctor’s roaring and shrieking, but one that is
equally spellbinding. Directly(archatc>— on the brighter side of the
coin – are the horn which plays in O beron , Papageno’s bells in The
M agic Flute and the magic flute itself, carved by Pamina’s father
‘from the very base of the thousand-year-old oak, by thunder and
lightning, gale and tempest’. And once again it is not just their
melody which casts a spell, but the fetish to which it is given. In Dr
M iracle’s case the flute of the Pied Piper is suggested, and in
Tam ino’s the zither playing o f Arion and Amphion, not only in
respect o f their great artistry but in respect o f the magic instrument
in each case as well. Ultimately, to be sure, this element o f the magic
object appears to be only a surrogate for the supreme concentration
o f music at a crucial point in the action, viz. a point effecting some
transformation. Takeftfó^ nuhp^^^nm y-scarcely still remembered
in terms of ^ m a g ic’D h je ^ — which, purély as melody, echoes down
into Florestan’s dungeoíTin Beethoven’s Fidelio. At that stage the
142
MAGIC RATTLE, HUMAN HARP
trumpet is almost completely forsaken, and in it we hear solely the
M inister’s voice or the realm o f salvation which has suckled
music since it first became Christian. But even the archaic enclaves
which preceded it (in Papageno’s bells, in Tam ino’s flute) no longer
qualify as vestigial proofs o f ancient magic instruments. They fail to
do so at all in terms o f the philosophy o f musical metal, and they
barely qualify in terms o f the gnostic-celestial instrumental sym-
bolism that Abert rediscovered in his book, M usikanschauung des
Mittelalters. It is not the paucity o f the remains that contradicts the
‘pagan’ interpretation o f music in the context of magic objects (for
the remains are nonetheless lodged in a high, if not the highest,
musical authority). But even f a consecrated. befll which is another
revealing remnant, rings not from the depjh o f the metal but down
from aloft; essentially it is within the church tower that the metal
thus dedicated is hanging and operative.|ln an orchestra, jh e belPs
‘metal mouth’ is, rather̂ a subject for ridicule. Indeed, even music
whose áim is black magic (like Stravinsky’s Rite ofSpring) derives its
instrumented gruesomeness only incidentally from the clanging
brass or resonant bell. — . . „
W e are now coming to the actual purpose o f our enquiry. For all
this harking back brings us to the singer, the problem o f the singer
himself, especially in opera; he is the ‘remnant’ which shows the most
life. The singer really still is a beating drum, or rather, a harp that
plays its own music. Here — in contrast to those previous remnants
which were rarely found and faintly bizarre, though important — the
musical note has a definite site, namely a body that sings in the ‘
cess o f acting. Here music, in the process o f floating, alights onxa
visible instrumental provenance; it thereby adheres at the same timé\
to other visible objects and denotes them. By means o f t h e singer, the ]
melismatic and even the symphonic elusiveness of music finds itself ̂
placed in an operatic scene and the action which this localises. In solo
song and in oratorio.ÍE eliu m an voice isju st an instrument like other 3
instruments, and the singer just a mildly stressed vehicíe for the
“n^sidC’But as for opera, this is not only a place where tKehm han
voice will not cause an obstruction. Here — and not just to further the
expansión of human song — it is a form of dramatic action which is
necessary and supremely justified in a specifically musical context.
And in using singers.it puts instruments on the sta^e that ^ phys.i-
cally active and significant once more: not sacred instruments, but
ones^wítK’a’ natural emphasis. Here w e have on the stage a Ttind o f
143
MAGIC RATTLE, HUMAN HARP
human harp. This instrument’s voice signifies purely the instrument
itself; only on a secondary level do its actions signify the instrumen-
tally detached musical content. Certainly the origins o f opera are not
archaic in the aforesaid sense. Yet the opera singer is distinctly in
touch with the archaic remains, with music as the attribute of an
object set in motion and emitting its own sound. Although this
parallel may seem an odd one, it is one o f the facts of music and has
proved to be effective in practice. Apárt from the music within whose
fabric they are found, soprano and alto, tenor, baritone and bass do
in fact ‘attend on’ the instrument which they represent and with
which théy are linked through a personal bond. Henee the associ-
ation of individual vocal categories, i.e. individual ‘instruments’ in
the vocal depártrnent, with specific natural qualities o f the m ale ór
fem ale ch a ra ctersc o rresponding to them ispartly extra-musical and,
indeed, pre-musical. Here music is not ‘formallylfserf-suFfícient|at all.
Ñ or does it articúlate in an expressive role — as still applies in the solo
song — suffering, longing, anger or love ‘in general’, to cite Schopen-
hauer’s abstract and too emancipated interpretation. Rather, it is
now the case that music is attached to the material o f its instruments,
being the sound and speech o f this very material. In the role of
soprano or bass, it renders the ‘soul of the natural condition’ which
thus sings o f itself|A nqteqnthe recorder will never revea!, thew opd,
,and a trumpet note..will neyer. reveal the ‘souP of the metal, whereas
in the casé offthe lyricaí instrument o f th e body^ the voice is also
manifested as a revelanon of its soprano, álto, tenor or bass character
there on the stage. And the associations change in accordance with
the way an age views thé ’̂‘charáctefs of voices’. W ith M ozart the
tenor, for instance, representeS a rather gentle cantabile quality:
Tam ino has it, as does the less powerful Ottavio. The baritone voice,
in contrast, represents the very condition o f Don Giovanni, the
actual hero o f an amorous intrigue, as well as the condition of Count
Almaviva. Even Figaro’s own bass voice is readily reconciled with
Susanna the soubrette, whose youth he shares. But when ideas
changed in the nineteenth century, the tenor carne to the fore. With
Verdi, and more ‘victoriously’ still with Wagner, he became the
aurum p o ta b ile of youth and erotic power — heroic power in fact.
Blond hair, a sunny radiance, strange white gods and their fetishes
were now the prestige, as it were, o f the tenor instruments, the way
they looked, and they had a character not previously associated with
them. Today, this heroic resolve is fading again. The baritone is
144
MAGIC RATTLE, HUMAN HARP
regaining ground once more as the instrumental sound o f masculine
vigour, minus all cello-like smoothness, bland benevolenee or
acquiescence. This matches our aversión to a high imperious voice
expanding in song – even if the actor’s ñame is Radames. But none of
these changes would have occurred had it not been for the singer’s
representation o f himself, which affords a lover of operatic voices the
presence o f a distinct vocal object. Here the instrument’s association
with a character revealingly arrests, at any rate, the elusiveness which
music, even non-Rom antic music, still normally possesses. For vocal
trumpets, unlike the ones that you blow, really are kings. Indeed
W alther Stolzing’s high trumpet, which is a sheer paean to himself,
almost finishes — even without the m elos — by transforming the
awakened Mastersingers into a love nest’s attendant spirits. W hat
has replaced the archaic magic of the object, to some extent at least,
is a kind o f intrinsic magic o f the material, the singing love-poetry or
the instrumental live Ímpetus of an especially effusive display o f the
material itself, which is localised. T o be sure it is also this, for its own
part, which enables opera music to have its issue in a human world
portrayed on the stage and to transform the pan-pipe, or the harp,
into actors; these, however, now voice human m aterial. They are
serving the music above them – with the limiting ideal o f the entire
theatre itself as a magic flute. ,
145
The Paradox of
Timbre
Author(s): Cornelia Fales
Source: Ethnomusicology, Vol. 46, No. 1 (Winter, 2002), pp. 56-95
Published by: University of Illinois Press on behalf of Society for Ethnomusicology
Stable URL: https://www.jstor.org/stable/852808
Accessed: 30-01-2019 20:12 UTC
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.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
University of Illinois Press, Society for Ethnomusicology are collaborating with JSTOR
to digitize, preserve and extend access to Ethnomusicology
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
VOL. 46, No. 1 ETHNOMUSICOLOGY WINTER 2002
The Paradox of Timbre
CORNELIA FALES / University of California, Santa Barbara
Introduction
n 1950, Alan Merriam toured the Belgian Congo and Trust Territory of
Ruanda-Urundi to record music representing various ethnic groups of the
region. Among the 67 reels of tape that resulted are the earliest recordings
of Burundi Whispered Inanga or Inanga Cbucbotde, a genre of which only
one other substantial collection exists and few performers remain. For this,
Merriam’s tapes are priceless; but they are also memorable as an example
of how the simple act of recording immortalizes a notion of music that may
reflect the researcher more than the musicians represented.
Inanga Chuchotee consists of a whispered text, accompanied by the
inanga, a trough zither of eight strings. As is often the case with African
music, the Barundi assign primary importance to the vocal text, though its
real significance according to musicians, lies not so much in its meaning,
but in the whisper that articulates it and particularly in the effect of the
combined timbres of the noisy whisper and the inanga. Merriam’s tapes,
however, show a consistent tendency to position the microphone so close
to the inanga, that the text is often muffled and inaudible. It is difficult to
escape the conclusion that Merriam was more interested in virtuosic inan-
ga playing-in the accompaniment, that is-than he was in the whispered
vocals by which the Barundi define the genre.
In obscuring the central effect of Whispered Inanga, Merriam’s record-
ings of the music betray the subtle bias of what has come to be called “pitch-
centrism” or “timbre deafness,” a perceptual proclivity on the part of west-
ern listeners, including ethnomusicologists, to focus on melody in music
where the dominant parameter is timbre. Listeners from a culture where
pitch is governed by law while timbre is governed by taste, where musical
execution is judged correct or incorrect according to variations in pitch,
while variations in other parameters of music are judged pleasing or dis-
pleasing’-such listeners would be surprised and perhaps disoriented to
? 2002 by the Board of Trustees of the University of Illinois
56
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 57
find the opposite polarity in evaluations of Whispered Inanga. A perfor-
mance of Inanga is judged incorrect if the expected timbral effect is impre-
cisely executed, whereas wide deviations in pitch are considered ornamen-
tal, expressive or, if unsuccessful, in bad taste or inappropriate. Especially
in a genre where the instrument, inanga, retains its tuning with difficulty,
where the whispering voice is pitchless, and where the meter changes
according to patterns of long and short syllables in the text, timbre is the
single element that is a fairly stable and predictable standard of correctness.
I will return to Whispered Inanga later.
Timbre
In the last fifteen years, the citing of timbre as an important feature of
African music has taken on the same aura of banal truth that once charac-
terized the association of rhythm and African music. Both associations are
true for a great deal of African music, but unlike rhythm which continues
to attract scholarly attention, the role of timbre even now is often no sooner
mentioned than forgotten.2 As scholars, indeed as listeners, we have a
difficult time describing timbre. Though we can talk about it in large gen-
eralities, as though it were a conceptual abstraction–“timbre is important
to African music, community is important to African music”–it is only by
deliberate effort that we conceptualize it as a distinctly ongoing, dynamic
feature of music with the same clarity as pitch or meter. So to describe
Merriam’s misinterpretation of the Whispered Inanga aesthetic as timbre
deafness is imprecise. We have a peculiar amnesia3 in regard to timbre, but
we’re not deaf to timbre: we hear it, we use it-no one has much trouble
telling instruments apart-but we have no language to describe it.4 With
no domain-specific adjectives, timbre must be described in metaphor or by
analogy to other senses, and this is true in many, many languages of the
world.
Through the study of timbre runs a sort of paradox that I will summa-
rize here briefly before embarking on the more detailed discussion to fol-
low. One of the objectives of hearing, as of all the senses, is to furnish lis-
teners with accurate information about the environment. We will see that
the dimension of timbre is particularly implicated in achieving this objec-
tive: not only does timbre carry the most information about a source and
its location (Butler 1973), but of all parameters of music, it is also carries
the most information about the environment through which the sound has
traveled. Human experience shows that in most cases, acquisition of audi-
tory information is accomplished with little difficulty. But, as the technol-
ogy of digital sound analysis becomes more sophisticated and accessible,
increasing evidence accumulates that what we hear-the source we iden-
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
58 Ethnomusicology, Winter 2002
tify so easily-is often different, especially in timbre, than the sound that
digital analysis tells us was actually emitted. The difference is blatant be-
cause sound analysis can reveal only the physical characteristics of a sound,
and if it shows features different than those we perceive, we know that the
act of perceiving the sound has changed it. The paradox emerges with the
observation that while timbre is a dimension of central importance to iden-
tifying sources, it is also the dimension that is most divergent from the
sound in the physical world.
How can we reconcile the success with which we identify a source
with the fact that the sound we hear is demonstrably different in timbre
than the sound emitted by the source in question? The resolution of this
paradox is complex and will hopefully emerge in the remainder of this
paper, but for now, enough to say that auditory system does indeed iden-
tify sources, but it identifies a version of the sources that may not always
coincide with the version existing in the physical world. Instead, it per-
ceives sources according to its own expectations, sources that are consis-
tent with similar sources identified in the past, or that have characteristics
typical to an environment, though digital analysis might show them to be
completely anomalous by any measurable standard. The paradox exists
because however different the perceived version of a source might be to
the physical signal it represents, it is a version that works in our world, it
is a version that is consistent with versions of other listeners, it is a version
“real” enough that it allows us to deal with the physical environment. The
version is that source to us, and as long as it continues to work, we need
no other. That the paradox of timbre is rarely appreciated by ordinary lis-
teners is a sign that the system that processes sounds for perception is also
careful to keep its transformative operations outside of the awareness of
the listener.5 One of the proposals of this paper, however, is that in music
something happens to timbre that makes listeners aware of its paradox.
Timbre is a slippery concept and a slippery percept, perceptually mal-
leable and difficult to define in precisely arranged units. Though human
auditory acuity, on one level at least,6 is greater for the discrimination of
timbre than pitch, contrasts in pitch register on a conscious level more
immediately and starkly than timbre contrasts of equal magnitude. To the
general listener, pitch and loudness are variable characteristics of sound,
timbre is a condition; pitch and loudness are things a sound does, timbre
is what a sound is. Given that timbre is critical to human contact with the
environment and a sonic dimension we track with peculiar sensitivity, giv-
en that timbre is routinely cited as one of the four parameters of sound,
the fact that it attracts so little attention7 becomes itself part of the mys-
tery: timbre seems to do its considerable work with secretive discretion.
This paper proposes a theory of timbre that addresses its role in the
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 59
larger sensation of musical sound. The theory begins with recent research
in timbre perception and applies it to music recorded in various contexts.
While the most persuasive validation of a theory like the one presented here
would include a number of extensively analyzed case studies, the limits and
objectives of this paper argue for breadth rather than depth of evidence.
For broadly confirming evidence, therefore, I draw on musical phenome-
na widely distributed throughout ethnomusicology literature, returning to
my own work in Burundi for deeper analytical demonstration. If the result
is a bit anecdotal or less than ideally thorough, my aim is to roughly map
out a neglected area of the musical experience, establishing coordinates to
be filled in with richer ethnographic study.
Timbre and the Unconscious
An interesting and revealing exercise is to play a short excerpt of
monophonic instrumental music several times to listeners who are utterly
unfamiliar with the instrument they are hearing. If asked to describe the
sound and its source, most listeners are surprised at the quantity of specific
information they are able to deduce from tiny, immensely subtle details,
all without conscious effort or reflection. But though they can easily deter-
mine that the unknown sound is, for example, an impulsive rather than a
sustained tone,” they must be prodded for this information with questions
as to the actions (plucking, bowing, blowing, etc.) the musician might be
making to play the instrument. They can, in other words, more easily de-
scribe the production of the sound than the perceived features of the sound
that allow them knowledge of its production.
The point of the exercise is to demonstrate three aspects of auditory
cognition. First, ordinary listeners with no special training possess an ex-
traordinary amount of knowledge about sound and its sources. Second,
while some source characteristics are implied by the pitch range and in-
tensity of its sound, listeners generally seem to base their knowledge of a
source and especially of its location, largely on qualities of timbre-on the
abruptness of an attack, for example, or a brightness of sound indicating
resilient resonating material-so much so, in fact, that timbre comes to be
identified with its source. Third, the information listeners are able to de-
duce from a sound is derived from a multitude of unremarkable acoustic
details of which a great deal have been processed and interpreted preat-
tentively9-that is, without listeners’ conscious awareness. I propose in this
paper that the dimension of timbre in particular is preattentive both in
processing and in the qualities that result from that processing.
In what sense is timbre unconscious? If we are conscious of hearing
first one instrument and then another, are we not conscious of timbre?
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
60 Ethnomusicology, Winter 2002
While researchers continue to grapple with the facts that there seem to be
several kinds of consciousness and that it is a phenomenon of degrees, not
an all-or-none cognitive state, most seem to acknowledge a broad categor-
ical difference between consciousness as phenomenal experience and con-
sciousness as reflective self-awareness. Phenomenal consciousness might
include the warm sensation of the sun on one’s face, while reflective con-
sciousness would consist of the intellectual awareness of felt warmth, and
of the sun as the source of the sensation. Phenomenal consciousness is often
sensory or emotive, reflective consciousness is often verbal or representa-
tive-that is, one who is reflectively conscious can usually express the
nature of the experience. Though most often the two kinds of conscious-
ness occur together, they can occur individually as well.’o Without the viv-
idness of phenomenal consciousness, we must trust a less convincing reflec-
tive consciousness of our experience; without reflective consciousness, we
may be phenomenally conscious of an experience that is richly sentient,
but at the same time, hazy, ill-defined, and inexpressible. If indeed the
experience of timbre is preattentive, then it appears to be a case of phe-
nomenal but not reflective consciousness.
While relatively little of the research on consciousness has focused on
timbre perception specifically, work on auditory perception in general and
unconscious perception in other sensory modalities has been extensive,
yielding a set of characteristics common to both the process and the results
of preattentive perception. Many of these features-of which two are es-
pecially relevant to this discussion-are present in the experience of tim-
bre. For example, the hallmark characteristic of preattentive or unconscious
perception, a feature Merikle and Reingold call “indirect measure sensitiv-
ity” (1992), is obvious in the listening exercise above. Indirect measure
sensitivity alludes to the fact that preattentively processed information can
not be directly examined or evaluated, though it may be a source of input
to relevant problems. We may have difficulty describing, or even concep-
tualizing timbre as an independent musical parameter on the basis of di-
rect examination, but we use it easily to distinguish or characterize sounds.
A second characteristic of preattentive processing relevant to timbre
is the tendency of perceivers to attribute the effects of such processes to a
conscious phenomenon of outstanding perceptual salience, and to do so
with absolute conviction (Merikle & Daneman, 1998). For example, when
laboratory subjects are exposed to subliminal pictures of colored geomet-
ric shapes, then asked later to pick out the shapes they “saw” earlier, they
are unsuccessful, as predicted by the indirect measure sensitivity of uncon-
scious events. However, if the same subjects are asked to choose from
among several shapes those they find most attractive, not only do they
choose the shapes they were exposed to previously with a probability far
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 61
greater than chance, but they provide thorough, reasoned explanation for
their choice, insisting on factors such as a habitual preference for certain
shapes, colors, symmetries, etc. Their very knowledge of themselves is
influenced by the need to explain a choice whose real explanation has lit-
tle to do with aesthetic preference. Similarly, listeners may point to a change
in pitch or dynamic level in music to describe what is actually a variation
in timbre; it may be, in fact, that the disregard of timbre mentioned earlier
is as much the result of misattribution as true neglect.
Auditory Processing
With whatever degree of auditory unconsciousness, the fact remains
that listeners are amazingly and confidently good at deriving information
about the acoustic world from the perceived world it inspires. The distinc-
tion is important: the acoustic world is the physical environment where
sound as acoustic signal is produced and dispersed; the perceived world
is the subjective, sonic world created by listeners as a result of their trans-
lation of signals from the acoustic world. The acoustic world is available
through deduction and calculation, but never directly experienced. Sepa-
ration of the two domains is at least conceptually necessary, because as
noted earlier the correspondence between the two worlds, between sig-
nals from the acoustic environment and the auditory perception that results,
has proven to be far from exact. A given acoustic stimulus may not excite
the same percept in all listeners, nor even in the same listener across mul-
tiple exposures. Auditory perception more often than not depends on fac-
tors external to-sometimes in contradiction to-the acoustic stimulus that
provokes it. In this paper, I maintain that many of these factors are embod-
ied in the parameter of timbre.
Outside of a musical context, auditory perception has been shown to
be precisely geared to source identification. The goal of listening, at least
ostensibly, is to maintain an identity relationship between acoustic source
and perceived source. Research has shown that the auditory cortex-with
no help from the conscious mind which may know nothing about acous-
tics-uses hardwired information about the behavior of sound in the phys-
ical world to accomplish most lower- and much higher-level processing of
complex sound. The auditory world of a normal listener, it seems, is found-
ed on a canonical knowledge of sound and its sources, knowledge which
has to do largely with issues of timbre.
How does source-orientation shape the mechanics of sound process-
ing, and why is timbre the parameter most effected by this orientation?
From minute to minute, the auditory cortex is faced with the formidable
task of grouping by source, an immense collection of individual acoustic
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
62 Ethnomusicology, Winter 2002
elements-frequencies of specific amplitude and duration-each group to
be deciphered as a perceptual unit characterized by its mixture of ele-
ments.1″ A generally accepted thesis as to the brain’s organizational strate-
gy is that it uses what Bregman calls “heuristics”, competing clues or hy-
potheses regarding possible groupings, based largely on characteristics of
source behavior.12 Those clues whose combination describes the most likely
source are used to organize a set of stimuli into auditory percepts. Timbre,
of course, is a primary result of this grouping. In theory, all the informa-
tion needed to determine pitch and loudness for natural sounds is available
before the sound has even left the inner ear; timbre must wait until signals
from all elements reach the auditory cortex where they are grouped and
subjected to the process of perceptual fusion into the unitary sensation of
tone quality. Since the perceived timbre of a steadystate tone is determined
largely by the relative amplitudes of its constituent frequencies,’3 there is
no single property or component of an acoustic signal that corresponds to
that sensation; thus, perceived timbre exists in a very real sense only in the
mind of the listener, not in the objective world. It is here that the paradox
of timbre begins to show itself.
For obvious reasons, accurate and efficient source characterization is
an essential tool for the survival of the listener. But it is also vital to the
perceptual equilibrium of the listener. When Bregman uses the term “heu-
ristic” for the brain’s organizational strategies, he means it as computer
scientists use it to contrast with “algorithm”. An algorithm is a procedural
solution to a problem; faced with a problem, to discover the relevant algo-
rithm is to overcome the problem. A heuristic, on the other hand, is a hy-
pothesis; it is time-dependent, the best of several possible strategies in-
formed by current information to advance from one stage to the next in
progress toward a solution. Given the nature of human hearing as current-
ly understood, therefore, identification of a sound source is by definition
based on incomplete information. If we notice furthermore that auditory
processing as described above is profoundly tautological–that is, its ob-
jective is to discover source information, while using preexisting informa-
tion about sound sources to pursue its objective-we begin to understand
how truly precarious the perceived world of an ordinary listener can be.
Nevertheless, under unexceptional circumstances listeners are altogeth-
er unconscious of the heuristic nature of auditory perception, oblivious of
the acoustic world whose relation to the perceptual world they inhabit is
one of “best guess”. The consistency of perceived information is so vital
to the equilibrium of listeners, that they will report hearing missing pieces
of an acoustic pattern in order to maintain an auditory gestalt (Warren
1976). Similarly, when exposed to audio/visual stimuli constructed so that
what is heard conflicts with what is seen, subjects ingeniously and uncon-
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 63
sciously combine elements from both aural and visual stimuli to form a
percept in which they have absolute faith; the sensory conflict is resolved
with no more apparent effort than ordinary listening (McGurk and Mac-
Donald 1976). Confident of their auditory acuity, listeners feel themselves
directly and aurally linked to a source in the acoustic world. So strong is
the source orientation of environmental listening, that listeners project the
fundamental premise of their auditory logic onto the data it is meant to
interpret and subjective auditory sensation onto a world of sources until
sound equals source. We say-I hear a cricket; not-I hear a sound that
may indicate the presence of a cricket. As though to colonize the acoustic
world, we project even our reaction to a sound onto its imputed source;
we say-I hear a sad violin-when it is a sound we hear, and we who are
sad. Our entire auditory world, it seems, is based on the substitution of the
indexed for the indexical and the effect for the cause.
Perceptualization
As a kind of shorthand in this paper, I will use the word “perceptual-
ization” to mean any cognitive operation or feature that contributes to the
perceptual outcome of a signal beyond the actual acoustic elements of the
signal. If one conceives of the perceived world as shared responsibility, the
result of contributions from the acoustic world and the perceiver’s mind,
then perceptualization is the process by which necessary interpretive ele-
ments are identified, created, and combined with acoustic properties of the
environment to create auditory percepts. Theoretically, perceptualization
is a measurable quantity: the difference between the percept predicted in
consideration of its raw acoustic input and the percept actually experienced
by a listener-or, the divergence of the perceived sound from the acous-
tic signal-is a measure of perceptualization. For example, to the extent
that there exists no component of an acoustic signal that by itself provokes
or corresponds to the sensation of timbre, timbre can be said to be less
predictable from a consideration of an unprocessed acoustic signal, and thus
more highly perceptualized, than the parameters of either pitch or loud-
ness. For a number of reasons, timbre appears to be the parameter most
frequently and intensely implicated in perceptualization. First as perceived
quality, it is already the most perceptualized, therefore the most mallea-
ble parameter of sound. Second, it is the parameter that carries the most
information about sound sources, and thus accrues the most benefit from
various perceptualizing operations. And third, it is reflectively unconscious
to the listener, thus representing a domain in which perceptualization can
work most invisibly.
Experimental work on auditory illusion using stimuli created to provoke
perceptualization has uncovered many of the strategies and priorities by
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
64 Ethnomusicology, Winter 2002
which it functions. Perceptualization appears to act according to consistent
and transparent logic, largely-but not entirely-geared to promoting au-
ditory efficiency. For example, a common laboratory illusion is created when
a listener’s perceptualization assumes incorrectly that it is confronted with
a degraded signal, which it efficiently attempts to correct by filling in or re-
storing “missing” segments of sound where no acoustic stimulus exists (War-
ren 1976). A great deal of perceptualizing energy is directed toward remov-
ing inconsistencies, enhancing or subduing sounds according to their
importance, and arranging other alterations that ensure a confidence-inspir-
ing transformation of the acoustic into the perceptual world. One of the iro-
nies of perceptualization is that many of its functions are geared to convinc-
ing the listener that it doesn’t exist, that signals from the acoustic world
literally are the percepts they provoke. In this sense, perceptualization is
the guardian of sensory equilibrium, a state that depends as much on a lis-
tener’s obliviousness to the acoustic world as on the evasive tactics of per-
ceptualization. Thus a subset of perceptualized effects seem to break with
the logic of auditory economy, to be directed instead toward camouflaging
the operations of perceptualization from the listener. Since the creation of
some of these effects consume considerable cognitive resources, their pres-
ence in situations where they contribute little to perceptual success sug-
gests that the perceptualizing mind has an interest in self-concealment only
slightly less urgent than its mandate to facilitate audition.
For our purposes here, a particularly important characteristic of per-
ceptualization concerns its relative potency and durability in different con-
texts. If the role of perceptualization generally is to effect some useful dif-
ference between auditory percept and acoustic signal, then its strength can
be measured by the degree to which it successfully overcomes the natural
characteristics of the acoustic world in constructing a perceived world. In
this sense, the power of perceptualization is inversely proportional to the
feature intensity of the acoustic signal. For example, a common efficiency
mechanism of perceptualization is the removal, from a listener’s auditory
field, of sounds in the environment-the buzzing hum of neon lights, for
example-that are continuous and unchanging. In most cases the buzz of
neon lights is dynamically soft enough that it is easily eliminated as percept;
but the more intense or louder the neon buzz, the less successful the op-
erations necessary to remove it.
Another factor influencing the robustness of perceptualization is its
relative reliance on acoustic information in creating its effects. A compar-
ison of effects suggests that perceptualization seems to create the strongest
effects when it can work from within the perceived world, avoiding as
much as possible any reference to the acoustic world. The perceptualized
filling in of the degraded signal described above requires intense and on-
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 65
going attention to the acoustic signal in order to restore a sound that is
precisely consistent with the existing sound. The result is an illusion that
is striking but relatively fragile and easily disrupted. On the other hand,
perceptualizing operations that allow listeners to hear the sound of their
own names standing out clearly and distinctly from an undifferentiated mass
of sound (a large crowd all conversing, for example) that otherwise defeats
audibility-these operations create a more robust effect, one that persists
even when listeners are informed of its illusory quality. The durability of
the effect depends on the special significance to listeners of their own
names, to which a correspondingly heightened sensitivity is an integral part
of the very structure of their perceived worlds. Such sensitivity is activat-
ed automatically by the sound of the name without particular reference to
its acoustic context. Thus, unlike the restored sound in the first example
whose perceptualized production is highly dependent on the existing
acoustic signal, the name-sensitivity of the second example allows the per-
ceptual amplification of a name with only the smallest assistance from the
acoustic signal to which it responds. At an extreme, the effects of percep-
tualization may be created almost entirely from the cognitive substance of
listeners’ minds-that is, from pre-existing information or structures in the
perceived world-but they must implicate the acoustic world to some
degree at least-else auditory perception becomes auditory hallucination.
Music
This is the process as it occurs on the part of unimpaired listeners in
unexceptional circumstances. But substantial evidence suggests that the
experience of music is something other than unexceptional. That there are
different modes of listening is a fact long recognized by auditory scientists.
More recent research in brain laterialization (Liegeois-Chauvel et al. 1998;
Peretz 1988) has found evidence suggesting that the mode applied to mu-
sical listening is more flexible-both procedurally and in perceptual result-
than the modes applied to speech and certain environmental sounds. It
appears that part of the difference between musical and other modes of
listening resides in the degree to which source characteristics are ignored
during the regrouping process in favor of other schemes of perceptualiza-
tion. Depending on the skill of the musician and the willingness of listen-
ers, such subversion may disrupt perceptual equilibrium profoundly or
fleetingly, but disruption is inevitable. An overview of traditional perfor-
mance styles suggests two broad categories of timbre manipulation that I
call respectively timbral anomaly and timbre juxtaposition. Within the
category of timbral anomaly, there are two techniques that I will call tim-
bre anomaly by extraction and timbre anomaly by redistribution.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
66 Ethnomusicology, Winter 2002
Timbral Anomaly
The creation of timbre anomalies generally involves a disorder of per-
ceptual grouping. We observed earlier that the perceptual grouping of dis-
crete acoustic elements in the cortex occurs according to heuristic logic
based on the nature of sources in the acoustic world. The success of a tim-
bral anomaly depends on the ability of a non-sourcebased organizational
schema to supplant the natural tendency to organize perceptually by
source.'” Thus timbre manipulation works directly contrary to the efforts
of perceptualization. There are two techniques for creating timbral anom-
alies in music (schematicized in Figures 1 and 2), differing in the methods
by which they incite reorganization. The creation of timbral anomaly both
by extraction and by redistribution occurs by changing the relationship of
individual components of a tone to the timbral group of which they are a
part. The extraction technique leaves the global organization of the origi-
nal timbre in tact, while emphasizing one or a small group of those elements
to stand out in relief against the remainder of the tone. The redistribution
technique takes the global organization of a tone’s timbre, and breaks it
apart so that a subset of its elements are free to join or form other groups,
each to be perceived as a separate timbre.
Timbre Anomaly by Extraction
The extraction method is best typified by overtone singing or didjeri-
doo music. In both overtone singing (Figure 3), and didjeridoo music (Fig-
Figure 1: Timbre change by EXTRACTION, presented schematically in vi-
sual domain. The diagonal line visible in timbre two is also present in tim-
bre one, though not immediately visible until emphasized in timbre two.
Notice in particular that the technique of extraction has involved a physi-
cal change to the constitute elements.
~’~4? ~F?’~
~~~es~~2 ~c t~ :;: ~p
4:: ~ ?t?~~o ~ ~ ?I?~
Timbre One Timbre Two
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 67
Figure 2: Timbre change by REDISTRIBUTION, presented schematically in
visual domain. The constituent components of timbre one can be seen as
a single grouped unit; in timbre two, the presence of additional timbres has
caused two subsets of the larger group to redistribute into groups with new
timbres. Notice in particular that the original constituent elements have not
physically changed; technique of redistribution causes only the viewer’s
perception to change.
Timbe Tw 9
@2E3 ?
s~~~~p ~ @ @i @8 5IE~~~ ~ jj
~~~a~ ~ ?IE @~ @ 4~~~~~IZ~ s e~ B
@~ ~ @ @@E @@ @@ @@ ?~3 @k~ @i~ @@? @ @ @
3@ @@@ @@ ;SG @ @@S j@ ?zt @If~ @@@ @@ @Ff 3@ @~ ?3 @s 3@5 ~~~ J
@ia @~ @3 @3 @i @Ti i@~ j@ @~ @ @38 @ @@@@
@s @b @E @it @i @i~ @ @i ?i @k @s @ @ @@
@c @~5 ?7 3L@ @2 @i @ 235 @3~ @F @5 @ @ @@@
@i x@f 3@ j@3 @~F @a j@t @ @L @9~ @ @3 @s @@ @s @ @ 1@ @Z3 @~ @ld j@~ @
iss @4 @ ic @~~t @8~B @3 @ @~3 @1 i@ j@3 @a @@E @e @s @ @F @3 @ @ S@ ?
@~ @~ ?~ *2@ @a @~i j@ 4@5 @~ @ ‘I @ @ @@@
Timbre One Timbre Two @ @@ ~~
ure 4), a subset of the harmonics comprising a tone’s timbre is amplified
to a point of sufficiently greater intensity than surrounding harmonics, that
they stand out in relief from the remainder of the tone’s timbre. In the case
of overtone singing, a very few harmonics are extracted by amplification,
provoking the sensation of a second pitch sounding above the drone of the
primary tone. In didjeridoo music, the coupled resonators of the musician’s
mouth and the tube of the instrument permit the extraction of an entire
formant-a “clump” of harmonics-that moves in frequency relative to the
drone of the instrument to give the sensation of two timbres moving against
each other.”I
It is important to note that timbre effects like those demonstrated by
both kinds of music often result in percepts that would not otherwise oc-
cur in the acoustic world. Both produce multiple tones from what are or-
dinarily single-phonation instruments. In overtone singing, furthermore, it
often happens that the successive harmonics producing the melody, con-
stitute a series of sinetones-single frequencies without harmonics of their
own which are ordinarily produced only synthetically. The didjeridoo of-
fers one of a very few occasions to hear frequentially-configured timbre”6
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
68 Ethnomusicology, Winter 2002
Figure 3: Spectrogram of Tuvan overtone singing. Horizontal axis is time;
vertical axis is frequency; amplitude is indicated by greyscale (darker is
higher amplitude). Note that harmonics are relatively constant in frequen-
cy, except for a pronounced vibrato, indicating little fundamental pitch
variation, but that amplitude of emphasized harmonics varies intensely in
order to create overtone melody.
Ow::.”.,
without pitch, since the moving formant is made up of the same harmon-
ics as the primary tone but lacks its own fundamental (and the harmonics
that would go with it), and thus has no discernible pitch. When acoustic
elements like pitched overtones or the didjeridoo formant break free from
the perceptual fusion of timbre, they lose a degree of perceptualization. To
“hear out” a high-frequency harmonic is to reverse the auditory process,
including whatever contribution has been made by the listener’s own per-
ceptualization. Ordinarily, loose harmonics are inaccessible to us, since they
only ever make an appearance as part of timbre.17 Released from the effects
of perceptualization and untempered by the refining influence of the lis-
tener’s mind, these elements must stand on their own, their singular char-
acter in the perceptual world nearly identical to-or at least, directly dic-
tated by-their character in the acoustic world.
Timbre Anomaly by Redistribution
The anomalies above result from the extraction of a harmonic or for-
mant from a timbral gestalt. Other forms of timbre-shaping, however, re-
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 69
Figure 4: Spectrogram of didjeridoo tone. Horizontal axis is time; vertical
axis is frequency; amplitude is indicated by greyscale (darker is higher
amplitude). Note that harmonics are relatively constant in frequency, while
formant moves rapidly over time.
i: ….
, :-:?.:;i-i– , .i. ; : -i :: ………….:::: i::iiii ” : …….. ?”…
S .::– . . . . : i ?. . I i ..
::-:::-?:::
distribute a subset of gestalt components to new groups whose member-
ship may be determined by a nonsource-based organization. Such a redis-
tribution creates an anomalous sound, an example of what Bregman (1994)
has called a “chimeric” sound, a percept made up of components from
several sources with little resemblance to any of them; or one might call it
an example of “inherent timbre” by analogy to G. Kubik’s notions of inher-
ent melodies and rhythm (Kubik 1989). Examples of these kinds of effects
are plentiful. The Ghanaian balafon, for example, produces tones in a bass
register whose harmonics fuse with the long-resonating noise of a mirla-
ton, to produce a pitched, buzzy drone that transforms a percussive sound
into a sound of sustained resonance (Figure 5). Another kind of chimeric
sound results from two simultaneous tones of a steel pan whose inharmonic
partials combine to provoke the percept of a third tone of different pitch
(Fenn 1997). Another example described by Gage Averill (1999) is the bell
tone in barbershop singing, which appears to result from the blending of
a few harmonics from each of the singers. Finally, in this category is Bu-
rundi Inanga Chuchot6e (Figure 6), a description of which follows short-
ly; for now, we note it as a genre that rests on an illusive melodicized
whisper-the fusion of inanga harmonics with whispered noise.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
70 Ethnomusicology, Winter 2002
Figure 5: Spectrogram of Ghanaian balofon tones, showing rapid, percus-
sive strikes of the lamels, overlaid with the sustained buzzing of mirlaton-
produced noise.
i:~::_:::
-,s~’s-“-~~(?~”~s –::-:::,::
~i:: ::;.:: ::: I: ::: -: -‘ : -_~-?-:: i-i:
–::-?-:i?- .-: ::: :::::: i :
.-:. ?:: :: ::-::~i:_ : ii?-:: ::: :: :-‘:–?- i:r;:ii~ii:i:cil ?:: : :
:~-~;-~~—-:~~- :-~I::–:
.,.
::
:…:- ai”.~~ii:’ ::i~i~iii:i:::;:::
:.l–ii_ -:_-Rli–i-i’: ..:: -iiiiiiii-ii-:_:iiij-::i:- -:T-: (::~..i li:-~(/ ~ ::–:: -::: : j:?i:/i:–:j.i:– _
:::-?:: : ::;:: -:::::: ‘i.i4i~i~:i:i:i- -li -?Qi–?.::::_:l -:~?,:,::?-::-:iii—‘1 :iidi:i,::i,:- :-,,:-_:iw-r;i :-:-?-,.::i?:i:?ii; ::::-::::: : : :’i”i:;,li:iai::–:-:i :::::::?::-:::::: :i: . :?~ril:liiiii :.:,- :::1:; _ ‘:’i.j:_i:i:: :; ii:?-:iiiii::~-?a- : : :-n-i- :i5:
.i’;,iil’ai:~iii_ :::i-:I 1:;::~:,:_– :::i,: :-i-li:i’-~:-:-i~~~: :~~i”:~:’-;:- ::–? :i”:jp?:::::: :’-:::-i-:-:
,: :r:~:: :::;::i- :::I-:: : :: ?i::?::–:1:I:::; : : :-C : :
:i-:::?!_:;:~~–~i:-i:r~~-laiii?:-‘li-: :: :-:–?a-?iiiii-L
.- ::::: :: :::::–~:;:? ? ?? i~ i::i~:?i:i?ii :_~-i?i::?.ii ~ji:i ? i:.,:i-~’i~rii~i _ :. ‘-“-‘- ” -::’:-~:
?:l-iii-i?::-i– :,-iie_:i :- -:?:::-_ii-iliiiii::ili;i:?-:ii?:_
:::
iii i?i:i:-i–:: -: a~i?i:i:iiiii?: -: :: i-i”~~i-ii-i- – -:-i
i??.”L’ii- j:~
i::i:-:M,~:-:r:::: ::~::-:i i:_~~9~~:l:~::?::: .:j__–I~R”;
:i-:::: :’::;:’:I:::: :_ : _:_ ::: ::_- :i:- :: :::- : :-:-::*: :::: :’i-:::_i’:ii:_-:-ii:l-i:-.r;::?::: : :::,:-::::?::: -i.iii :’i__::i’- ii- i–i:-i:i: i:: i ::::::- :i- –:?: ‘?-i:-i
… ii iiiiiiii,_-_-i:ii-iii- :.:: -::::::::: -:ii:ii_-:;-i
::;:::-:::::-:; i–:_,:-::? -. r:’—“-‘::k’:’i-M:.j- :1-~ 91:~?~.U~::~-iBii-‘:i?i~ii::~i ii?::~:-i?—_:i ?:::
~:j~: :?:~~-~:: ~~::: ::~-lil~:
-:I_:::~::
I-‘””::’::~: : :
.::: :: :: :::: :::::::i?::a::-1:_: -:~-: i ;-_?_ i: : i:ti:: ::?::i
:_ –::::: _:::~:::: ; :i?:::::i-i:8`i:~:-:.:- ?~-:: -:?::: -:I: -?il-:”-‘~-::::: :ii_?— :: ::-:: :: -::: -:iri-ii – i?i–
;i-i–i?*in-~::I
An important difference between the extraction and redistribution
methods of anomaly is the degree to which they use acoustic factors to
incite perceptual reorganization. As schematicized in Figures 3 and 4, anom-
aly by extraction requires an acoustic change in the component to be ex-
tracted, whereas anomaly by redistribution occurs with no change at all to
the original sound. Thus, while overtone singing requires that the extract-
ed overtones be physically amplified until they are heard out as a separate
pitch, the melodicized whisper of Burundi Inanga Chuchot&e requires no
acoustic change in the sounds from which the components are contribut-
ed; whatever change occurs, occurs solely in the mind of the perceiver.
From the point of view of musicians, the creation of a timbral anomaly in
overtone singing requires that they change the sound; the creation of a tim-
bral anomaly in Inanga Chuchot&e requires that they change the listeners,
inducing them to change their mode of perception. Thus a performer’s
control over the effect of a timbral anomaly is correspondingly greater with
the extraction than with the redistribution technique; extraction of rele-
vant harmonics is largely a matter of musical skill, redistribution of acous-
tic elements is more a matter of musical persuasion.18 If indeed timbre
manipulation works by defeating perceptualization, and if perceptualization
is strong to the extent that it avoids the acoustic world, then the acoustic
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 71
Figure 6: Spectrogram of an excerpt from Inanga Chuchotee. The first five
broad vertical stripes are the attacks of the inanga playing alone; the last four
vertical are the inanga together with the noisy whisper of the musician.
!;JX ::= . ? ::: : : . :i~i:Fd ::<::: :: i:: i l:i::-i:s:::-:. .. . .
? 4 i:::;: i .:ili
;?”IT
… ….. ………… . ..iiii
i: i } :!!!i – – :
changes required by the extraction method give it a potent weapon against
the forces of perceptualization.
What is the effect of timbral anomalies? Shaped by a skillful musician,
the borders of a timbral unit become fluid, provoking a different percept
with each context. Among other things, techniques such as these disrupt
perceptual complacency, presenting anomalous sounds that subvert a lis-
tener’s instinctive knowledge of source behavior, creating instead a pro-
found if momentary disorientation. Whether, for example, the absolute
certainty that a human vocal tract produces only one pitched sound at a
time is innate or learned early in life, the disruption of that certainty with
first exposure to overtone singing must constitute a brief violence to a lis-
tener’s faith in perceptual constancy. For all the drama of these methods
of timbre shifting, however, one suspects that they quickly decline in effi-
cacy with listeners’ exposure, as anomalous sounds solidify into generic
attributes or become typical of new and exotic sound sources.
Timbre Juxtaposition
A second category of timbre manipulation, produces an effect that is
more subtle but perhaps less easily dulled over time. Here, the musician
juxtaposes sounds that fall on opposing ends of a continuum of timbral
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
72 Ethnomusicology, Winter 2002
structure. Evidence from digital analysis of a broad range of musical tones
suggests a timbre continuum extending from sounds that are harmonical-
ly-based to those that are formant-structured (Figure 7). A formant-struc-
tured timbre is one whose spectrum consists of several formants, broad
bands or clumps of high intensity harmonics similar to those described
earlier in didjeridoo music. Generally speaking, the softer the material from
which an instrument’s resonator is constructed, the broader the spectral
peaks or formants of its timbre. In formant-structured timbres, the location
and size of the formants determine the timbre of a sound. In harmonically-
structured sound, timbre is dominated by several primary harmonics that
characterize the tone’s timbre. In the world of acoustic sources, the epito-
me of formant-structured timbre is demonstrated in vocal sound, charac-
terized by formants whose position determines vowel quality. At the oth-
er end of the timbre continuum, a typical harmonically-structured sound
might resemble a flute tone in a high register, whose audible spectrum may
consist of as few as two harmonics.
The difference between harmonically-structured and formant-based
timbres centers on the unit of perceptual salience-that is, on whether
single harmonics or entire formants dictate timbral quality. In the process-
ing of harmonically-structured sounds, source information must be derived
from the relative characteristics of individual harmonics, whereas in for-
mant-based timbres, source information is contained in the relative inten-
sity and position of formant units. In harmonically-structured timbres, each
harmonic contributes individually to the quality of the tone, whereas in
formant-based timbres, entire formants constitute a single contribution to
the tone quality.”9 An important distinction between the two kinds of
sounds lies in the differing degrees of perceptualization inherent in each
sound. Given that the two sounds imply different units of perceptual sa-
Figure 7: Schematic representation of spectra of two kinds of timbre, rep-
resenting two ends of timbre continuum. In the formant-structured sound
three formants are shown outlined with a dotted line and numbered.
F1
….. ….. F 2
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 73
lience, formant-based timbres can be said to require a greater magnitude
ofperceptualization than harmonically-structured timbres. A formant con-
sists of harmonics that have already been perceptualized into a salient unit,
whereas harmonically-structured timbres have undergone no prior percep-
tualization when they are examined for grouping and source characteris-
tics. In a sense, then, harmonically-structured timbres are purer, less “con-
taminated” by perceptualization, and closer in nature to the raw acoustic
form in which they were emitted.
Examples of timbre juxtaposition are as plentiful as those of timbral
anomaly. We have already seen one such contrast: A review of the over-
tone singing represented in Figure 5 reveals that the music actually employs
both techniques of timbre manipulation. Not only does the resulting sound
constitute a timbral anomaly, but it also juxtaposes the quintessential for-
mant-structured sound with the quintessential harmonically-structured
sound by taking from the vowel-formant timbre a single harmonic, a sine-
tone, to be perceived as pitch. Thus the musician produces the anomaly
of multiple pitches from a single-voiced instrument, while simultaneously
including both ends of the timbre continuum in his vocalization. While
overtone singing is remarkable in using both timbre techniques in a single
sound, its real power lies in the fact that it encompasses distant points on
the timbre continuum in the same sound. A classic example of juxtaposi-
tion over time occurs in Indian sitar music with the ornament of “minde”
by which a string is bent on its way to a target pitch. Like many stringed
instruments, the sitar produces a timbre that is typically formant-based;
regular use of minde, however, produces a sudden prominent harmonic that
fleetingly rearranges the instrument’s timbre into a harmonic-structure (Fig-
ures 8a and b), an effect that appears to be constant in spite of consider-
able variability in the timbral quality of sitars overall.20
Perceived Timbre Juxtapositions
The implication of the timbre continuum is this: A juxtaposition of
formant- and harmonically-structured timbres entails a redistribution of
perceptualization. If as noted above, a harmonically-structured sound is
shaped by a less intense perceptualizing influence, then it is more easily
predicted from a consideration of its acoustic components-that is, it is
more similar to the signal which it represents-than is a formant-structured
sound. A sound’s position on the timbre continuum, then, reflects not only
magnitude of perceptualization, but the relative similarity of the perceived
sound to the acoustic signal that provokes it, or the proximity of the
sound-and by extension, of the listener-to the acoustic world.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
74 Ethnomusicology, Winter 2002
Figure 8a: Spectrogram of sitar ornament, minde.
.8i- i ~ -::_:: :;:: i-ii i:-:i– :
:ii
.. :::::.i:: ::, i–~I::~:: .. ~I:; :: iZI -:i.-i-i :ii::i:::
:-:- -i-
: —
?i~?~?~u?~–~:i ~~~i;”?:: 6-ir—-~-s~~
: :
:::: -:1
~~: ~ i: i :. i:::-: i :iii: : a- :
i:: i? ::-?: ?::: ;:: -: :::::;i–:.:::
: ::_-:cai—-:-:::::
:: i:
~a a: i;i:
:: i:: ::: -: .. : E.::: i?
::.. :i: -lz -_ —
?~ :,,:,::~:~::,:i~
~-8L’i~
‘:”‘ :
.. ~:ii:i ‘-~iiii i~iirir,_l~~::i~ … : : :::::_s:~ii9I:~i,_~:i`-il:dii
?ii~: ~~ :6-;:~: ::::~.:::;~-: ~~-~:~i 3:2~1-:
i?’ti
‘li-
-‘::-i?::: ::: ::
??:: :::::::::::r:: -:a~i-Ei: :::~: :::_:_i:::::r_: ;:–$-:i- :::-:-::::::: -~,:`i-:i:id-:-~’ia—-_:- :: ?-~;~?
An Analogy, With Apologies to Plato’s Cave
Imagine sitting in a spot for most of a lifetime at a precise distance from
an impermeable screen behind which exists a glowing source of constant
warmth. If you never moved from the sitting position, you might wonder
about domains beyond your awareness, but you would have little suspicion
of a defined region containing a heat source behind the screen. However,
if you were to move forward and backward relative to the screen, you
would feel a temperature fluctuation with changing distance from the
screen. With no other information, you would remain oblivious of the glow-
ing source behind the screen; but you would know that something on the
invisible side of the screen emits warmth, since its effects on the visible
side of the screen-as measured by your sensory thermometer-alters with
proximity to the screen and the invisible world behind it. And by means
of your changing reaction to this “thing” behind the screen, you would have
ongoing indication of a domain where the thing resided.
In the same way, the skillful arrangement of contrasting timbres moves
listeners to different perceptual positions relative to the acoustic world,
positions that reflect auditory processes differing in quality and degree, and
resulting in different sensory reactions. Part of the experience of listening
in these situations extends beyond the sounds themselves, to an awareness
of differences in timbral processing, in degree of perceptualization and
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 75
Figure 8b: Spectral slice of sitar tone during minde, at point marked by
arrow in spectrogram 10a.
……… ……———————……………——————–… .. ………….. …………. ….
proximity to the acoustic world. Because this awareness occurs in the pass-
ing of musical time, it is most likely nonspecific. Just as the moving sub-
ject in the analogy remains unaware of the glowing source, so also musi-
cal listeners receive no information from processed sounds as to the nature
or contents of the acoustic world, but only the awareness of a realm re-
vealed in the changing magnitude of processing it requires. Though subtle
and brief, the moment offers listeners phenomenal evidence that the world
is a facsimile, a humanized rendering-as though perception were noth-
ing more than a projection of features onto a domain whose real character
is indecipherable. In this moment, the paradox of timbre is revealed.
Implications : Timbre Manipulation v. Perceptualization
We have already observed that a major difference between the broad
categories of timbral anomaly and juxtaposition lies in the relative durabil-
ity of their effects: timbral anomaly can be startlingly dramatic, but is quickly
absorbed into a listener’s repertoire of familiar sounds; timbre juxtaposi-
tion, on the other hand, is ongoing, cumulative, and quietly relentless in
its assault on the barriers between the perceptual and acoustic worlds. As
noted above, timbre manipulation in all its forms stands in direct opposi-
tion to the efforts of perceptualization whose function is nothing if not to
shore up the walls enclosing a listener’s perceived world. The difference
in potency of the various timbre techniques, then, can be understood as
an ongoing tally of wins and losses in the struggle between perceptualiza-
tion and musical timbre manipulation. A comparison of the three methods
of timbre manipulation is represented in Figure 9. Notice particularly the
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Figure 9: Comparison of techniques of timbre manipulation.
TIMBRE ANOMALY TIMBRE JUXTAPOSITION
By extraction By redistribution
Examples Overtone singing; didjeridoo; mouthbow. Steel pan tones; Whispered Inanga. Panpipes; sitar minde.
Effect One or more harmonics extracted from Timbre broken apart, elements Juxtaposition of formanted and
timbre and heard out separately. redistributed to other groupings. harmonically-based timbres.
Result Perception of an extra sound in absence Perception of sourceless sound. Rapid changes of degree/kind of
of extra source. perceptualization.
Role of Musician Alters musical sound to emphasize Induces listeners to perceive sound in Alternates timbres appropriately
relevant elements (acoustic & non-source mode (perceived world). (acoustic world).
perceived worlds).
Reliance on acoustics Moderate to intense. Slight. Intense.
Robustness of effect Moderate to intense. Depends on listener. Intense.
Listener susceptibility Moderate to intense. Controllable by listener. Intense.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 77
relationship between degree of reliance on acoustic factors in creating tim-
bre effects and the strength of the effect.
If, as described above, auditory perception generally is the result of
acoustic stimulus plus perceptualization, then timbre anomaly by extrac-
tion works on the acoustic stimulus, timbral anomaly by redistribution
works on perceptualization, and timbre juxtaposition works by revealing
the interplay between the two. What we find is that the three techniques
of timbre manipulation-timbre anomaly by redistribution, anomaly by
extraction, and timbre juxtaposition-form a scale of progressively more
effective control of perceptualization as reliance on the acoustic world
increases. If the technique of extraction relies on the acoustic world to a
greater extent than the technique of redistribution, the technique of tim-
bre juxtaposition is totally centered in the acoustic world. And if the mu-
sician who employs the extraction technique is therefore more in control
of its success than the musician employing the redistribution technique, a
performer who uses the juxtaposition technique has even greater power
over its success. Put another way, a listener exposed to all three methods
of timbre manipulation has progressively little power to resist the effect as
its production moves from the redistribution to the juxtaposition tech-
niques. In terms of the musical effect of the different techniques, the abil-
ity of each to disrupt perceptual equilibrium, to fracture the listener’s per-
ceived world becomes more intense and predictable with movement from
juxtaposition to extraction to redistribution.
The theory presented here proposes that when musical timbre is han-
dled in specific ways, it works against the ongoing efforts of perceptual-
ization to maintain a seamless perceived world. If it cannot show us the
gross discrepancies between the perceived and acoustic worlds by present-
ing us with timbral anomalies, fictional sources, and proof that the same
acoustic signal yields different percepts from one moment to the next, then
timbre forces an awareness of perceptualization as it engineers those dis-
crepancies, then hides all evidence of its work. It would be foolish, of
course, to suggest that every time listeners experience the right combina-
tion of timbres in music, suddenly they understand all the principles of
acoustics. Like many auditory phenomena, timbre effects pass swiftly and
with subterranean impact. If indeed timbre is a parameter of sound that is
unconsciously processed and experienced, listeners at a concert will not
be conscious of each perceptual tremor delivered through their ears. Rather,
I suggest that the very subliminality of these sensations is a clue to their
power. If one recalls the immunity of unconscious phenomena to direct
examination and the inevitable misattribution of unconscious effects, if one
recalls that especially reflectively-unconscious sensation may lack phenom-
enal definition, it is not surprising that listeners might feel the power of
music without ever considering its timbral basis.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
78 Ethnomusicology, Winter 2002
Instead, the cumulative effect of a musical experience of dramatic,
subtle, sustained, or sporadic timbre manipulation–the flashing into view
of the acoustic world-may move listeners with a range of specificity and
emotion: from a more or less vague sense of perceiving something normally
imperceivable, to an unaccountable feeling of transcendence or separation
from the earthly world (of sources), to a more general notion that music is
an imperfect translation of something, to the apprehension of something
absent or lost of which a hazy essence is returned through music. It may
be, that is, that the inaccessible effects of musical timbre elicit a kind of
rorschach subjectivity: detached from the mooring of perceptual constan-
cy, listeners make sense of their experience in whatever expressive cur-
rency is around.
Inanga and Kubandwa
If the effects of timbre manipulation are phenomenal but not reflective-
ly conscious, then one should not expect musicians to be able to describe
their use of timbre, nor listeners their response. But it would be entirely
consistent with what is known about unconscious processes to interpret
patterns of timbre variation as indirect evidence that musicians are aware
of and using timbre effects strategically, though they may be conscious only
of choosing sounds that seem to “work” better in one context than anoth-
er. A comparison of three kinds of Burundi song that use the inanga as
accompaniment will reveal timbral patterns that suggest strategic variation
and demonstrate the characteristics of timbre manipulation discussed
above. Two of the genres that use inanga also employ respectively the tech-
niques of timbral anomaly and timbre juxtaposition, so that their compar-
ison will reveal primary differences between the techniques and typical
contexts of use.
Voiced Inanga
The first kind of song, called simply Inanga (I will call it “Voiced Inan-
ga” here to distinguish it from Whispered Inanga) consists of texts from
traditional declamatory styles of poetry sung to an inanga accompaniment.
These texts are sung in full voice, just as they are when sung to other in-
struments like the umuduri (musical bow) or the sanza (mbira). As a musi-
cal form, this kind of inanga music seems to constitute a catchall category,
consisting of songs sometimes sung a cappella or to other instruments, by
men, women, and children, in no particular context, and for no particular
audience; in common parlance, the category includes virtually any song
sung to the inanga with the exception of the music specifically classified
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 79
under another genre of inanga music. Lacking any generic homogeneity,
Voiced Inanga is too unwieldy for exhaustive description here. Of impor-
tance to the argument however, is the quality of the combined timbres of
the instrument and singing voice (Figure 10). All things being equal, Voiced
Inanga shows no generic specification for timbre effects. As seen in the
figure below, both the voice and the instrument with its soft-wood resona-
tor are formanted. In view of its generalized usage, Voiced Inanga is useful
as a kind of default or baseline timbre for comparison with the other two
genres of inanga music.
Whispered Inanga
If Voiced Inanga is a heterogeneous genre, Inanga Chuchotee is the
opposite. The music was introduced in the beginning of this paper, men-
tioned again in the section on timbral anomaly by redistribution, and I have
dealt with its special qualities at length elsewhere (Fales 1995a, 1995b,
1998). Relevant to discussion here is Whispered Inanga’s manipulation of
timbre to create a pronounced anomaly, an auditory illusion around which
performance of the music is shaped. The essential effect is this: The text
of the genre is whispered in the tonal language of Kirundi. Because a whis-
per is acoustically pitchless, the vocal component of Whispered Inanga is
limited in two ways: it cannot produce a melody, and it cannot produce
the linguistic tones that distinguish grammatical or lexical features. Never-
theless Burundi listeners claim with absolute conviction that the musician
“sings” to the melody of the inanga; the musician’s voice must “meet” the
strings of the inanga by going up and down in pitch in coordination with
the melody of the accompaniment. In regard to the tonal contours of the
language, listeners point out that if the voice failed to go up and down with
the inanga, the singer would be “mispronouncing” or “speaking wrong” the
text. Part of the illusion of Whispered Inanga depends on features of the
genre that function to link the instrumental and vocal parts of the song by
giving linguistic functions to the instrument, and musical functions to the
voice. For example, the tonal contours that are absent in the whispered text
are instead produced by the inanga, which simultaneously matches the
moraic rhythm of the text. Thus tone and vowel length are two features of
the text that are reflected in the instrumental melody. In addition, a hier-
archy of generic rules-some linguistic, some melodic-is built into the
combined sounds of the whisper and instrument, until the relationship
between the inanga and voice is entirely synergetic.
The synergy of the voice and instrument allows interaction between
features of the two components to produce a timbral anomaly. Specifical-
ly, primary elements from the harmonic complex of the inanga fuse per-
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
80 Ethnomusicology, Winter 2002
ceptually with the vocal (whispered) component of the genre to produce
the sensation of a melodicized whisper, an example of anomaly by redis-
tribution and a true auditory chimera. Notice however, that the voice/in-
strument synergy is based on constructs existing only in the mind of the
listener, not in the physical sound to which they apply. It is the listener
who expects the voice to produce a melody, the listener who expects
melodic and tonal correspondence, the listener who expects vowel length
to be reflected rhythmically.21 As described above, the redistribution tech-
nique does nothing to change acoustic properties of the relevant sounds,
it is the listener who must experience the changes. As the whispered stro-
phes alternate with short instrumental interludes, the illusion turns on and
off with each verse, each time requiring listeners to rearrange their percep-
tual grouping strategy.
The work of the Whispered Inanga musician is subtle and indirect. He
intends his music to introduce a timbral anomaly into the perceived world
of his listeners. But whether by design or necessity, he cannot change the
acoustic world of his sounds, and he cannot enter the perceived world of
his listeners. Instead, he carves out from their shared acoustic world a
smaller world-call it an acoustic arena, a performative space-where
perhaps listeners can relinquish their source-ridden perception without
physical danger. Or perhaps it is simply a place where the musician sub-
stitutes other concerns for source identification. Following the generic rules
of Whispered Inanga, he creates a place where the distinction between
sources is less clear cut and exacting: musicians insist that they can only
perform at night when auditory and visual distraction is at a minimum, and
they position themselves with faces as close to the inanga as possible to
minimize the distance between sources and obscure localization cues that
result from their separation. But these efforts represent the limit of the
musician’s control over his music’s reception. He can create the smaller
arena of Whispered Inanga, he can beckon to the audience, but finally, lis-
teners enter or not as they will. One might suppose that part of the expe-
rience of Inanga Chuchotee is the implicit decision on the part of the au-
dience to yield to the musician, to the perceptual transformation of his
music. For the illusion to succeed, listeners must be willing to surrender
their concentration and imagination in service to the illusion.
Inanga musicians are well aware of the importance of the listener’s
willingness to fall under the spell of the music:
Ni mwijoro nyene ikunda kuvuga mwijoro. Ku murango
It [inanga] likes to speak only at night. In the daytime,
baca imyaga nyinshi abantu bakavuga ari benshi. Nayo abandi
there is a lot of chatter, many people are speaking, while
inanga urabona mw’ijoro, abantu baba bahoze bakababicaye bamwe
bakumvira
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 81
the inanga-you see, at night people are silent and sitting together, and
they’d listen to
uno murya w’inanga …
to that string of inanga..
(Etien Biranguza, Rusaka Commune)
nayo abandi bari abatumva neza bokwumva upfa
Otherwise there are those who don’t hear well, they only hear
kuvuza mu murya, batazi gufata ico uriko uravuza. ..
what you make the strings speak, they don’t know how to hear [the
substance] of what you’re causing-to-speak…
bobitora batevye kuko abantu ntibategera kimwe
They discover/feel it [the effects of the music] slowly, because people don’t
understand in the same way.
urabona kwiga kwumva inanga biba mu maraso.
You see to learn to hear Inanga it has to be in the blood.
(Joseph Torobeka, Bujumbura)
The phenomenon of Whispered Inanga then, is not so much an illusion as
a collusion between audience, performer, and instrument.
Inanga ya Kubandwa
The last kind of inanga music to be examined here is the least hetero-
geneous of the three and the most constrained in performance. It is per-
formed only by designated individuals and only in very particular circum-
stances. The music is a ritual part of an initiatic possession cult called
kubandwa, and it demonstrates the second large category of timbre ma-
nipulation, the juxtaposition of timbres from opposite ends of the timbre
continuum. Kubandwa, is a phenomenon of all the intralacustrine countries
of Africa (Berger, 1973). In other countries, the cult worships a spirit called
Ryangombe; in Burundi, the spirit is called Kiranga or occasionally Kiran-
ga-Ryangombe. In all countries where kubandwa is practiced, the spirit
Kiranga-Ryangombe, is understood to have begun as an ordinary mortal; this
fact is important, since it means that Kiranga-Ryangombe is actually one of
an infinite number of imizimu or ancestor spirits who are ordinarily in-
clined to malevolence at the slightest sign of neglect from living relatives.
One of Kiranga’s functions is to protect the living from dead.
By all accounts, the rituals marking the celebration of kubandwa are
more often arduous, draining of resources, and fearsome than they are joy-
ous or purely celebratory. Initiation, in particular, occurs in the dead of
night and involves a fair amount of unpleasantness for the initiate. Part of
the apparent lugubriousness of kubandwa ceremonies begins with the pre-
cise role of Kiranga in regard to his human charges. Though he may grant
favors-and a large part of various rituals seems to consist of listing the
wants and needs of supplicants-Kiranga and his spirit attendants act pri-
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
82 Ethnomusicology, Winter 2002
marily as protection against a vibrant world of sometimes vicious spirits.
And Kiranga too can be exacting and vengeful if an umubandwa (a follow-
er of Kubandwa) promises something to the spirit in return for some de-
sired outcome, and then fails to deliver. But most often Kiranga is enlisted
to help ward off attacks by the intezi or abaganza, spirits whose particu-
lar role seems to be to afflict human victims as a way of bringing them to
Kiranga. On being called by an igishegu or adept of the cult,
Kiranga yaseruka uno munsi kuri uno (muntu) yavyiteguriye. . .emwe
[Kiranga] agabanga.
Kiranga makes himself be seen that day at the house of the one who
commanded the ceremony, and then he forms a screen (between the
protected and the assailing spirits).
(Marita Ntawenganyira, Sous-colline Mija)
There is little hesitation on the part of ababandwa to admit that one
does not come willingly to kubandwa, a sentiment mirrored in numerous
kirundi proverbs. The term “kubandwa” in fact derives from the passive
form of the verb kubanda, to push or press, so that the very name of the
cult-to be pressured–implies the reluctant submission of adherents. The
sequence of events leading to initiation seems to be as follows: A man or
woman becomes sick; perhaps s/he consults a traditional healer who of-
fers no relief, or a European doctor who is equally powerless, or both. By
now, it is clear that the victim has been assaulted by some unnatural force-
a spirit of some kind-and s/he hopes fervently to find that the spirit is
umuzimo-an ancestor spirit-since these are among the least noxious and
costly to appease of all spirits. Next s/he consults an umupfumu, a divin-
er or sorcerer, who investigates the condition and announces- alas–that
the victim is not beset by imizimo, but instead “arategwa intezi”, s/he has
been thrown or attacked by the intezi or abaganza spirits whose appease-
ment demands initiation to kubandwa. Bernard Zuure points out that while
Rwandans come to Ryangombe willingly and with celebration, the Barun-
di are coerced, chosen by Kiranga who shows his choice by sending the
intezi: “Utaterwa intezi ntiyobandwa-he who is not thrown by (possessed)
the intezi will not go to kubandwa” (quoted in Zuure, 1929).
However, both in live interviews and in the little existing literature
about kubandwa, there is considerable suggestion that initiation provokes
conflicting feelings. The initiate has, after all, been personally chosen and
betrothed by Kiranga, and whether male or female, becomes Kiranga’s
“wife” for the duration of the initiation ritual. There appears to be a fair
amount of pride on the part of initiates at being the center of what is un-
doubtedly an impressive ceremony, and at joining a community which has
previously excluded them. Furthermore, there is large evidence that once
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 83
initiated, ababandwa are eager to proceed to a higher stage in kubandwa,
a stage where they become ibishegu, are privy to esoteric knowledge, and
have the authority to initiate new members or to conduct other ceremo-
nies which praise or call Kiranga for protection or special requests. Though
it is true that ibishegu are often the recipients of various gifts and offerings,
it is unlikely that they accrue enough real wealth from their position to
motivate them to higher levels in the cult, especially since the position also
subjects them to a fair amount of responsible and sometimes onerous ac-
tivity in the community. More likely, those who become ibishegu feel a real
calling to the position, a kinship to Kiranga and his spiritual attendants, and
a desire for ongoing contact with the world they inhabit.
If kubandwa inspires conflicting reactions, the conflict sharpens in the
experience of Inanga ya Kubandwa. Inanga ya Kubandwa differs from the
other two kinds of inanga in both sound and purpose. It is not entertain-
ment; it is sung prayer. Interviews suggest that questions concerning tal-
ent or skill in performance make little sense to the ababandwa, though this
seems less true in regard to the ability to play the inanga than the ability to
sing. Typically, all participants in a kubandwa ritual sing, though usually
the leader of a song in call-and-response form is the one who plays the inan-
ga, almost always a man. With the exception of a better or worse ability to
remember texts to many songs, there seem no real criteria for a good or
less good performance. The sole characteristic that seems to distinguish one
performance from another is the quality of ubuguruguru, a term difficult
to translate, but meaning something like “noise”, “agitation,” or “turbu-
lence.” The term actually applies less to a performance of Inanga ya Kuband-
wa, than to a ritual or ceremony at which the music has been performed
in a particular manner. Not only does ubuguruguru distinguish different
calibres of music within the context of kubandwa, but it also distinguishes
Inanga Chuchotee from Inanga ya Kubandwa.
inanga ifise kuvugirwa basi apana gusobora ijwe urishire bejuru
Whispered Inanga has to be whispered for with a low voice, not a high
voice,
Harya inanga uyongoreye usbobora kuyumva na we nyene. Nayo barya
ushira
That inanga you whisper for it, you can hear it in the right way. While when
you put
ijwi bejuru, canke ukayivuza cane ukaturutsa umurya. Birya
birashobora
a high voice, or you speak in a high voice, you make the string cry out loud.
It [Inanga ya Kubandwa] can
kwanduruku nko ku buguruguru nyene.
cause that turbulence exactly [that we spoke about].
(Lazare Nkurikiye, Colline Biganda)
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
84 Ethnomusicology, Winter 2002
From this and similar comments in recorded interviews, it is clear that
the difference between Inanga Chuchotee and Inanga ya Kubandwa is not
simply that the first is whispered and the second sung in full voice, but that
the singing in the second is high-and there is no question here that “high”
means “high in pitch.” Inanga ya Kubandwa requires a kind of falsetto that
not only sets it apart from Inanga Chuchotee, but also from Voiced Inan-
ga. And it is exactly the falsetto of the voice combined with the timbre of
the inanga that produces the effect described earlier as timbre juxtaposi-
tion. Unlike the whispered and voiced components of the other two genres,
the high sparse sound of the falsetto-both for male and female singers-
produces a stark harmonic that pierces the thick formants of the inanga.
An even greater contrast is implemented with the addition of inyagara,
gourd rattles filled with seeds or pebbles that are essential to rituals in all
the kubandwa cults of east/central Africa. Inyagara produce pronounced,
noisy formants that alternate with the harmonic structure of the voice.
A comparison of the three kinds of inanga music presented here shows
that Voiced Inanga combines an instrument whose timbre has the relative-
ly pronounced formants of most soft-wood, stringed instruments, with ful-
ly-voiced vocal formants. All things equal, Voiced Inanga requires no regu-
lar timbre effects. Whispered Inanga shows the same elements as Voiced
Figure 10: Spectrogram of an excerpt from Voiced Inanga. Horizontal lines
are inanga harmonics; curving lines are voice harmonics. Both are well-
defined and formanted.
::::I
::
.::.:.
.::…
::::
::… :::.:::
:: ::::::
::::
:..
::,:.:
::x,:i: ……
:::. …… :.:. :::,:i-:
::~:: -::::?: -::-:: :::::-: :_:
: ,:. –:~::: :~i-::
.::.:.- :-:_::::
: ::~-: : ::-
: :::::-: : :::
::,::::
:~–~:: ::~::~a~as~;::: :::.:?::
~;i~–~~iiii:- .. ?i~–:~::::::::- :i~i:i::-~~i;i9:l-?”i~:::~~”-?–? :- -i~-i::;i
:i-~i”: _~a~i-~:’ fi~iiii –:i~: ‘:’~:::_i-:’:- _,iiiiiii i:i~:~ii~ii-:~i~: i~;i8ii _::~:iiiiii~i-
:::
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 85
Figure 11: Spectrogram of an excerpt from Inanga Chuchotee. Hazy shad-
ows are the noise of the whisper.
-i-i:i-i
:::F:::-: ::i::: :i::::_-:
.:
:: ::::
::-i :
i-i-i:
:-:-::_
_::E:i:-:::
:–iBls— ::-:-I:
:-:i::
::: :: : i- : :
:::::–:::ii:i: ,:_iiiii:,: -:i–i-i –uiii?:-ii
:::::8:i:i:: ::::i::_:::::::-:::
::: : ::
::~i’:i:::::
1::::i:::j::: :: ::::_::::::-:: :::a-:::
::: :::
: :: : :
: :-i:i-i—i:i- ….. -: : ::::
-:::-‘:::-:i-:_:::::8i:: ::::::: _ :- –
-:::: :::?::?::
-::-:::
-:::-:
.::::::_
:::i:::i:i :
:- : :
::::: :
:::: :
.::::..
:.. :-:
::: :::
Inanga, except that the vocal formants consist of noise rather than the dis-
crete harmonics of voiced singing. The timbre effect is dramatic in Whis-
pered Inanga-the entire genre revolves around it-but the effect is invis-
ible to acoustic analysis, since it is purely perceptual. In Inanga ya
Kubandwa, the impetus for the timbre effect is clearly visible from acous-
tic analysis. Indeed, as described above timbre juxtaposition depends on
purely acoustic elements of the sound, so that spectral analysis of Inanga
ya Kubandwa shows clearly the features that qualify a timbre as formanted
or harmonically-structured. Thus, an important distinction between Whis-
pered Inanga and Inanga ya Kubandwa lies in the degree to which their re-
spective methods of timbre manipulation adhere to the perceived world.
In Whispered Inanga, the musician uses a technique that appeals directly
to the forces of perceptualization, so that the entire effect occurs within
the perceived world; in Inanga ya Kubandwa, the musician relies on acoustic
features of juxtaposed sounds to dictate the behavior of perceptualization.
A corresponding distinction exists in the susceptibility of listeners to
the techniques in each genre. As noted above, the use of anomaly by re-
distribution in Whispered Inanga results in an effect that is largely volun-
tary on the part of listeners, whereas the use of timbre juxtaposition in
Inanga ya Kubandwa results in an effect that is almost inescapable. If Whis-
pered Inanga charms listeners into a timbral anomaly, Inanga ya Kuband-
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
86 Ethnomusicology, Winter 2002
Figure 12a: Spectrogram of an excerpt from Inanga ya Kubandwa. Notice
single prominent harmonic, second from bottom.
V.
Vi
wa captures them and assaults them with acoustic awareness; if the Whis-
pered Inanga audience drifts in and out of the appropriate perceptual mode,
modulating the impact of the effect, the kubandwa audience is riveted by
music whose effect intensifies relentlessly. Perhaps the most important
difference between the two kinds of inanga is that whereas Whispered
Figure 12b: Spectral slice of tone in figure 12a at point marked with arrow
on spectrogram 12a. Notice spike of prominent harmonic.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 87
Inanga is performed by a musician using all his artistry to persuade listen-
ers to perceive a timbral anomaly, Inanga ya Kubandwa is enacted by the
listeners themselves. Together the ababandwa create a juxtaposition of tim-
bre, each offering a sound to contrast with the others. As a group, they build
a musical momentum that no one individual can halt or resist, until the
contents of the perceived and acoustic worlds overflow into each other. It
is in this overflowing that is the significance of ubuguruguru appears to lie.
In kubandwa, then, the combination of the inanga and inyagara with
the falsetto voices of the ababandwa brings ubuguruguru into the music,
and leaves the border between the acoustic and perceived worlds perme-
able. Whether or not adherents of kubandwa actually understand the phe-
nomenon in terms of worlds and borders, it is surely no coincidence that
the metaphor of expanded perception plays well into the objectives of the
ritual. For just at the moment when the border between the acoustic and
perceived worlds weakens, Kiranga appears, having crossed over from the
world of spirits into the world of the living. Kiranga yabonetse-lets him-
self be seen–Kiranga yasokoroka-manifests himself. It is as though over-
exposure to the effects of timbre juxtaposition allows the ababandwa to
open a window between the worlds through which Kiranga enters. Al-
though it is unlikely that the arrival of Kiranga is conceptualized by the
Barundi specifically in terms of the ritual music that accompanies him, still
the analogy presented here is more than just a figurative paralleling of acous-
tic and spiritual worlds; rather it presents a precise resume of a listener’s
stance in regard to the known perceived world and the unknown acoustic
world. In the relationship between the spirit Kiranga and his human wor-
shippers we see in condensed form all the dimensions of a listener’s un-
derstanding of a world of acoustic sources he knows is there but cannot
experience directly.
Just as a listener infers a source he cannot see from its impact on his
auditory system, so also an umubandwa infers the existence of Kiranga from
the spirit’s impact on his life; as a listener constructs the nature and condi-
tion of the source from qualities of its impact, the umubandwa judges Kiran-
ga’s moods, his pleasures and peeves by the kind and intensity of his im-
pact. Though a listener has only auditory-thus circumstantial-evidence
of a source, if he determines it to be an object of consequence, he responds
with conviction; he does not wait for proof of an oncoming truck to step
back onto the curb. Similarly, a confirmed umubandwa responds directly
to Kiranga, though the evidence from which the spirit’s will is inferred may
seem indirect, random, or ambiguous to a nonbeliever. Indeed, neither the
listener nor the umubandwa respond at all to the immediate evidence, but
rather to the entity that the evidence confirms. Only an impossibly naive
listener steps back on the curb to avoid the sound of a truck, rather than
the truck itself; and only a noninitiate would go to a healer or sorcerer to
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
88 Ethnomusicology, Winter 2002
cure the illness which is only a symptom of Kiranga’s attention (“Si mur-
wayi, yaganiye na Kiranga”-he’s not sick, he’s chatting with Kiranga).
Just as a listener conflates sound and source to hear a cricket, an umuband-
wa takes the medium of Kiranga’s message to be Kiranga himself.
Like elements of the acoustic world, Kiranga can be experienced only
at one remove, and part of the esotericism of the cult appears to reside in
the secrecy surrounding the recognition and interpretation of that remove.
A notable consistency in interviews of ababandwa and in literature on the
cult is the conflation of Kiranga with signs of his influence. The birth of
twins or the striking of lightening are not simply omens or afflictions im-
posed by Kiranga to show pleasure or disapproval. They are rather Kiran-
ga himself, present at a greater or lesser remove from the perceiver. Often
an event constituting a greater remove of Kiranga’s presence seems to call
for the ceremony at which Kiranga is present with the least remove, in his
possession of an igishegu, an adept. Even in expository interviews detail-
ing the step-by-step procedures of kubandwa, informants elide the stage
between preparing for a ceremony, gathering the faithful, and the moment
of Kiranga’s first appearance; ababandwa seem never to distinguish between
the possessing spirit and the possessed conduit. According to most inter-
views, Kiranga indicates his arrival with a characteristic sound, ukuvumera,
a word best translated as “lowing” or “mooing”, the sound a cow makes to
its calf. At the sound of ukuvumera, Kiranga is present.
Jewe, usanga ari jewe, Kiranga-kiri-umweru, emwe tera imbere
nk’umuzinga.
Me, you have found me, Kiranga-painted-white, so progress forward like a
hive of bees.
(Felix Agahimbare, Colline Musanga)
Neither our interviews nor the interviews documented in available litera-
ture record details of the possession process-no trauma, no signs that the
possessed is undergoing a personality transfer. Simply Kiranga is there and
the chosen igishegu is not.
Interview questions meant to explore the juncture between the pos-
sessed and possessor-what if one igishegu gets tired, can Kiranga move
to another body? What if there are ceremonies on two hills at once; can
Kiranga be in two places at once?-questions such as these seemed to make
no sense to the ababandwa.
Ntibaba bagira ngo ni ugubita gusa… .ngo urya ni Kiranga.
It doesn’t have anything to do with some person only identified
(recognized), this one is Kiranga.
ahasigaye Kiranga agaca asokoroka, akabonekera aho, hano kuri uno
munsi.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 89
There is nothing left but Kiranga appears, he makes himself seen, there, on
that day.
(Marita Ntawenganyira, Sous-colline Mija)
The possessed does not become Kiranga; rather the possessed is Kiranga,
and when Kiranga is gone again, something else exists in his place, just as
Kiranga is present in the birth of twins, whereas when he’s not there, the
birth results in only a single child. The trick is not so much to recognize
the signs of Kiranga, but to recognize the forms of Kiranga, and the degree
of the perceiver’s remove from that form. Kiranga is perceivable in each
form, but the indirectness of that perception, the degree of remove between
the perceiver and Kiranga as presented in that form is measured by the
obscurity of the connection between the form and Kiranga. Though these
degrees of obscurity seem randomly assigned to an outsider (Kiranga-as-light-
ening presents him at a smaller remove from the perceiver than Kiranga-as-
an-act-of-bravery-in-battle), the correlation of Kiranga’s proximity with type
of representation seems so longstanding to the ababandwa as to require no
explanation; Kiranga is closer in the form of lightening than he is in the form
of bravery in the same way that Burundi is closer to Rwanda than to Gabon.
The most direct, least removed experience of the spirit is the perception
of Kiranga incarnate in the body of an igishegu. All encounters with Kiran-
ga-whatever the perceptual remove-lead to this ultimate presentation of
the spirit incarnate, though even it is one remove from direct experience
of the spirit. Direct experience of Kiranga would require a sojourn in the
spirit world, a world unattainable to living beings in a mortal world. In a
sense, the entire cult of kubandwa consists of movement toward and away
from Kiranga depending on the form in which he presents himself.
The variable proximity of the ababandwa to Kiranga, of course, adds
depth to the analogy between timbre juxtaposition and kubandwa. Just as
a harmonically-structured sound at the extreme end of the timbre continu-
um invokes the least amount of perceptualization, resulting in the greatest
similarity between the perceived sound and the acoustic signal that pro-
vokes it, so also the presence of Kiranga incarnate in a human body results
in an experience that approaches most nearly a direct encounter with Kiran-
ga in the spirit world. And just as the similarity between perceived sound
and acoustic signal reflects the proximity of the listener to the acoustic
world, so also the nearly-direct encounter with humanly incarnate Kiran-
ga reflects the proximity of the umubandwa to the spirit world. A final sim-
ilarity between kubandwa and musical timbre juxtaposition is the quality
of submission each requires. Both phenomena in a sense define a field of
power, and having crossed into that field, participants relinquish the abil-
ity to resist the power. Questions concerning the relative strength of Kiran-
ga and his human conduits elicit notions of fatalism and resignation:
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
90 Ethnomusicology, Winter 2002
Interviewer (Rosemarie Kadende):
None nk’ubu, umuntu yatewe tuvuge, arashobora… .umuntu
Well, like this, a person who’s been thrown (possessed by Kiranga), can he
… a person
ataraterwa arashobora [kwanka]… kwankira nk’ico cari
can he [refuse to] not be thrown . .. refuse with those amulets
kumuzamwo? Nk’ubu comatuma nko kwica umuntu, yobigira?
to come over to them? For example, if they (Kiranga and/or his attendants)
ordered him, say, to kill a person, would he do it?
Arashobora kwanka? Urumva aho mvuze nyene ikuntu gihambaye cane,
Could he refuse? You understand I’m taking here (an example of) something
very terrible/complicated
mugabo tuvuge nk’ubu kimutumye kugira ikindi kintu nyene kigoye. ..
but let’s say, for example, he ordered him to do some other thing/act
delicate…
Umubandwa:
Oya… .ico ntibiba.. .oya, ntibiba. Oya Ntibiba. Oya…
No, this, it can’t be, no it’s not possible, no it can’t be no…
aba agomba kwiyahura. . Bimatumye gutryo, kuri iryo jambo
He wants to kill himself. When the things (spirits) order him like that, to
that word
kwica umuntu, aba abagomba kwiyahura.. .Indome ziwe ziba
To kill a person, he wants to kill himself.. .his words/his term
zashitse Imana yamwandikiye.
has arrived there (at the end of the plan/story) that God has written for him.
(Bitama Marcien, Ecole Primaire de Busimba)
Just as the only retreat from timbre juxtaposition is to remove oneself from
the music altogether, escape from Kiranga’s will seems to correlate with a
loss of vital will.
Without pursuing the analogy further, I point out that the parallelism
between timbre juxtaposition and kubandwa emerges in part because the
two phenomena may be seen to have a common basis: both proceed from
an impetus toward what is unknown. It has perhaps been evident from the
beginning of this paper that the perceived world/acoustic world distinc-
tion is simply a subregion of the subjective/objective dichotomy that em-
braces the contributions of all the sensory systems to a perceived reality
as constructed from the objective reality of physical stimuli. To the extent
that the subjective/objective distinction describes the line between the
known and unknown, many formulations of an essentially epistemological
duality end in metaphysics: the ultimate unknown is God. Thus, it is per-
haps not inappropriate that music that explores the subjective/objective
duality in one domain should become an expressive instrument in explor-
ing the same duality in another domain.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales.: The Paradox of Timbre 91
Conclusion
I have proposed in this paper three characteristics of timbre that con-
tribute to its role in music. First, timbre constitutes a link to the external
world in containing the descriptive clues important for source identifica-
tion and for deciphering aspects of the terrain between the source and the
listener. Second, it functions as perceptualization’s primary instrument
in accomplishing its various objectives. And third, it is a parameter of mu-
sic that we experience phenomenally, but without informational con-
sciousness. The first two of these characteristics are responsible for what
I have called the paradox of timbre: that timbre is the parameter of sound
most implicated in source identification but also most implicated in the
discrepancy between an acoustic signal and the percept it provokes. The
third characteristic of timbre is the one that allows it the malleability and
elusiveness to wield its power in music. As an informationally unconscious
phenomenon, timbre is restricted to indirect measure and its effects are
likely to be misattributed to some other outstandingly conscious aspect of
the experience. Thus, timbre is free to operate with little direct scrutiny
by a listener, creating effects that are intense but also hazy in definition,
difficult to articulate, and freely attributable to other features of the musi-
cal context. Without addressing the issues of universalism or comparativ-
ism, then, we might note that though the principles of auditory cognition
presented in this paper imply a certain uniformity of process, they also allow
wide variation in result. Indeed, the “misattribution” that is so clear a clin-
ical sign of unconsciousness becomes an unmarked conceptual and musi-
cal field to be shaped with all the creativity and variability whose study is
the substance of ethnomusicology.
I do not mean to imply that musicians in command of the various tech-
niques of timbre manipulation understand the acoustic or cognitive impli-
cations of what they do in the way they have been presented here; it may
well be, in fact, that they are as informationally unconscious of timbre as
their audience. But it is clear that a gifted musician knows that certain
sounds or combinations of sounds carry a power that others do not, and
that part of a musician’s art includes a sense of when to make use of that
power. A skillful use of timbre betrays the stratagems of perceptualization,
disrupts its efforts to maintain an unfractured perceived world, and subjects
listeners to relentless small pricks from perceptual to acoustic awareness.
If there is doubt as to the ability of musical timbre to reveal the exis-
tence of acoustic and perceived realities, one might consider the frequen-
cy with which musical sound is represented in many traditions as a kind
of “double medium”, a place-holder for some absent entity: music as the
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
92 Ethnomusicology, Winter 2002
voice of the ancestors, as the sounds of nature, as the cry of lost children,
as the wisdom of totem animals. Western music as well has consistently
conceptualized music with a kind of binary structure: in a sense, the en-
tire history of western music aesthetics is shaped by the effort to explain
why perceived music is not simply itself-an art form with sound as a
medium; rather, music is an earthly reflection of divine proportion, a ren-
dition of celestial harmonies, an imitation of nature, or an expression of the
passions. Perhaps conceptions of music such as these reflect the efforts of
diverse cultures to recover a “sourceness” lost from music, or to make sense
of the disjuncture between acoustic and perceived worlds unveiled by
musical timbre.
Notes
1. Rhythm, of course, is also subject to determinations of rightness or wrongness, though
for expressive purposes, broader deviations in meter than in pitch are tolerated.
2. It is artificial to talk about one musical parameter in isolation from the others. They
all interact and are at times perceptually interchangeable. But since in most respects, timbre
is distinct from and covaries with pitch, loudness, and duration, and because it must be iso-
lated conceptually to understand its unique functions and effects in music, I will continue to
refer to it in isolation.
3. We should note from the outset that the ability to use timbre in music or to hear tim-
bral change is not the same as the ability to describe or directly examine timbre: it is perfect-
ly possible that even musicians who are particularly sensitive to timbre might still be unable
to describe what they are hearing.
4. Even in literature on Western music in fact, from the presocratic philosophers until
the early romantic period, descriptions of timbre are practically nonexistent, except rarely
and often elliptically in regard to performance technique. For example, Charles Burnett points
out that throughout the middle ages “whenever the perception of two different sounds is
discussed. .. differences in pitch rather than in any other quality are at issue. Pitch is the
primary differentiating factor in sound-rather than say, brightness, volume or sweetness”
(1991:48); and as early as the twelfth century, the translator Dominicus Gundissalinus noted
that there were no words to describe differences in sound quality; different sounds “had no
names of their own”, but were described by analogy with other senses (63).
5. The neurobiologist Walter Freeman (2000) describes the same paradox as a more
general phenomenon:
The conclusion is that the only knowledge that animals and humans can have of the world
outside themselves is what they construct within their own brains. … This finding could
not have been obtained by introspection, because the process of observation contains
within it some well-known operations that compensate for accidental changes in appear-
ances of objects owing to variations in perspective, context, and so forth. …. Each ex-
posure to a stimulus changes the brain’s synaptic structure so that it cannot respond
identically over time, although it may appear subjectively to be so. (414)
6. Recent MNN experiments show listeners able to preattentively detect timbral chang-
es in sound segments so short in duration that a pitch determination is impossible.
7. Recent attempts of theoreticians (cf. Erickson 1975, Slawson 1985, Cogan 1984, Cogan
and Escot 1976) to establish analytical systems of timbre have sadly met with little response.
Also, an anonymous reviewer for this paper points out that among ethnomusicologists, Ger-
hard Kubik, David Dargie, and David Rycroft have also addressed issues of timbre.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 93
8. Impulsive sounds are those produced by discontinuous excitation of a vibrating ma-
terial (plucking or striking, for example); sustained sounds are those produced by a continu-
ous stimulus (blowing or bowing, for example).
9. For ease of expression and consistent with current (perhaps deplorable) practice, I
will use the words preattentive and unconscious synonymously to refer both to auditory pro-
cesses that occur without the active awareness of the listener, as well as to the results of that
processing.
10. An example of phenomenal consciousness without reflective consciousness might
become apparent if, after a few days of feeling uncharacteristically pessimistic and negative,
one suddenly realizes, “Oh yes, this is a general sadness that occurs every year near the anni-
versary of my father’s death”; the realization of the exact content and cause of one’s unchar-
acteristic emotion marks the return of previously absent reflective consciousness; an example
of reflective without phenomenal consciousness might be a situation in which one recogniz-
es intellectually an object or person to be dangerous, while continuing to feel unthreatened,
even to the point of carelessness.
11. The necessity of grouping acoustic elements by source results from the fact that
incoming signals from multiple sources undergo a form of natural Fourier transform by which
the composite signal is “unravelled” at the basilar membrane into discrete frequency compo-
nents. Each component sends a signal to the brain, traveling along neural pathways that are
predominantly tonotopic, with the result when news of the signal reaches the brain, its ele-
ments are mapped by frequency rather than by source.
12. E.g., heuristics might group together partials that are harmonically rather than non-
harmonically related, that have the same onsets, that modulate in frequency together, etc.;
all of these features are characteristic of sound in the acoustic world.
13. In addition to the relative amplitudes of its harmonics, a sound’s timbre is also char-
acterized by its attack and time variant spectral features.
14. For example, a source-based schema might group together a collection of frequen-
tial components based on their harmonic relation to each other (e.g., frequencies of 220 Hz,
440 Hz, 660 Hz, 880 Hz, 1100 Hz, etc.), producing the percept of a tone of pitch A, with a
distinctive timbre; under the right circumstances, a simultaneous tone consisting of harmon-
ic frequencies of 165 Hz, 330 Hz, 495 Hz, 660 Hz, 825 Hz, and 990 Hz might succeed in “cap-
turing” not only the 660 Hz component of the first tone, which fits into its own harmonic
series, but also the 880 Hz component, not because it is harmonically related to other frequen-
cies in the second tone, but because it “follows” the 660 Hz component into a new group-
ing. The result of this capture would be that both the first and the second tone would have a
new timbre.
15. Among other generic variations in both overtone singing and didjeridoo music, the
prominence of the unfused elements may be varied or maintained at a level of intensity that
exceeds only slightly that of surrounding harmonics, so that inexperienced listeners are of-
ten unable to hear the extracted elements separately, while experienced listeners can shift
perceptual gears back and forth between “hearing out” the harmonic elements separately or
leaving them fused to the original timbre. When listeners “leave” the elements fused to an
original timbre, they can be said to be resistant to the effect of timbre manipulation.
16. As opposed to noise which can also be considered timbre without pitch, but which
does not consist of harmonically-related overtones, but rather of random frequencies.
17. The harmonics that are produced on stringed or other instruments to create pitches
in an upper range are not really single frequencies, but rather some subset of the original tone’s
harmonics; on a stringed instrument the specific hamonics included in this subset are deter-
mined by where the string is touched lightly enough to prevent certain modes of vibration.
18. In reality, of course, the difference between the extraction and redistribution tech-
niques is a matter a degree. As the overtone in overtone singing becomes less and less prom-
inent, the creation of anomaly becomes less a function of extraction and more a function of
redistribution, albeit of a single component; it is said that particularly gifted Tuvan children
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
94 Ethnomusicology, Winter 2002
can hear overtones in a simple drone even when none are emphasized, so that they are able
to hear whatever melody they like in any steady tone (personal communication).
19. More evidence for this difference lies in the fact that if any single harmonic is filtered
out of a formant-based sound, the overall sound quality of the sound remains unchanged; if a
single harmonic is filtered from a harmonically-structured timbre the difference in sound quality
is immediately apparent.
20. Thanks to Scott Marcus for information regarding the effect of bridge construction
on sitar timbre.
21. That these qualities are perceptual, not acoustic, is confirmed by the fact that they
are idiosyncratic to listeners who speak a language that is tonal with long and short vowels;
nonKirundi speakers are often hardpressed to hear an illusion that comes easily to Kirundi
speakers.
References
Averill, Gage. 1999. “Bell Tones and Ringing Chords: Sense and Sensation in Barbershop Har-
mony.” The World of Music 41(1):37-51.
Baars, Bernard. 1998. “Metaphors of Consciousness and Attention in the Brain.” Trends in
Neurosciences 21(2):58-62.
Barancira, Sylvestre. 1990. Possession par les Esprits: Baganza et Rituel Therapeutique du
Kubandwa au Burundi. Bujumbura: Ministere de la Jeunesse, des Sports et de la Cul-
ture du Burundi.
Berger, Iris. 1973. “The Kubandwa Religious Complex of Interlacustrine East Africa: An His-
torical Study.” Ph.D. dissertation, University of Wisconsin, Madison.
Bregman, Albert. 1994. Auditory Scene Analysis. Cambridge, MA: MIT Press.
Burnett, Charles. 1991. “Sound in the Middle Ages.” In The Second Sense: Studies in Hear-
ing and Musical Judgement from Antiquity to the Seventeenth Century, edited by
Charles Burnett, Michael Fend, and Penelope Gouk, 43-69. London: University of Lon-
don.
Butler, Robert A. 1973. “The Relative Influence of Pitch and Timbre on the Apparent Loca-
tion of Sound in the Median Sagittal Plane.” Perception & Psychophysics 14(2):255-58.
Cogan, Robert. 1984. New Images of Musical Sound. Cambridge: Harvard University Press.
Cogan, Robert and Pozzi Escot. 1976. Sonic Design: The Nature of Sound and Music. Engle-
wood Cliffs, N.J.: Prentice Hall.
Erickson, Robert/ 1975. Sound Structure in Music. Los Angeles: University of California Press.
Fales, Cornelia. 1995a. “Acoustic Intuition of Complex Auditory Phenomena by ‘Whispered
Inanga’ Musicians of Burundi,” The World of Music (37) 3:3-35.
. 1995b. “Auditory Illusion in Whispered Inanga of Rwanda and Burundi,” Systematische
Musikwissenschaft/Systematic Musicology/Musicologie systimatique 3(1): 125-54.
. 1998. “Issues of Timbre: Inanga Chuchotee.” In The Garland Encyclopedia of World
Music 1:164-207.
Fenn, John. 1997. Unpublished manuscript, Indiana University.
Freeman, Walter. 2000. “A Neurobiological Role of Music in Social Bonding.” In The Origins
Of Music, edited by Nils L. Wallin, Bjorn Merker, and Steven Brown, 411-24. Cambridge,
MA: MIT Press.
Kubik, Gerhard. 1989. “Subjective Patterns in African Music.” In Cross Rhythms: Papers in
African Folklore, Volume 3, edited by Susan Domowitz et al., 129-54. Bloomington, IN:
Trickster Press.
Liegeois-Chauvel, Catherine, Isabelle Peretz, Myriam Babar, Virginie Laguitton, and Patrick
Chauvel. 1998. “Contributions to Different Cortical Areas in the Temporal Lobes to Music
Processing.” Brain 121:1553-1867.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
Fales: The Paradox of Timbre 95
Mandler, George. 1991. “The Processing of Information is Not Conscious, but its Products
Often Are.” Behavioral and Brain Sciences 14(4):688-89.
McGurk, Harry, and John MacDonald. 1976. “Hearing Lips and Seeing Voices.”. Nature 264:
746- 48.
Merikle, Philip M. and Meredyth Daneman. 1998. “Psychological Investigations of Unconscious
Perception.” Journal of Consciousness Studies 5(1):5-18.
Merriam, Alan. 1951. Field Recordings, 66-127-F+, ATL 4490-527+, Archives of Traditional
Music, Indiana University, Bloomington.
Ndagijimana, Kosima. 1989. Inkoranyainagambo ngufi y’umuhango uwo kubandwa. Ru-
hengeri: Kaminuza y’u Rwanda.
Peretz, Isabelle, and Jose Morais. 1988. “Determinants of Laterality for Music: Towards an
Information Prrocessing Account.” In Handbook of Dichotic Listening: Theory, Meth-
ods, and Research, edited by Kenneth Hugdahl, 323-58. New York: Wiley.
Rodegem, Francois M. 1970. Dictionnaire Rundi-Frangais. Tervuren: Musee Royale de
l’Afrique Centrale.
. 1971. “La Motivation du Culte Initiatique au Burundi.” Anthropos 66:863-930.
Slawson, Wayne. 1985. Sound Color. Berkeley: University of California Press.
Velmans, Max. 1999. “When Perception becomes Conscious.” British Journal of Psychology
90(4):543.
Warren, Richard M. 1976. “Two Types of Auditory Sequence Perception.” Perception & Psy-
chophysics 29:323-35.
Zuure, Bernard. 1929. Croyances et pratiques religieuses des Barundi. Bruxelles: Editions
de l’Essorial.
This content downloaded from 149.31.21.88 on Wed, 30 Jan 2019 20:12:33 UTC
All use subject to https://about.jstor.org/terms
- Contents
- Issue Table of Contents
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
[76]
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
Ethnomusicology, Vol. 46, No. 1 (Winter, 2002), pp. i-vi+1-196
Front Matter [pp. i-134]
From the Editor [p. v]
World Music at the “End of History” [pp. 1-32]
The Rationalization of Intensity in Indian Music [pp. 33-55]
The Paradox of Timbre [pp. 56-95]
“Republic of China National Anthem” on Taiwan: One Anthem, One Performance, Multiple Realities [pp. 96-119]
Music in the Hong Kong Handover Ceremonies: A Community Re-Imagines Itself [pp. 120-133]
Dropping the Bomb: Steelband Performance and Meaning in 1960s Trinidad [pp. 135-164]
Book Reviews
Review: untitled [pp. 165-168]
Review: untitled [pp. 168-171]
Review: untitled [pp. 171-174]
Review: untitled [pp. 174-177]
Review: untitled [pp. 177-179]
Review: untitled [pp. 179-183]
Recording Reviews
Review: untitled [pp. 184-186]
Review: untitled [pp. 186-189]
Review: untitled [pp. 189-191]
Film and Video Reviews
Review: untitled [pp. 192-194]
Review: untitled [pp. 194-196]
Back Matter