Chapter 3

Hearing, Feeling, Grasping Gestures

Amic Cox

Musical gcstures are musical acts. and our perception and understanding of
gestures involves understanding the physicality involved in their production. Al
one leve! this is a rather straightforward mauer, but when it comes 10 using rhe
concepl of 'gesture' 10 analyse meaning construction, sorne of our assumptions
leave problematic gaps in our cxplanations. For example, consider the following
claims by David Lidov ( 1987: 82, emphases added):
The variables of pulse are speed and intensity,
lntensity is involving. The values of simple pulse
trong, forcground pulse as in folk dances and
11101c111c111 directly. Attcnuatcd pulse is a factor in

Speed is excuing.
are fairly obvious:
marches controls
the sublimation o/

somaticforce.

On the surfacc there may not be a problcm here, since pulse is sornething we oflen
feel when listening to music and it does seern lo control movement in certain
coniexts. But there is a circularity here, for we could jusi as well say that a feeling
of moverncnt generales a fecling of pulse, and that the sublimation of somatic force
is a factor in attenuated pulse - or so 1 would claim. Wherc does this feclng of
pulse originatc? lfwe say tbat it is a property oftbe music (of the acoustic stimuli),
which we fcel when listcning to music, thcn we are lcd bnck lo whcrc \Vestartcd:
wc fccl pulse bccause pulse is thcre 10 be fcll. Thc problcm hcrc is onc shnrcd by
other concepts rclatcd to embodicd meaning, including 'gcsturc': How is it that
music rnakes us feel anything al ali? (1 am not refcrring nccessarily to emotional
feelings but 10 the more visceral sensations related directly lo movement.) In the
context of folk music and marches it might not seem that this is a matter in need of
explanation: peoplc dance to dance rnusic and march to mnrch music, and the
qucstion o how music works in thcsc contexts may not scern to sorne a crucial
arca of scholarly inquiry. However, Lidov is using these exarnples of obvious
physicnl engagernent as part oan cxplanation of how similar types of engagernent
occur in musical experience generally, including 'art music', and sincc this
engagemenl and its cause are not as obvious as in other repertoires, and because a
great deal is at stake, this claim requires a more explicit understanding of how
music engagcs us. What is at stake, to my mind, is the claim that musical meaning
is generated by our embodied experience of it - that our embodicd experience is
not only nccessary for expcricncing meaning that is somchow inhercnt in the music

Copyl'l<ll"llXI
m~~'"

46

Muste mu/ Gesture

itsclf, but that meaning arises in our conccptualizations of cmbodicd musical

experience and that abstrae! meaning is the product of embodied reasoning. This
broad claim requires an examination of the concepts that we tend to take for
granted, including the far-reaching question of how we conceptualize relations
among musical events in tcrms of 'rnotion' generally, and io tcrms of 'gesture'
specifically, The topic of tbis chapter is tbc quesrion of what motivares and
strucrures conccptualization of music in tcrms of 'gcsrures', and what scnsc there
might be in using 'gesture' in addition to, or instcad of, the afien cocxtensivc
'motive' and/or 'figure'. To answer this, 1 first offer an account of how music
engages us bodily.

Thenretical Background: The Mimetic Hypothcsis

Do you ever ftnd yourself tapping your toe to music? Jf so, why should this be?
This very common response to music, along with other subtly overt embodied
responses such as head-bobbing or swaying, is not something we normally choose
to do (nlthough we are of course certainly capable of choosing to engage witb
music in overtly cmbodicd ways). lnformaUy conducting, playing 'air guitar', and
'beat boxing' (vocal imitation of the rhythm section in rap) are similar responses.
although these are perhaps more often a matter of volition. But we do more than
visibly move to music; we also sing along, in real time and in recall, aloud and in
our heads. By way of example, let me offer the following anecdote.
1was at a recital by the renowned bass Thomas Quasthof in which he sang
thc Kcms/Hammcrstein tune 'Old Man River' for an encore. As he sang 1 thought
that 1heard a sort of echo and theo realized that an elderly gentleman was actually
singing along. His wife shushed him, but when the cborus rerurned he started
singing along once more (much to her quite evidcnt cmbarrassmcnt). As 1read this
scenario. this man's impulse to cngage with the music as he did ovcrrode his social
inhibition: it was clear that his wifc thought he should listen quietly, but it was as if
he could not hclp himself. 1 believe that bis response is of the same sort as toetapping and other kinds of involuntary participation, and 1 believe that these are
overt forms of an imitative participation that is a regular pan of musical cxperience
and comprebension. Bascd on observcd mimetic panicipation of these sorts 1have
proposed thc following hypothcsis: pan of how wc understand music involves
imagining making the heard sounds for ourselves, and this imagincd participation
involves covertly and overtly imitating the sounds hcard and irnitating the physical
actions that produce these sounds.1 Mimetic participation occurs in three forms: ( 1)
covert and oven imitation of the actions of pcrformers (whether the performers are
seen and hcard, or heard only, or rccallcd); (2) covert and overt subvocal imitation
of the sounds produced, wbether the sounds are vocal or instrumental (likely to
include an imitation of timbre as well as pitcb, rhythm and dynamic lcvcl): and (3)
an amodal, cmpathctic, visceral imitation of thc excrtion partcms that would likcly
produce such sounds. Each of thesc forms occurs in real time and in recall.
sometimes within the same modality (as in vocal and subvocal imitation of a
singcr) and olcn across modalities (vocal and subvocal imitation of an

Copyncl l1't1 m1::~r1

1

Hearing, Feeling, Grasping Gestures

47

instrumental melody). The strength of each fonn of mimetic part.icipation vares

from person to person, which establishes a variability in the meanings that we
derive from musical experience, including the meaning of 'gesrure'. Understanding
the variability in different listeners' particular habits of mimetic engagement is
importan! for communication about music in tenns of 'gesture' and other bodily
based conceptions.
Before considering the evidcnce for the hypothcsis, 1 should compare lhis
to similar ideas on the subject. The rnimetic hypothesis is similar to ideas pul forth
by Patricia Carpenter (1967), Manfred Clynes (1977), Thomas Cliflon (1983),
David Lidov ( 1987), KendaU Walton ( 1993, 1997), Naomi Cumming ( 1997)
Andrew Mead (1999) and thc ninetecnth-ccntury philosopher Herbert Spenccr
( 1857). What cach of thcsc writcrs agrcc on is that imitation on thc pan of listcners
scems clearly lo play a role in musical expcrience; however, there are threc
importan! ways in which their conceprions differ from the mimeric hypothesis:
they provide liule orno evidence (beyond the anecdotal such as 1 have offered thus
far in a preliminary way); thcy do not examine the forms, or modalities, of mimetic
participation; and tbey do not consider the breadth of tbe implications.2 In this
cssay 1 follow this third rcstriction, although clsewhcrc 1 considcr briefly thc
various aspccts of musical meaning in which mimetic participation would scem to
play a role (Cox 2001 ). One implication beyond the topic of 'gesture ', which 1
must allude to here in the interest of contextualization, is the construction of the
concepl of musical verticality, or 'high' and 'Iow' notes. Most of us take this as
one of our most basic, quasi-literal concepts, but cven whilc acknowledging its
metaphoric basis there is not a generally accepted account of thc metapboric
rcasoning that gives us this conccpt. 1 havc offcrcd a preliminary account of how
this conccpt can be undcrstood to be motivated by mimctic panicipation (Cox
l 999a), which in turn activares the cmbodied metaphoric rcasoning described by
George Lakoff and Mark Johnson ( 1980, 1999; Jobnson 1987). The significance
for tbe present context is tbat the mimetic hypothcsis is something that has
implications eveo for our most basic, quasi-literal concepts - including concepts
such as 'asccnt', 'dcsccnr' and 'contour' which dcpend on thc concept of musical
verticality - as well as thc conccpt of 'gesturc', which is why this aspect of musical
cxperience and conceptualization needs lo be deiailed with carc.

Mimctlc Participation

in Dally Lifc and in Musical Expcriencc

Mimetic participation is fundamental to human comprchcnsion, and its role in

musical experience is very much simply a special case of how the cmbodicd mind
works generally. Since thc evidcncc is detailcd elsewhcre (Cox 2001) 1 will rcvicw
it only bricfly hcre. Clinical evidence for the mimctic hypothesis can be sorted into
four kinds: (1) studics of imitation in facc-io-face communicarion; (2) motor
imagery studies involving mirror neurons; (3) subvocalization studies for spcech
and for rnusic; and (4) non-vocal motor imagery studies for music. lt is Important
to note that each of these is a kind of motor imagery: hand and limb movernents,
adjustmcnts of posturc, facial gesrures and vocal 'gesrures' (the motor actions thnt

Co1..yr .1htcdir,-.1011

48

Mvsic and Gesture

produce speech and song). To pul it another way, in each case thc sounds are
evidence of the motor actions that produce them, and our comprchension of the
sounds involves comprehension of the relevant motor actions. In addirion to the
clinical evidence, we also have (5) the indirect evidencc of our vocal descriptions
of non-vocal sounds: we regularly describe instrumental sounds in terrns of vocal
sounds (such as cantabile for an instrumental mclody), and thc voicc is rcgularly
uscd as a modcl for mclodic playing in instrumental pcdagogy. One implication of
this is that instrumental sounds are rcgularly conceptualized in tenns of vocal
cxpcrience, and that this is motivated by subvocal imitation of non-vocal music,
l.
Imitation in face-to-face comrnunication includes, among other tbings,
srudies of infant-parent intcractions.3 Babies imitatc ihosc around them and this is
pan of how we lcam to understand otbcrs: wc scc and hcar things - facial, vocal,
gestural - and wc imitare thcsc actions. Our specics sccms to have evolvcd in such
a way that succcssful imitation is crucial for succcssful communication, including
leaming to reproduce thc vocal sounds of spcech. Signi ficantly, when infants and
parents imeract it is not only the infants who do thc imitating but the parents as
well. Why should it be that parents spend any Limeirnitating their bables? One way
10 understand this is that mutual imitation fosters mutual understanding. In mutual
imitation we bccome likc those we are trying to undcrstand - we undcrstand (in
sorne mcasure) what it must be like to be thcm because we are being like them.
This son of empathy is fundamental 10 being human, which brings up another
point. While it might seem thai the imitation we praciise as infants is something we
outgrow. it appears instead that our imitation becornes more covert as wc mature and occasionally becomes overt in certain situations. We can understand this as
resuhing from thc gradual development of motor imagery: the capacity to
remcmber and plan motor actions in imagination, without having to rehcarse these
actions ovenly.
2.
Some of the best evidence for mirnetic participarion gcnerally comes from
studics of mirror 11e11rom. These are neurons that firc not only when a goaloricruated action is pcrformed - particularly grasping gestures - but also whcn
similar actions are obscrved. Care must be takcn in applying this evidencc to
music-producing actions (which have yet to be studied direetly). but the findings
are suggestive.
3.
Perhaps the most spccialized motor imagery is that rclatcd to speech
production and comprehcnsion, and part of how wc comprehend speech is through
mimetic subvocalization (Gibson & Levin 1975; Gathercole & Baddcley 1993).5
Wc can vicw subvocalization in aduhs as a covcn form of what infants and
children do in acquiring language. We can also see this as a special case of motor
imagery: this aspee! of speech comprehension is comprehension of the motor
actions that produce thc sounds of spccch. In addition to the speaking voice, not
surprisingly, a similar process occurs in comprehending the singing voice,6 so that
pan of how wc undcrstand song (in real time and in reeaU) involves subvocal
imitation. Since subvocalization is groundcd in the physical cxpcricncc of ovcrt
vocalization, comprchension of heard song thus appears 10 involve comparison
with our own experience of singing or otherwise vocalizing. This subvocal
cmpathy is part of what wc fccl when listcning to singing, and difTerent kinds of

OL.Yr Jh"

Hearing, Feeling. Grasping Gestures

49

singing (e.g. Dusty Springfield, Maria Callas, Janis Joplin, Ella Fitzgerald) can be
understood to generate dilTerent kinds of feeling in pan because they invite
differcnt kinds of mimetic participation.
4.
There is a similar story to be told about instrumental musical sounds but it
is twofold. There is little dircct evidencc for mimctic motor imagery regarding thc
motor actions of instrumental musicians - we have, for example, the fingcr
movemeots of pianists when listening to piano music (Haueisen & Knoschc 2001)
- but 1 believe that it is only a matter oftime before we have more evidence in this
domain. Clynes's (1977) work indicares a degrec ofisornorphic physical responses
across modalitics. By itself his work rnight be interpreted as indicating little more
than thc fact that music generales a physiological response that somchow marches
the intensity of the rnusic heard; however, in combination with thcse other kinds of
cvidence his work can be interpreted as part of an impulse to understand through
imitation - by matching (unconsciously, or normally so) the physiological intensity
that is somehow isomorphic with the rnusic and its meaos of production.
5.
Of panicular interest is the cvidcncc for subvocalizaticn for instrumental
mclodies (Baddeley & Logie 1992; Smith, Reisberg & Wilson 1992). Since most
of us havc a voice and havc used it to make and imitare sounds for most of our
lives, it should not be surprising that we would draw on vocal irnagcry to
understand instrumental musical sounds generally. But this means that musical
sounds generally are understood partly in terms of our embodied vocal experience,
making the exertions of specch and song relevant for understanding music
generally. In other words, pan of what we fccl when listcning to music are the
imagined, imitativc (sub)vocal cxertions along with the imagined, imitative intramodal exertions (the exertions specific to a given instrument). We can see this as
being reflected in our vocal descriptions of instrumental sounds: 'cantabile",
'cantilena' and 'mezza vocc' in music for strings and for piano; polyphonic
'volees': 'voicings' of piano hammers and organ pipes; 'choirs' of strings;
'scrcarning' jazz trumpets and rock guitars; and getting notes to 'speak' on wind
instrurnents. To feel what 1 mean, try the following: recall a favourite instrumental
mclody - perhaps the slow movcmcnt of a symphony or a chambcr work - and
while doing so, ask whether your voice is engaged in any way at ali: thc fccling of
singing along, or only the urge to somehow participare subvocally. With few
exceprions peoplc repon that they do indeed feel sorne son of subvocal
cngngement. (There are several ways in which this could be and should be tesred
cmpirically, along with the question of whether people subvocalize without being
aware of it.) This is a remarkablc phcnomcnon - that rccall of an instrumental
mclody should cngagc thc voice - because it indicates that pan of how we
cornprehend music (at least in rccall), is in terms of vocal cxpcriencc, rcgardless of
the medium of the sourcc. This would makc the exertions of vocal experience
relevan! for comprehension
and conceptualizarion
of music generally and
regularly, and this would give us une way of showing explicitly how musical
meaning is embodied. For this to be dircctly relevant to 'gesture', however, it
would have to be shown how exertions in the vocal modality are felt in other
modalities. 1 addrcss this in thc next scction.

Gopyr .ihtodm i1cn

50

Music and Gesture

Grasping Gestores Within and Across Modalities

If the physicality of gestures were grasped only within the same modality in which
they are produced, their meaning would have quite a different character than 1
believe they do in fact have. When we hear, say, the sound of a violin, the mimetic
hypothesis holds that we comprehend it in three ways: (1) via (covert) imitation of
the arm and finger movements; (2) via subvocal imitation of the sounds produced;
and (3) via an amodal, visceral imitation of the exertion dynamic evident in the
sound (a pattem of exertions that would produce the same or similar sounds). The
intra-modal imitation will vary according to experience: a violinist, for example,
will comprehend the finger and arm movements of a violin performance differently
than a musician who has never played a string instrument, who in tum will
comprehend these differently than someone who has never played a musical
instrument. Yet even having never played the same instrument or any instrument at
all, we will automatically have sorne idea of what it must feel like to move one's
fingers and arms in a certain way.
1 have already discussed the cross-modal example of subvocal imitation,
but there is another kind of mimetic comprehension that is akin to subvocalization
and this is the comprehension of one instrument (or voice) in terms of another
instrument - for example, representing violin sounds in terms of the motor actions
that produce essentially the same sounds on the piano. 1 believe that this is more
leamed than innate, and it may also be more intentional and less automatic than
other kinds of representations, but it is common enough among musicians that it
ought to be considered here. The ability to represent one modality in another is
possibly (or perhaps likely) dependent upon an amada/ representation of the
exertion dynamic that would produce a sound or a sound pattern in one domain or
another, and this is the third kind of imitation listed above. This is a very difficult
matter to discuss because the relevant feelings are ineffable and, related to this,
their location is invisible. But it is a visceral response to music which is not located
in either the limbs or the voice and yet is fundamental to the embodied foundations
of musical meaning. To understand the roles of these various kinds of imitation,
consider the gesture that opens the final movement ofBeethoven's Violin Concerto
(Example 3.1).

Violin
p

Example 3.l

-~

==~

Beethoven, Violin Concerto in D Major, Op. 61, 111,bars 1-4

Let us focus on the initial two-note gesture. According to the hypothesis,

this event (or event pair) would motvate imitation of the arm and finger
movements of the violinist (whether the violinist were seen and heard or heard
only, or whether the performance were only recalled or otherwise imagined), as

Hearing, Feeling, Grasping Gestures

51

well as subvocal imitation of the pitch, contour, rhythm, accents (phenomenal and
agogic) and dynamic level. But I believe that most of us would also feel something
that was not located in either the limbs or the voice - something in the gut that
somehow matched the energy pattern of the music. This might well be manifest in
the modalities of toe-tapping, head-bobbing and/or conducting, but none of these
are specified by the music (the violin playing). The question of where in the body
an amodal representation might be located is irrlportant, and it may be that this is
only a phantasm resulting from the fact that embodied representations can be and
are manifest in any of several modalities.7 But the more crucial point here is that,
according to the mimetic hypothesis, a musical gesture motivates imitative
representations that are not confined to the modality in which they are produced
(for example, the finger and arm movements of a violinist). This means that a
gesture has a meaning which is at once in accord with its mode of production and
transcendent of its mode of production. The melodic 'sigh', for example, retains its
vocality while being comprehensible as a sigh in any instrumental medium. It is
not sigh-like only by exterior (acoustic) association with its origins in vocal
repertoire, because even in an instrumental realization it is comprehended in part
via subvocal imitation: it is sigh-like because it is comprehended by the sighproducing medium of the voice. At present, however, although the mimetic
hypothesis shows that comprehension is cross-modal, it can only suggest that there
might be an amodal, visceral representation.

Robert Hatten and the Mimetic Hypothesis

If the mimetic hypothesis has relevance for theories of musical gesture, then there
is no better place to demonstrate this than in the context of Robert Hatten's work.
In this section I consider how the mimetic hypothesis bolsters the claims and
explanatory power of Hatten's proto-theory as set forth in a series of online
lectures (Hatten 2001).
In Lecture 2, 'Embodying Sound: The Role of Semiotics', Hatten offers a
set of eight 'presuppositions for a semiotic theory of gesture'. Number 1 reads:
'Gesture is movement interpretable as a sign, whether intentional or not, and as
such it communicates information about the gesturer (or character, or persona the
gesturer is impersonating or embodying).' If applied to danceor theatre, there is no
immediate difficulty here: the gestures referred to are those of the gesturer
(including gesturer-as-character/persona). But when we speak of musical gestures,
to what extent are we speaking of the gestures of the performers? Without question
a performer's gestures communicate information about the gesturer (the
performer), but the role of the musical performer in reproducing signifying
gestures is different than that of other performers. The same musical gesture - for
example, the 'melodic sigh' - can be produced by a great variety ofphysical means
- for example, on the violin, the oboe, or the piano - and yet be classified the same
(as a 'melodic sigh'). These various modes of performance of what we take to be
the same thing must then have something in common, and we ought to be able to
say precisely what this is. If we focus on music's acoustic features (pitch, contour,

Muste and Gesture

52

rhythm) thcn 1scc lcss scnsc in using tcrms such as 'sigh' and 'gesturc'. In thc case
of the 'sigh , we might say that in each performance medium it sounds like a sigh.
This is true enough, and an exterior mapping - this event sounds like that event
sounds - might suffice to account for the metaphoric tcrm 'melodic sigh', \Ve
could also specify thc identity in terms of shared acoustic features in order to
cxplain the rcasoning. But 1do not fiad this satisfactory. for whcn 1 hcnr a 'sigh' 1
feel a sigh, or somcthing very much like the feeling of a sigh. 1 believe that this
feeling motivares and grounds the meaning ofterms like 'sigh' and 'gesturc', and 1
believe we ought to seek a theory that accounts for the viscerally afTective
dimension of musical meaning.
\Ve might apply Johnsoa's theory of image schemata (Johnson 1987) and
note that cach mclodic sigh performcd by vnrious instrumcnts maaifcsts. somchow,
thc samc irnagc schcrna; and wc rnight furthcr note that imagc schcmata are
grounded in ernbodied experience, and that this ernbodied grounding is thus the
basis for using the terms 'sigh' and 'gesturc'. Such an application of Johnson's
work gives usa very good conjecture, but we would still need to specify the means
by which musical gestures are ernbodied. This, thcn. is preciscly where thc
mimetic hypotbesis bccomcs relevan! for gcsture theory. First, sincc thc tcrm
'rnclodic sigh' makcs seosc in sorne measurc as a sigh, regardlcss of the
performance medium, and since the term 'musical gesturc' makes sense in sorne
measure as gesture, regardlcss of the performance medium, thcn we must considcr
what ihese have in common as sighs and other gestures. Second, since the acoustic
features by themselves motivare only a superficial understanding (this sounds like
tltat), thosc of us interested in embodied meaning must explain how events called
'rnelodic sighs' and 'musical gesrures' fee/ like gestures and sighs. According to
the mimetic hypothesis, events that wc call 'gcsrurcs' and 'sighs' not only souod
alike, but thcy also fccl alikc because thcy are ali comprehendcd in part via
mimctic participation. Specifically, cornprehcnsion of a mclodic sigh involvcs thc
following (in various rneasures, in both real time and rccall):
tire vio/in: ( 1) imitation of the fingering and the bowing; (2) subvocal
imitation of the musical sounds produced (the rwo-note dcscent, likely sbaped in
sorne mensure by timbre aod dynamic level); and (3) amodal. visceral imitation of
thc exenion dynamic of the evcnt.

011

011 tite oboe: (1) irnitation of the fingcring, cmbouchurc and blowing; (2)
subvocal imitation of the musical sounds produced (thc two-note dcsccnt); and (3)
amodal, visceral imitation of the exertion dynamic of the event,

on the piano: ( 1) imitation of tbe finger and arm movernents; (2) subvocal
imitation of the musical sounds produced (thc two-note dcsccnt); and (3) amodal,
visceral imitation of thc cxcrtion dyoamic of the cvcnt.
Notice that only the superficial medium difTers(the fingcrings and so forth), which
is not where the essence of 'sigh' tics. Rcgardless of thc performance medium,

Capyri<;bled rnot ri

Hearing. Feeling. Grasping Gestures

53

each is comprchendcd via subvecal imitarion, and in this way each not only sounds
likc a sigh butfee/s like a sigh.
11is worth noting here that in each case the fingcr movernents are nlso
imitated, and we ought to ask why this docs not motivare a compcting
conceptualization. There are severa! factors that work against a digitally based
conceprualization. Onc is that there is not necessarily a one-to-one correspondence
in each case, since a stepwise dcsccnt on the oboe might involvc a combinarion of
severa! fingers. Anotber reason is that the finger acrions here are out of proponion
with thc most salicnt features o the cvent - thc duration, contour, timbre, dynamic
levcl and harrnonic/tonal context - and are instead rcsponsible for only the vertical
dimension of the contour. Nonetheless, in another context the finger acrions could
easily bccome much more salient, as in a virtuosic passage work, and in sucb a
case a different sort of conceptualization would be motivated - sucb as the
rnctaphors of 'passagc' or 'run' (including thc corrente).
Returning now to Hauen's prcsuppositions. the first ofthesc also leads to
anothcr importan! qucstion: how might gcstures bccome abstractcd into gesturcs of
a charactcr or a persona? According 10 ihc mirnctic hypothcsis, we experience
pauerns of exenion by way of mimetic participation. and in this way it is as if we
are acting - acting in a way that is more or less isomorphic with the soundproducing actions heard (and sccn). In conceptualizing these patterns of exertion.
wc rnap these amo our own experiencc of making similar cxcnions. and among thc
most straighrforward rnappings is that omo gestures: wc do not fcel only abstrae!
sensarions of cxertion; wc also undcrstand these as the intentional, expressive
gestures that we have madc and have scen made in other domains of experience.
One result of this mimetic participation is that we enact the role of a character or
persona - that oan ideal gesturing perforrner (at once making music and not, since
tbc excrtions are not confined to tbe spccific domain of sound production), At tbe
same time. howcvcr, thc othemcss of the music rcmains, sincc wc are not thc
principie (original) souree of the sounds with wbich we are engaging. The fact that
the musie is produced by a source exterior 10 ourselves may be what motivares a
projcction of our rnimetic musical agency outwards, perhaps towards this exterior
source. 8111sincc this agency cannot be identified directly with the actlons of the
performers, it remains nn ideal agency 1ha1 is noi-us. We call this agent 'the
music', and its genesis might follow this pattern: (1) the sound-producing actions
ofthe pcrforrncrs are (2) henrd/scen and imitatcd by a listcner, which (3) morivatcs
a participatory agcncy within the listcncr, which (4) is thcn projectcd ourwards
towards the original sourcc but which (5) cnnnot be identified directly with this
source and so rcmains ideal.
The only oiher of Hallen's cight prcsuppositions that 1 wish to consider
here is that regarding posture (presupposition 8), which he says 'rnay be considered
as gesture "under a fermata." A "frozen rnotion" or pose may revea! the energy and
affect with which it is invested. Sucb momcnts can be among the most powerful in
music. Part of their affect results simply frorn the anticipation of what will follow.
hui part ofit results from the feeling ofbolding an imagined pose. According to the
mirnetic hyporhesis, in the case of a hcld mamen! we fecl vicariously thc cnergy
rcquired to sustain a sound - ar 10sustain a silcncc - in sorne cornbination of intra-

54

Muste and Gesture

modal, cross-modal (including particularly thc subvocal) and amodal (visceral)

imitation. The scnsation of holding a pose would result also from the same sense of
mimctic agency described above: among the gesrures made by our musical persona
(our own, or that projected away from us) are poses which may be held for
expressive purposes just as in othcr domains of expcricncc.
Let me apply and extcnd this to an exarnple from Hattcn 's analysis of the
opening of Schubert's late A major sonata, D. 959. 1 should note that the topic
from which the following excerpt derives is the question of 'resonance' as a
gcsrurc; however, for my purposes hcre 1 am interested in thc pcrsonification and
locomotivc metaphors. The analytical excerpt refers to the beginning of thc first
rnovement (Exarnple 3.2).

11
-"-"- 'll"'
1

.,

,._

ft,

. ft,*

I~

,,.---...

'

Example 3.2 Schubcrt,

==-
-

'

;1

=-

.,;.

'

'

w'~it

Piano Sonata in A Major, D. 959, 1, bars 1--0

Hallen (2001: Lccrure 4: 3, emphases added) writes:

Here, the le hand has the (...) gestural motive onc, and its 'stoic
hcroisrn' mny be scen to 'resonare' ihe right-hand's sustained chords,
which [...) strive in stepwise ascent until rclaxing into thc half cadence
on thc dominont in bar 6. The determined wil! of a persona is clearly
implied and gesturally projected.
1 havc alrcady described the role of mimctic participation in the
pcrsonification of music, but it is one thing to talk in general about gestures and
anothcr to talk spccifically about the strivings of a persona in a stepwisc asccnt,
That is, once we havc a musical persona, how do we spccify its acrions within the
metaphorical terrain of musical space? We have 10 start by rejecring the view that
therc is a bass line rhat ascends by step. along with the view that a personification
of this motion is a rnetaphoric interpretarion of the objcctivc facts. As Marion Guck
( 1994) has pointcd out, even such quasi-objective obscrvations are analytical
cuons based on a metaphoric understanding. Our musical fines, ascents and steps
are mctaphoric conccptualizations of relations among tones in tcrms of spatial
relations. Thc conception of musical motion and space and the conception of
musical personae are both products of the sarne embodied metaphoric reasoning, so
let me sketch an account of this embodied reasoning.
According to the mimetic hypothesis, when we hcar the opening of this
sonata we participare vicariously with the production of the music - specifically,
thc bass linc and the sustained chords - and our participation takes threc fonns:

o: 'Yr.1hl,

Hearing. Feeling. Grasping Gestures

55

imitation of the ann gesrurcs, subvocal imitation of the sounds produced, nnd a
visceral exertion that matches the excrtion dynamic manifest in thc music. Again,
the degree and combination of participation will vary from pcrson to pcrson, so
that more properly 1 should say that thcse are ihe specific kinds of participation that
the music affords and motivates. With respect to the bass-line chords, we feel the
rwo-part arm gcsturcs and thcir gradual progression 'up' the keyboard (that is, their
progression 10 thc right), Wc also fccl what it would be like to sing the sounds that
are produced by these gesrures, including the difference in tcrms of what wc cali
'pitch', which in this case gives us 'asccnt'. Finaliy, we also fecl an amodal,
visceral excrtion that is in this case more or less isomorphic with the imagined ann
and vocal excrtions. As in other domains of experience, the embodied feelings
motvate conccptualization in tenns of basic, concrete experience (Lakoff &
Jobnson 1980, 1999; Johnson 1987), and these include 'gesture as wcli as the
quasi-objective spatial conceptions. Tite pattern of cxcrtion in the Schubert, within
and between the succcssive lefl-hand gcsturcs, activares thc conceptual metaphors
greater is higher and states are locations, along with the related metaphors of
change o/ state is motion between locations (change is motion} and difference

between states is distance between locations (differencc is disumce).8

The conceptual metaphor greater is higher is onc of our most common
metaphors, undcrlying conccpts such as 'highcr' priccs, 'higher' quality, 'highcr'
cducation and 'higber' frequencies, ali of which are literaliy greater and
metaphorically 'higher'. In tite present case, this mctaphor is activated by the
greater and lesser exertions fclt in imagining producing the piano sounds by way of
ubvocalization, where the phenomenon that we call 'ascent' norrnaliy involves
greater exenion. The state-locations meraphors are also very cornmon - giving us
emotional, financial, musical and othcr mctaphoric locarions - and in this context
thcy givc each musical evcni a location, with small, discrctc differcnces conccived
as 'steps' (on successive downbeats in the bass). and greatcr, discrete dilTcrcncc
conceived as 'leaps' (the crotchet octave leaps in each bar). These mctaphors
combine to give us musical locarions, distances and motion within the vertical
musical dimensin. while the horizontal dimensin is the product of these and
other conceptual rnetaphors that simiJarly becorne rclevant va mimetic
participation, Although thc dctails of our conccptions of musical rnotion and space
cannot be cxplored Itere, thc crucial clnim is that ali such spatial conceptions begin
with the feeling of an exenion dynamic that rcsults from mimctic participarion.
Once felt, and then conceived in terms of kinds of motion, we can then srep back
and adopl a third-person analyrical perspccrivc, positing an ideal agcnt in our place.
lt is this musical agent that thcn exerts in this way and that way, performing
gesturcs and moving in various ways through and between meiaphoric locations.
Our conceptual and linguistic habits thcn perpetuare the analytical perspective - for
example, the common fonnulation, 'here thc rnusic does this' - but language
depcnds on the ernbodicd mctaphoric rcasoning 1 have skctched (both originally
and as a reinforcemcnt of language's meanings). lt makes scnsc to describe thc
opening of this Schubcrt sonata in terms of 'striving' and otherwise 'gesruring ' not
because we cmpathize with the actions of sorne already-given musical persona, and
not because there is a musical motion already given, but because we feel the

56

Music ond Gesture

excrtions involvcd in thc sound's production, and the exertion dynamic that we fecl
is more or lcss isomorphic with the fecling of striving, progressing and gesturing.

Hcaring, Feelng, Comprchcnding, Concciving and Grasping 'Gcsturcs'

My primary intcrest gcncrally is in how music engages us and in the rclationship
berween experience and conccptualization. Although ali of our rnetaphoric
conceprions of motion and space can be understood to bear thc residue of their
embodied origins, 'gcsiurc' secms to foreground Lhiscmbodiment more than most.
My final point in this chapter concems the place of 'gesturc' in music
cpistcmology and thc diffcrcnccs bctween Lhis and othcr conceptualizations in
terms ofthe kinds ofknowlcdgc thcy hclp cstablish.
Our cmbodicd engagcmenl wilh music afTords an intimatc, visceral and
intuitive way of knowing music. A good portion of Lhis type of knowlcclgc is
ineffable (RafTmaa 1993), but our embodied expericnce also motivares
concepruatizations whose mcaaings remain tied to this embodied cxpcricnce by
wny of the pauerns of cxcrtion shared betwecn music and other domains of
cxperience, 'Gesture' is one such conccpt, and whilc it might be extended 10
includc largc-scale relationships, and perhaps to other less immcdiate events and
rclations, the more immcdiatc cvents and relations that are roughly coextensivc
with 'motives' and 'figures' are in a sense more viscerally cngaging and thus more
rneaningful." These irnrnediate events nnd relations, as in the opcning of the
Schubert A major Piano Sonata, are those which we can most rcadily 'grasp'
(comprchend) and thus know in a way that is different from large-scale
rclationships. Thc use of thc term 'gesture' highlights this diffcrencc, but what
advantage is there in the use ofthis tcrm over 'figure' and 'motive'?
lf one examines the concepts in English related lo knowing, one quickly
finds that thesc are dominated by visual terms (Johnson 1987; Swcetser 1990).
aturally, this extends to music analytical discourse (Cox l999b; Johnson 2002),
where we 'reveal ', 'show', 'clarify' and so forth. lndeed, rhe very word 'theory' is
itself a visual term (rneaning 10 show', with the same root ns theatre). Siace
human perception is dominated by visual iuformation, this visual bias in our
cpistemology is not surprising; and since most of the non-visual perccptual
infommtion we den) with also has a visible sourcc - invisible smclls, lastes and
sounds come from othcrwisc visible objccts - in most cases this bias <loes not
present a problcm. But for music epistemology this prescnts a considerable
problern, the gis: of which is that, that which is rnost crucial about musie is aeither
visible nor visualizable (without tuming it into somcthing other thnn musical
sound), nnd this leaves music episternology in an awkward position: if an
undcrsianding of somcthing cannot be shown, then it cannot align with our more
general epistemology, which would mean that musical knowledge would havc to
be clnssificd ns a spccial kind of knowlcdge - a more intuitivc kinclof knowledge
that would not carry the same clout as knowlcdge in other fields. By focusing on
music's 'structure' and on 'figures' we make visualizable representarions, which

CopyrllJltlodmal "

Hearing, Feeling, Grasping Gestures

57

have the advantage of aligning with our more general epistemology but which also
have the disadvantage of objectifying a non-objective, embodied experience.
But we have another way of conceptualizing knowledge in English, and
that is in terms of 'grasping' (Johnson 1987; Sweetser 1990; Cox 1999b).
'Comprehend', 'conceive' and 'perceive' are each grounded in the experience of
grasping. Much like our visual bias, the importance of grasping objects in the
history ofour species makes it unsurprising that 'grasping' should structure our
conception of conceiving and comprehending. But grasping has a different feel
than seeing: grasping is more immediate; and while it still objectifies that which is
grasped, our knowledge of the thing grasped is more intimate and visceral than it is
when we simply regard it. The large-scale 'structure' of a musical work is perhaps
not easily graspable, but in contrast to this, we know a work more intimately in our
moment-to-moment experience of its more easily graspable events and relations
(Levinson 1997), and these events and relations are at the level of motives, figures
and gestures. But 'motive' and 'figure' do not reflect that quasi-tangible feature
indicated by 'gesture', which focuses our attention on a more physically intimate
understanding of how the music works.
To see this - to feel this - compare the following conceptualizations of
the opening two-note event of the finale of the Beethoven Violin Concerto: (1) a
figure, (2) a motive, (3) a leap and (4) a gesture. 'Figure' suggests something that
is external to us and fixed. 'Motive' highlights the dynamic of change, conceived
as motion, but still leaves the event(s) as extemal. 'Leap ' suggests a more
embodied sense of motion, but it is too big for us to feel directly in terms of our
own experience of leaping: to leap requires a much greater exertion than the
exertion reflected in this opening event; it has an analogous dynamic, or exertion
'contour', but it is out of proportion. However, 'gesture' suggests not only an
analogous exertion dynamic, but one of the same proportion, and this isomorphism
affords a closer comparison with our own embodied experience.
'Gesture' seems to match best the level at which we grasp (comprehend)
music most viscerally and intimately, and in this way it highlights a kind of
musical knowing that is distinct from our more visual and quasi-objective
conceptualizations. By focusing on musical gestures we draw attention to a crucial
area of musical meaning, and by understanding how musical gestures are grasped
and conceived we strengthen our understanding of how musical meaning is
constructed.

Notes
1. This is a slightly different formulation than that in Cox 2001.
2. One exception is Clynes (1977), who offers a particular kind of empirical
evidence which I discuss below. Lidov (1987) bases his arguments on Clynes's
work.

58

Music and Gesture

3. A good sample of such research as it bears directly on music can be found in

Papousek (1996), Malloch (1999/2000) and Trevarthen (1999/2000).
4. Much work has been done in this area; a representative sample includes Gallese
et al. (1996), Rizzolatti et al. (1996), Gallese and Goldman (1998) and Fadiga et al.
(1998).
5. 1 take subvocalization to include both silent vocal imagery and sotto voce
rehearsal and imitation.
6. Evidence for this can be drawn from Baddeley and Logie (1992); Smith,
Reisberg and Wilson (1992); Smith, Wilson and Reisberg (1995); Vaneechoutte
and Skoyles (1998).
7. Daniel Stem (1985) has argued for amodal perception in infants. Although he
suggests that we eventually outgrow this, it may be that amodal perception
becomes disguised by the gradual dominance o modally specifc perception.
8. For a fuller explanation of how this and how other conceptual metaphors
structure musical thought, see Cox l 999a.
9. Compare Hatten (2001), Lecture 4: l.

References
Baddeley, A., & Logie, R. (1992). Auditory Imagery and Working Memory. In D.
Reisberg (ed.), Auditory imagery (pp. 179-97). Hillsdale: Lawrence
Erlbaum.
Carpenter, P. (1967). The Musical Object. Curren! Musico!ogy 5: 56-87.
Clifton, T. (1983). Music As Heard: A Study in Applied Phenomenology. New
Haven: Yale University Press.
Clynes, M. (1977). Sentics: The Touch of Emotions. New York: Anchor
Press/Doubleday.
Cox, A. (1999a). The Metaphoric Logic of Musical Motion and Space. PhD
dissertation, University of Oregon.
Cox, A. (l 999b). Imagined Meaning in Immediate and Reflective Musical
Experience. Music Perception 16/4: 467-73.
Cox, A. (2001). The Mimetic Hypothesis and Embodied Musical Meaning.
Musicae Scientiae 512: 195'f209.
Cumming, N. (1997). The Subjectivities of 'Erbarme Dich'. Music Analysis l 611:
5-44.
Fadiga, L., Buccino, G., Craighero, L., Fogassi, L., Gallese, V., & Pavesi, G.
(1998). Corticospinal Excitability is Specifically Modulated by Motor

Hearing, Feeling. Grasping Gestures

59

Lmagery: A Magnetic Stimulation Study. Neuropsychologia 37/2: 14758.

Gallese, V., Fadiga, L.. Fogassi, L. & Rizzolaui G. (1996). Action Recognition in
the Premotor Concx. Brain 119: 593-609.
Gallese, V. & Goldman, A. (1998). Mirror Ncurons and the Simulation Theory of
Mind-Reading. Trends In Cognitive Sciences 2112:493-50 l.
Gathercole, S. E. & Baddeley, A. D. ( 1993). Working Memory and Language.
Hillsdale: Lawrence Erlbaum.
Gibson, E. J. & Levin, H. (1975). The Psychology of Reading. Cambridge, Mass.:
MJT Press.
Guck, M. ( 1994). Analytical Fictions. Music Theory Spectrum 16/2: 217-30.
Hallen, R. S. (2001). Online Lecturcs on Music and Gesture. Cybcr lnstitutc of
Semiotics (http://www.chass.utoronto.ca/cpc/srb/cyber/hatout.html).
Haueisen, J. & Knosche, T. R. (2001). lnvoluntary Motor Activity in Pianists
Evoked by Music Pcrception. Journal o/ Cognitive Neuroscience 13/6:
786-792.
Johnson, B. (2002). Unsound lnsights. ln K. Karki, R. Leydon & H. Terho (eds.),
Lookiug Dock. Looking Airead: Proceedings o/ the 11th Corference o/
Tire lntemational Association for tire Study o/ Popular Music (pp. 70412). Turku, Finland: IASPM-Norden.
Johnson, M. (1987). The Body in tire Mi11d: Tire Bodily Oasis o/ Mea11i11g.
lmogination, and Reason. Chicago: University ofChicago Press.
Lakoff. G. & Johnson, M. ( 1980). Metaphors We Live By. Chicago: University of
Chicago Prcss.
LakofT,G. & Johnson, M. ( 1999). Philosophy i11 tire Flesh: Tire Embodied Mi11d
and its Challenge to Western Thought. New York: Basic Books.
Levinson, J. {1997). Music in tire Mo111e111.
lthaca: Comell University Prcss.
Lidov, D. (1987). Mind and Body in Music. Semiotica 6611: 66-97.
Malloch. S. N. (1999/2000). Mothers and lnfants and Cornmunicarive Musicality.
Musicae Scientiae Spccial lssuc: 29-57.
Mead, A. ( 1999). Bodily Hearing: Physiological Metaphors and Musical
Undcrstanding. Journal o/ Music Theory 43/1: 1-19.
Papousek, M. (1996). lntuitivc Parcnting: A Hiddcn Source ofMusical Stimulntion
in lnfancy. In l. Deligc & J. Sloboda (eds), Musical Begi1111i11gs:
Origins and Development o/ Musical Co111pe1e11ce
(pp. 88-112). Oxford:
Oxford University Press,
Raffman, D. (1993). Language, Music, and Mind. Cambridge. Mass.: M!T Press.
Rizzolani. G., Fadiga, L.. Gallese, V. & Fogassi. L. (1996). Prernotor Cortex and
the Rccognition of Motor Actions. Cognitive Brain Research 3/2: 13141.
Smith, J. D., Rcisbcrg, D. & Wilson, M. ( 1992). Subvocalizarion and Auditory
lmagery: lnteractions between the lnner Ear and lnncr Voicc. In D.
Reisberg (ed.), Auditory lmagery (pp. 95-119). Hillsdale: Lawrence
Erlbaum.
Smith, J. D., Wilson, M. & Reisberg, D. (1995). The Role of Subvocalization in
Auditory lmagery. Neurospychologia 33/11: 1433-54.

60

Music and Gesture

Spencer, H. ([1857] 1951). The Origin and Function ofMusic. In his Literary Style
and Music (pp. 45-106). New York: Philosophical Library.
Stem, D. N. (1985). The Interpersonal World of the Infant. New York: Basic
Books.
Sweetser, E. (1990). From Etymology to Pragmatics. Cambridge: Cambridge
University Press.
Trevarthen, C. (1999/2000). Musicality and the Intrinsic Motive Pulse: Evidence
from Human Psychobiology and Infant Communication. Musicae
Scientiae Special Issue: 115-211.
Vaneechoutte, M. & Skoyles, J. R. (1998). The Memetic Origin of Language:
Modem Humans as Musical Primates. Journal of Memetics:
Evolutionary
Models
of
Information
Transmission
2
(http://www.cpm.mmu.ac.uk/jomemit/l 998/vol2/vaneechoutte_m&skoy
lesjr.html).
Walton, K. (1993). Understanding Humor and Understanding Music. In M. Krausz
(ed.), The Interpretation of Music: Philosophical Essays (pp. 259-70).
Oxford: Clarendon Press.
Walton, K. (1997). Listening with Imagination: Is Music Representational? In J.
Robinson (ed.), Music and Meaning (pp. 57-82). Ithaca: Comell
University Press.