TIGER C. ROHOLT

Musical Musical Nuance

“I remember when, I remember, I rememberwhen I lost my mind. There was something sopleasant about that place, even your emotionshave an echo in so much space.”1 With every word,Gnarls Barkley’s Cee-Lo Green is on the vergeof falling off the beat from the back end, pullingagainst our timing expectations. In a different con-text, this is the same effect Frank Sinatra achieves:“Fly me to the moon; let me, swing among thosestars. Let me see what spring is like on, Jupiterand Mars. In other words, hold my hand. In otherwords, baby kiss me.”2

A musical nuance is typically defined as a noteperformed slightly raised or lowered in pitch orslightly early or late in time (music theorists andcognitive psychologists prefer the term ‘expres-sive variation’). A cellist or vocalist may performa raised A-natural that we perceive to be slightlyhigh yet not high enough to perceive as an A-sharp; a drummer or pianist may strike certaineighth notes slightly late, and we may perceivethem to be late yet still perceive them to be eighthnotes (as opposed to dotted eighth notes).3 Inwhat follows, I emphasize that a musician per-forms this kind of minute variation for a reason,namely, in order to alter the way the music sounds;for example, nuances may alter a listener’s per-ception of musical structure, “brighten” a chord,result in rhythmic tension (as in my opening exam-ples), or result in grooves such as a swing rhythm’sfeel of being “in the pocket” or of leaning for-ward (a groove is, roughly, the perceptual feel ofa rhythm).4

Traditionally, regarding classical music, nuancesare taken to be features only of performances.5

Although nuances are obviously important in thisrole, I believe that nuances are even more impor-tant in rock music (broadly construed to include

rockabilly, hip-hop, et cetera). Ted Gracyk has ar-gued that rock musical works just are the record-ings.6 Since nuances are ubiquitous in recordings,they are features of rock musical works them-selves. As Stephen Davies writes, regarding Gra-cyk’s ontology, “If the primary works in rock arerecordings, then these works are very thick withproperties. Every aspect of the sound capturedby the recording technology is constitutive of thework.”7 Nothing in this essay rests on that view; Imention it here, by way of introduction, in orderto highlight the likely importance of obtaining acorrect account of musical nuance for examiningrock music.

Diana Raffman’s account of musical nuance,put forward in her Language, Music, and Mind,is the predominant account of musical nuance inanalytic philosophy of music.8 In what follows, Imove through a critical analysis of Raffman’s viewtoward a suggested reorientation of the examina-tion of this phenomenon.

i. raffman on musical nuance and ineffability

Raffman adopts a cognitivist view of perception,drawing from Jerry Fodor’s modularity of mindthesis: musical experience results from a multi-leveled structure of computations over mentalrepresentations. She understands the final clus-ter of computations in terms of Fred Lerdahl andRay Jackendoff’s Chomsky-like generative gram-mar of tonal music (“M-grammar”), which offersan explanation for an underlying competence thatallows listeners to make sense of a string of tones;it is an account of just what a listener “does,”subpersonally and unconsciously, to organize themusical surface into structured music.9 But most

The Journal of Aesthetics and Art Criticism 68:1 Winter 2010c! 2010 The American Society for Aesthetics

2 The Journal of Aesthetics and Art Criticism

relevant for our purposes is the shallowest levelof conscious perception, where our pitch sensa-tions are disorganized and chaotic. Prior to be-ing input to the M-grammar, our sensations gothrough an organizing phase, which Raffman clari-fies by drawing upon the work of the psychologistsJohn Anderson, J. J. Bharucha, et al. on schemas.Schemas are mental organizing structures thatare conceived as enduring, long-term represen-tations; they act as grids or templates that sort theincoming pitch sensations into type-identifiablecategories.10

Schemas transform our initial sensations intopitch-time events; they transform uncategorizednuance pitches into tokens of chromatic pitches,such as C-natural or F-sharp. These chromaticpitches are the pitch input to the M-grammar, rep-resentations “instantiating the sequence of pitchesspecified by the score . . . the input to the M-grammar is a mental recovery of the score. Callthis the mental score” (LMM, p. 64). “Thus themental score—the shallowest grammatical [struc-tural] level of representation—is already an ab-straction from a still shallower level at whichthe nonstructural nuance values are recovered”(LMM, p. 67); the latter level is what Raffmancalls the nuance-level. But note that our abilityto recognize transposed melodies suggests thatwe hear melodies not in terms of fixed pitch cat-egories, but relationally; our schemas must sortnuance pitches relationally rather than in termsof fixed pitch categories. A perceived melody ismentally represented first and foremost as a se-quence of chromatic intervals. As Raffman writes,“My point is that what transforms your N-level[nuance-level] into a mental score is the represen-tation of incoming acoustic events as instantiatingcertain relationships—specifically, relationships ofthe kind embodied in your C-interval [chromatic-interval] schema” (LMM, p. 72). Thus, accordingto Raffman, musical nuances just are those per-ceptions of pitch that we are conscious of but havenot yet schematized, that is, pitches that are per-ceived at the shallowest level of conscious repre-sentation.

Raffman begins the book with quotations fromStanley Cavell, Susanne K. Langer, and JohnDewey, which she interprets as having to do withart’s ineffability. Raffman writes, “Despite con-siderable differences in ideology, objective, andstyle, these theorists join in giving voice to oneof the most deeply rooted convictions in mod-

ern aesthetics: our knowledge of artworks is, insome essential respect, ineffable. In apprehend-ing a work of art, we come to know something wecannot put into words” (LMM, p. 2). The main ob-jective of Raffman’s book is to “develop a cogni-tivist explanation of musical ineffability” (LMM,p. 3). “To put it simply,” she writes, “I want to seehow these empirical disciplines [psychology, psy-cholinguistics, et cetera] might explain the appar-ent fact that conscious musical experience givesrise to claims of ineffable knowledge” (LMM,p. 3). When all is said and done, Raffman con-cludes that one central kind of musical ineffabilityconsists of unschematized pitches. We can discrim-inate many more pitches than we can conceptual-ize; the pitches that we cannot conceptualize justare musical nuances. Thus, Raffman’s ineffabilityclaim comes through her account of musical nu-ance: she develops an account of musical nuanceand argues that nuances are ineffable.11

To be sure, Raffman’s ineffability claim doesfollow directly from her characterization of nu-ances: musical nuances are ineffable becausenuances are perceived at a level prior to schema-tization. Our schemas make our perceptual infor-mation effable; schemas enable us to rememberand reidentify that which they organize. Our lim-itations regarding categorization are due to lim-itations in perceptual memory: “we can’t namethem [nuances] because we can’t recognize them,and we can’t recognize them because we can’t re-member them” (LMM, p. 84).12 Furthermore, ourinability to report nuances rests upon this ineffa-bility due to lack of schemas. Raffman succinctlyexpresses the point in a section heading, “WhyWe Cannot Report the Nuances: No Verbaliza-tion Without Schematization” (LMM, p. 83).

ii. direct versus indirect description

For Raffman, achieving the goal of explaining acentral kind of musical ineffability rests on ac-curately describing our perceptual experiences ofmusical nuances. I want to begin the critical por-tion of this essay by calling into question the ad-equacy of Raffman’s description of nuances. InW. E. Kennick’s 1961 essay, “Art and the Ineffa-ble,” he criticizes the ineffability claims of Dewey,Langer, and D. W. Prall (two of the same philoso-phers Raffman begins by quoting).13 Kennickdraws a distinction between directly and indirectly

Roholt Musical Musical Nuance 3

describing a feeling or quality; direct description isessentially naming, while indirect description in-volves characterizing the circumstances and con-text in which a feeling is experienced. While di-rect description typically falls short of adequatelycharacterizing subtle feelings, indirect descriptionis more effective. Objecting to Langer’s claim thatfacts about feelings cannot be depicted by lan-guage (discursive symbols), Kennick writes,

Mrs. Langer makes the mistake, often made in such dis-cussions, of supposing that describing a feeling is thesame as naming a feeling. This is because she takes asthe prototype of all descriptions the sort of “direct” de-scription we frequently give of people and objects, e.g.,the sort of description one might find on a “wanted”poster in a post office: “Height, 5"11""; weight, 170 lb.;color of hair, dark brown; eyes, blue; complexion, ruddy;small horizontal scar over the right eye.” This kind of de-scription can be given of feelings, but usually it is not,either in daily life or in novels. More frequently we em-ploy a sort of “indirect” description which includes adescription of the circumstances in which the feeling isfelt.14

Kennick borrows this distinction between directand indirect description from Ludwig Wittgen-stein, who, in The Brown Book, considers thefeeling or experience of familiarity.15 Wittgen-stein points out that there are different experi-ences of familiarity. (To anticipate where I amgoing with this, note the analogy between dif-ferent experiences of familiarity and differentF-sharps.) Wittgenstein claims that in order tocorrectly describe a particular experience of fa-miliarity, we must describe the circumstances orcontext.

Different experiences of familiarity: a) Someone entersmy room, I haven’t seen him for a long time, and didn’texpect him. I look at him, say or feel “Oh, it’s you”—Why did I in giving this example say that I hadn’t seenthe man for a long time? Wasn’t I setting out to describeexperiences of familiarity? And whatever the experiencewas I alluded to, couldn’t I have had it even if I had seenthe man half an hour ago? I mean, I gave the circum-stances of recognizing the man as a means to the end ofdescribing the precise situation of the recognition.16

Following this passage, Wittgenstein refers to adirect description of a table (giving the shape,dimensions, and so forth); he then notes that an

indirect description of the table might be the kindof description one finds in a novel: “‘It was a smallrickety table decorated in Moorish style, the sortthat is used for smoker’s requisites’ . . . if the pur-pose of it is to bring a vivid image of the tablebefore your mind in a flash, it might serve thispurpose incomparably better than a detailed ‘di-rect’ description.”17

If indirect description can render musicalnuances effable, then perhaps we can rejectRaffman’s ineffability claim and simply fault herfor failing to consider indirect description. Afterall, Raffman does characterize nuances in termsof applying names to these perceptual properties,and as we have seen, she then concludes that lim-itations of perceptual memory prevent us fromsuccessfully reapplying these names. Consider thispassage:

Recall the numerical N-pitch names ‘A-natural(1),’ ‘A-natural(12),’ ‘B-flat(17),’ and so forth introduced inpreceding chapters; these could serve perfectly wellfor enunciating the [representational] content in ques-tion. . . . The ineffability of the content of nuance repre-sentations derives not from the absence of terminologyadequate to its verbal expression, but rather from thepsychological impossibility of applying such terms “byintrospection.” (LMM, p. 140)

In a review of Raffman’s book, Georges Rey ex-presses a criticism related to the one we havejust considered.18 Rey does not discuss the effi-cacy of including considerations of context—thatis, he does not seem to have indirect description inmind as a solution (he does not invoke Kennick orWittgenstein)—but he does criticize Raffman forlimiting herself to something like direct descrip-tion. Moreover, he suggests that literary devicessuch as metaphor could be employed to rendernuances effable: “Discussions of ‘ineffability’ of-ten tend to focus on cases where there aren’t sin-gle words for the job and neglect the possibilityafforded by complex descriptions. Where we lackschemas for nuances, why mightn’t we evoke themby exploiting standard compositional resources,including simile, metaphor, and the like?”19 Wewill see that such descriptive devices can, indeed,be effectively employed in an indirect descriptionof nuances.

In sum, our being unable to employ termssuch as ‘B-flat(17)’ due to limitations of percep-tual memory does not necessarily establish the

4 The Journal of Aesthetics and Art Criticism

ineffability of musical nuances, because an indirectdescription may render these seemingly ineffablefeatures effable.

iii. structural objectives of nuances

Are we justified in criticizing Raffman for not con-sidering indirect description? As I mentioned atthe outset, it is crucial to emphasize that a mu-sician performs a nuance for a reason, namely, inorder to alter the way the music sounds, to give riseto some quality or element in the music; call thisthe nuance’s objective. Now, an effective indirectdescription ought to include (at least) (1) a refer-ence to the nuance’s objective, (2) a descriptionof the musical context, and (3) a reference to thenuance itself. Given what I have said so far aboutRaffman’s account, and given Rey’s critical com-ment, you will be surprised to learn that Raffmanherself offers a detailed indirect description of anuance that includes all of the elements I have justmentioned, as well as a thought-provoking use ofmetaphor:

Many fine-grained differences in interval width—indeedthe most interesting and important ones, for our pur-poses—are fully intended expressive features, as whena flutist ever so slightly raises (“sharpens”) an F-sharpsustained over a D-natural across a modulation from Bminor to D major. . . . The flutist’s objective is to widen(“brighten”) the major third between D-natural and F-sharp, thereby emphasizing and strengthening the newkey of D major. (LMM, p. 66)

This is an example of the depth of Raffman’s book,and one reason why it rewards close attention. Un-fortunately, in the very next paragraph, before ac-knowledging and developing the import of this in-direct description, Raffman turns back to the taskof developing her account of musical nuance interms of direct description. Why didn’t Raffmanexplore the possibility that indirect descriptions,such as the one she offers above, render nuanceseffable? What is it about her assumptions that ledher away from believing that nuance objectivesand musical context are relevant to rendering nu-ances effable? I believe an answer can be foundby considering the influence on Raffman of theaccount she adopts of nuance objectives.

In the indirect description above, Raffman ex-plains that the flutist’s objective is “to widen

(‘brighten’) the major third between D-naturaland F-sharp, thereby emphasizing and strength-ening the new key of D major.” For Raffman, theimportant aspect of the objective is the emphasis,the strengthening of the new key. The view of nu-ance objectives that she adopts, the standard view,is that a performer employs a musical nuance in or-der to lead a listener to hear the musical structureas he, the performer, hears it.20 Raffman writes,“As I have repeatedly noted, the performer’s ob-jective is to mold these fine-grained features insuch a way as to communicate his hearing of awork’s structure” (LMM, p. 86).

I want to suggest that Raffman does not take theabove nuance objective to be relevant to describ-ing the nuance because she takes the objective tobe characterizable in terms of structure; the nu-ance objective is categorizable, conceptualizable.She believes that the F-sharp itself is the onlyelement relevant to the musical nuance becauseit is the only element that is not conceptualiz-able. Raffman explicitly distinguishes between thestructural and the nonstructural features of music;one of the defining characteristics is that struc-tural features are conceptualizable, or, in her ter-minology, “type-identifiable”—and the nonstruc-tural features are not.21 (Another example of astructural nuance objective is a nuance’s resultingin our hearing a passage of music in one meterrather than another.) According to my interpre-tation, Raffman sees nothing in the indirect de-scription that can bring us closer to grasping thenuance, nothing that can render it effable, becauseall of the elements—other than the raised F-sharpitself—are conceptualizable; they can be describedin terms of music-theoretic concepts, or terms suchas “emphasis,” which she also takes to be concep-tualizable.

iv. nonstructural objectives of nuances

Let’s be clear about two positions you may adoptat this juncture. You may agree with Raffman thatnuance objectives are conceptualizable, and youmay agree that indirect description is irrelevantto an account of musical nuance; nuances remainineffable. Or, you may conclude that Raffman’sindirect description inadvertently undermines herown ineffability claim; after all, in this indirect de-scription, we are given a very detailed descriptionof the musical context, the nuance objective, and a

Roholt Musical Musical Nuance 5

reference to the nuance itself (that is, you may holdthat she is not wrong about nuance objectives be-ing conceptualizable, but still, she is wrong in herbelief that an indirect description cannot rendernuances effable).

Adopting either position would be prematureat this juncture; we need to turn our attentionto the fact that some nuance objectives are notconceptualizable; to be consistent with Raffman’sterminology I will call these nonstructural nuanceobjectives. The existence of nonstructural nuanceobjectives is, perhaps, especially obvious to thoseaware of the role of nuances in rock music, be-cause, when it comes to the effects of nuances,rock musicians are more concerned with qualitythan structure.22 However, we do not need to drawfrom rock examples to make this point; Raffmanherself mentions a nonstructural nuance objectivein the indirect description we have been consid-ering. Recall the latter portion of the indirect de-scription she offers of the raised F-sharp: “Theflutist’s objective is to widen (‘brighten’) the ma-jor third between D-natural and F-sharp, therebyemphasizing and strengthening the new key of Dmajor.” It is thought-provoking that the metaphorof brightening is placed in parentheses. The bright-ening is surely an effect of the nuance, but it isnot what Raffman typically refers to as a nuanceobjective; it is not merely a matter of hearing mu-sical structure in one way or the other, and it isnot conceptualizable. It is crucial for our purposesthat one consequence of the raised F-sharp is abrightening of the interval. But what should wesay about this kind of nuance objective?

Notice that the second position floated above—indirect description renders nuances effable—nowseems questionable, because this brightening, andother nonstructural objectives, seem descriptivelyelusive. We now find ourselves with something likea new ineffability problem, since F-sharps raisedto slightly different degrees will make the samesegment of the performance sound different: aslightly raised F-sharp in this context may, indeed,emphasize the new key, but an F-sharp(13) and anF-sharp(14) will each emphasize the new key ina qualitatively different way. Music-theoretic con-cepts are too coarse-grained to adequately char-acterize this difference. One way to character-ize the difference in the effect of these slightlyraised pitches, which Raffman’s indirect descrip-tion seems to invite, is to say that each differentlyraised pitch will result in differently brightened

major thirds. But it seems that our indirect descrip-tions, including metaphorical description, will fallshort of rendering these qualitative differences ef-fable. In order to accurately characterize the dif-ferences in terms of brightness, we will need toinvoke descriptive terms that we will not be ableto remember and reliably reapply, in Raffman’ssense. There will be distinguishable brighteningstoo subtle to be captured by the metaphors wecan reliably remember and apply (“bright,” “shim-mering,” “radiant”). However, next, we will seethat these nonstructural nuance objectives do notadd to the air of ineffability; rather, acknowledg-ing the existence of nonstructural objectives helpsus to discover just how indirect description is use-ful in understanding musical nuance.

v. reorientation

The reorientation begins by considering the wayrock musicians deal with the practical difficultyof communicating about musical nuance.23 It isnot unusual for a serious band of rock musiciansto invest a significant amount of time in attempt-ing to communicate about nuances. Why? At theoutset, I mentioned Gracyk’s view that rock mu-sical works just are recordings, and I pointed outthat since nuances are captured in the recordings,they are features of the musical works. Here, Iwould like to highlight the importance of nuancesin rock music in a different way, namely, throughthe perspective of the creative process. For rockcomposers, nuances are of central concern in thecreative process. Consider an example in whichthe nuance objective is nonstructural. It is oftenthe case that a composer envisions not merely acertain rhythm (which is a matter of structure)but a certain groove (which is the feel of a rhythm,a nonstructural objective). She might imagine arhythm that feels as though it pulls against the gui-tars to a specific degree (which is brought aboutby certain notes in each measure being repeatedlydrummed slightly late). If, upon trying to createthis groove with her band, she discovers that thepulling between the drums and guitars cannot beachieved to her satisfaction, the composer may optfor a different rhythm altogether. In such a case, anonstructural nuance objective matters more thanstructure for the emerging musical work.24

Above, regarding brightness, we preliminarilyconcluded that this nonstructural kind of nuance

6 The Journal of Aesthetics and Art Criticism

objective seems to be ineffable in Raffman’s sense.But there is a common way that rock musiciansmove beyond such limitations of indirect descrip-tion for practical purposes. Rock musicians (andother musicians, of course) share a fine-tuned fa-miliarity with a large number of recordings; byreferring to these recordings, they add compar-isons to indirect descriptions. They often beginwith an indirect description and then, in order toadd specificity, refer to an example: “the bright-ness I have in mind is a brightness just like the oneso-and-so achieved on the recording of that song.”A composer may say to a drummer, about a tar-get groove, “Lean the beat forward—not like therecorded version of The Beatles’ ‘I Saw Her Stand-ing There’ but like the Washington, D.C., live per-formance of the song from 1964.” By adding acomparison to indirect description we have addeda degree of specificity to our ability to communi-cate about nonstructural nuance objectives.25 Inthis case, musical nuances are effable enough forthe practical purposes of rock musicians, and Isuspect, through similar devices, for the practicalpurposes of musicians in general. After all, musi-cians do manage to communicate about nuances.Thus, ineffability seems to be relative to the taskat hand, and as far as the perceptually rich, prac-tical task considered above, nuances are effableenough.

Objecting to Raffman’s ineffability claim by in-voking comparison in addition to indirect descrip-tion is not the pivotal idea to emerge from theconsideration of rock musicians’ practical dealingswith nuances. A crucial point has come to light: inorder to acquire the improved specificity awardedby comparison, rock musicians do not focus onthe nuances themselves (the raised or loweredpitches, early or late notes); that is, they do notfocus on nuances characterized in terms of directdescription, such as ‘F-sharp(13),’ and often, theydo not focus on the structural objectives of nu-ances. Instead, in many cases, rock musicians arefocused on nonstructural nuance objectives. Thatwhich is placed in parentheses (in other words,downplayed) in Raffman’s indirect description isoften the emphasis of the rock musician’s practicalconcern. In the rhythm example, the composer isnot focused on exactly how late those eighth notesshould be played, and she is not focused on howthe listener should hear the structure; rather, she is

concerned to convey to her drummer just to whatdegree the rhythm should feel as though it is lean-ing backward or pulling. This is a nonstructuralobjective of a musical nuance.

This shift in emphasis is also manifest in relationto noncomposing rock musicians. All good rockmusicians are reflective about the fact that theyare able to play notes slightly early, late, raised,lowered, et cetera; but their focus, like the focusof composers, is often on the nonstructural objec-tives of these minute variations rather than thevariations themselves. In fact, in these cases, theaspects of experience that confirm to a musicianthat she has accomplished what she set out to ac-complish in performing a minute variation just arethese nonstructural objectives. This requires not afocused perception of the slightly raised F-sharpbut a perception of what it accomplishes in its con-text. For a bit of variety, let’s consider a timbreexample. Consider harmony singing. A backupsinger makes numerous slight variations in pitchand timbre; the timbral variations are sometimescontrolled by the shape of his lips in pronouncinga word. His ultimate criteria for judging the effec-tiveness of the variations are not perceptions ofjust how the lip adjustment affects his timbre, butrather, perceptions of the effect of his variationsin context. The relevant effect is often discussedin terms of a metaphor of a blending of the voices,which is a nonstructural objective of this varia-tion.26

Here, now, is the principal question that leadsto the reorientation: Raffman grants that musicalnuances are perceptual properties, so which per-ceptual properties should be the focus of an exam-ination of musical nuance?—which perceptions?27

That is, what should we take to be our explananda?First, let’s remind ourselves of what we have dis-covered Raffman’s nuances to be. Raffman doesnot take her indirect description to be relevant toher task; she does not take nuance objectives andthe details of musical context to be relevant to de-scribing musical nuances. Moreover, she charac-terizes her explananda directly: “recall the numer-ical N-pitch names ‘A-natural(1),’ ‘A-natural(12),’‘B-flat(17),’ and so forth introduced in precedingchapters; these could serve perfectly well for enun-ciating the [representational] content in question”(LMM, p. 140). In other words, the term ‘B-flat(17)’ is her way of picking out and describing a

Roholt Musical Musical Nuance 7

nuance. What is a B-flat(17)? It is a raised B-flat.Music-theoretic context does not provide tractionfor distinguishing between a B-flat(17) and a B-flat(18). The numbers—(17) and (18)—representthe most important aspect of those pitches, forthe purposes of Raffman’s account of musicalnuance; the numbers represent the nuance as-pect of those pitches. Raffman characterizes thesenuance aspects as detectable pitch differencesthat can be directly described with terms such as‘B-flat(17).’

So again, which perceptual properties should bethe focus of an examination of musical nuance? Ibelieve that we should not follow Raffman in tak-ing the perceptual properties enunciated by termssuch as ‘B-flat(17)’ and ‘eighth note(4)’ to be theexplananda of the examination of musical nuance.It is illuminating to note that the kind of percep-tual experiences that inform Raffman’s character-ization of musical nuance—that is, the data, theperceptual experiences that ground her claim thatwe can distinguish between a B-flat(17) and a B-flat(18)—are, in some cases, not even musical expe-riences. Consider one of her “policy statements”:“Many of the experiments pertinent to our inves-tigation employ synthesized sine tones (that is,tones lacking the overtones inherent in any in-strumental or vocal tone), and stimuli are oftenpresented independently of any tonal or other-wise musical context” (LMM, p. 64). If we startwith real musical experiences, we are naturallyled to different explananda. The explananda of anaccount of musical nuance should be the raisedor lowered pitches and early or late notes as theyare perceived in their musical contexts. Now, whatis it to perceive a B-flat(17) in its musical con-text? Hearing a B-flat(17) in its musical contextis not to perceive that pitch as an isolated prop-erty; hearing a B-flat(17) in its musical contextjust is either to perceive its structural objectiveor its nonstructural objective. Clearly, structuralnuance objectives do not deserve to be called nu-ances, because they are conceptualizable; they canbe characterized in terms of music-theoretic (orother) concepts. Thus, on my reorientation, theexplananda of an account of musical nuance oughtto be the nonstructural objectives of raised or low-ered pitches, early or late notes, et cetera. A musicalnuance just is the nonstructural objective of a mu-sician’s slight adjustment: a brightened interval or

a leaning groove. A musical nuance is not the var-ied note itself, characterized in terms of direct de-scription. We are, after all, using the term ‘musicalnuance’; what is truly musical about a B-flat(17) iswhat it does in a perceived musical context.28 (Ifwe were following music theorists and psychol-ogists in using the term ‘expressive variation,’ Iwould make the same point: a B-flat(17) only be-comes expressive in its musical context.)

There are practical reasons to prefer this changein explananda. This approach to examining mu-sical nuance is more fruitful than Raffman’s be-cause it will yield observations that are more mu-sically relevant. After all, Raffman’s explanandaled to ineffability; this does not help us to under-stand anything about the music except that thereare some contributors to it that we cannot retainin our perceptual memories. On Raffman’s view,in attempting to say something about the bright-ened interval and the pitch variations that bring itabout, all we can say is that the F-sharp is slightlyraised, and that this difference in pitch is a differ-ence that we are able to perceive but not remem-ber and reidentify; it is ineffable. We are not pro-vided tools or leverage for describing the purposeor end of the variation in the music. The central-ity of the new explananda to musical practice isalso obvious in that nonstructural objectives arethe explicit goals of musicians who perform slightpitch variations and other nuances, and goals of (atleast) rock composers who envision them. Theseexplananda are not only central features of themusic itself but can function as answers to ques-tions about why a composer or musician createdthis particular feature at this particular juncturein a given performance or recording. Focusing onthese new explananda predisposes us to seek toelucidate elements of performances and record-ings that are relevant to musical descriptions thatcan ground evaluations.

It should be clear that I am not disputing the ob-servation that we can, in fact, discriminate morefinely than we can categorize (which has beenraised by psychologists, Raffman, and philoso-phers of mind writing on nonconceptual content—see note 11). In fact, it is useful to notice howthis observation fits into my reorientation. Thefact that we can discriminate so finely providesonly a causal explanation (at the level of ourperceptual systems) as to why it is that minute

8 The Journal of Aesthetics and Art Criticism

variations in pitch (et cetera) have subtle effectsin musical contexts. In other words, we are able tohear a difference between an F-sharp(12) and anF-sharp(13) (as psychology experiments demon-strate), and that is why such slight variations re-sult in subtly different perceived brightenings ina musical context; that is why the nonstructuralobjectives of minute variations are nuanced. Butas I have argued, the nonstructural objectives arethe proper explananda of an account of musicalnuance, not the minute variations themselves.

vi. conclusion

Diana Raffman’s account of musical nuance isdriven by direct description; her claim that nu-ances are ineffable follows as a result. By consid-ering the efficacy of indirect description, we wereled to the discovery that her account centers on thewrong explananda. For Raffman, musical nuancesare the perceptions of the pitches (et cetera) them-selves: F-sharp(13), eighth note(12). Consideringindirect description resulted in broadening ourperceptual perspective, which led to noticing themusical significance of the effects of these minutevariations. My reorientation rests on my claimthat the explananda of the examination of musi-cal nuance ought to be the nonstructural objectivesof minute variations, such as a brightened majorthird or a groove—these are musical nuances. Indi-rect description, including metaphor and compari-son, renders these musical nuances effable enoughfor practical purposes. In addition to the render-ing effable of nuances, on what other grounds doI propose this change in explananda? Musiciansperform and (rock) composers stipulate minutevariations for the purpose of the effects they havein a perceived musical context; minute variationsacquire their musical significance in their effects.Terminologically speaking, an A-natural(12)—callit a ‘musical nuance’ or an ‘expressive variation’—only becomes musical or expressive in a perceivedmusical context. Focusing on the ends of minutevariations is the key to understanding what mu-sicians and composers are up to when they per-form or stipulate variations. This kind of under-standing is required for grounding evaluations ofperformances and music that include these vari-ations; this reorientation predisposes us to con-struct musically relevant descriptions, because theeffects of minute variations are already embedded

in the music. (In this regard, Raffman’s accountyields only a cognitivist explanation of our percep-tual limitations with respect to minute variations,and a claim that minute variations are ineffable.)Raffman and I both take our explananda to be mu-sical nuances, but on her view, a musical nuanceis (for example) a slightly raised A-natural—themeans—whereas, on my view, a musical nuance isthe slightly raised A-natural’s nonstructural ob-jective in a musical context—the end.29

TIGER C. ROHOLTDepartment of Philosophy and ReligionMontclair State UniversityMontclair, New Jersey 07043

internet: tiger.roholt@montclair.edu

1. “Crazy,” Gnarls Barkley, St. Elsewhere (Atlantic,2006).

2. “Fly Me to the Moon (In Other Words),” Sinatra atthe Sands (Warner Brothers, 1966).

3. In this essay, unless otherwise specified, when I men-tion pitch or timing nuances I intend to refer to musical nu-ances in general. In her Language, Music, and Mind (MITPress, 1993), Diana Raffman focuses on pitch but acknowl-edges the import of other kinds of nuances: “as I notedearly on, I conceive the N-pitches [nuance-pitches] and N-intervals as just two among a constellation of nuance fea-tures. . . . There is every reason to suppose that musical per-formances sustain fine details of duration, loudness, speed,articulation, and timbre, among others” (p. 90). Althoughmy reorientation is applicable to nuances in general, I donot assume that nuances are interchangeable for the pur-poses of all examinations. On this last point, see Justin Lon-don’s criticism of Raffman for not examining research onrhythm perception: “The Fine Art of Repetition: Essays inthe Philosophy of Music; Language, Music, and Mind,” Mu-sic Theory Spectrum 16 (1994): 269–275.

4. Here are two rock music nuance timing examples: acontributing factor to the perceived power of Jon Bonham’sdrumming is his tending to play “behind the beat”; that is, heplays certain notes slightly late. A contributing factor to theperceived buoyancy of Ringo Starr’s drumming is that hedoes not play behind the beat; in fact, Ringo tends to play abit ahead of the beat. Regarding hip-hop, one of the genre’ssignificant innovations is the exploration of just how farbehind the beat notes can be placed without being perceivedas rhythmically chaotic. Hear Brandy’s “What About Us?”Full Moon (Urban Atlantic, 2004).

5. Diana Raffman adopts this view in Language, Mu-sic, and Mind (see chap. 4). If nuances have anything todo with classical musical works, it is most likely through acomposer’s assumptions about performance practices. Fora relevant discussion, see Stephen Davies, Musical Worksand Performances (Oxford University Press, 2001), pp. 110,116.

6. Theodore Gracyk, Rhythm and Noise: An Aestheticsof Rock (London: I. B. Tauris & Co., 1996), chap. 1.

Roholt Musical Musical Nuance 9

7. Stephen Davies, “Rock Versus Classical Music,” TheJournal of Aesthetics and Art Criticism 57 (1999): 193–204,quote on p. 200.

8. References to this book will be given parentheticallyin the body of the essay as LMM followed by the relevantpage number.

9. The generative theory of tonal music posits four sets ofanalytic rules that a listener “applies” to pitch-time events;these rules have been unconsciously abstracted from expo-sure to the music of the listener’s culture; the rules constitutethe musical grammar (“M-grammar”). See Fred Lerdahl andRay Jackendoff, A Generative Theory of Tonal Music (MITPress, 1983).

10. Schemas are formed as a result of repeated exposureto one’s culture’s music; a hierarchical structuring of pitchesis the result. As Bharucha writes, “The tonal hierarchies usedin Western music are the 12 major and minor keys. . . . Thesehierarchies are internalized by members of a culture, andfacilitate the processing of music of that culture” (LMM,p. 69).

11. Prior to Raffman’s work, some psychologists,philosophers, and others acknowledged that perceptual dis-crimination outstrips conceptualization; see, for example,Gareth Evans, The Varieties of Reference (Oxford Univer-sity Press, 1982). The import of Raffman’s book is to applysuch an observation to the consideration of musical ineffa-bility. Note that the thrust of Raffman’s ineffability claim hasresonance in the philosophy of mind literature on noncon-ceptual content (which has roots in Evans); see ChristopherPeacocke, “Does Perception Have a Nonconceptual Con-tent?” Journal of Philosophy 98 (2001): 239–264.

12. One might wonder whether experienced musicianscan improve their abilities to recognize nuance pitches. Someof Raffman’s critics have failed to notice that she addressesthis. Raffman cites findings to the effect that even Indian mu-sicians, whose scales are microtonal, are unable to reliablycategorize more finely than semitones. The main point beingthat, although we might acquire more finely tuned schemas,“it is overwhelmingly unlikely that we could acquire intervalschemas as fine-grained as the pitch and interval discrimi-nations we can make” (LMM, p. 84). Also consider thefollowing: “The limits of our schemas are the limits of ourlanguage, and qua perceivers we are so designed that thegrain of conscious experience will inevitably be finer thanthat of our schemas, no matter how long, or how diligently,we practice” (LMM, p. 136).

13. W. E. Kennick, “Art and the Ineffable,” Journal ofPhilosophy 58 (1961): 309–320.

14. Kennick, “Art and the Ineffable,” pp. 317–318.15. Ludwig Wittgenstein, The Blue and Brown Books,

2nd ed. (New York: Harper and Row, 1965).16. Wittgenstein, The Blue and Brown Books, p. 181.17. Wittgenstein, The Blue and Brown Books, p. 181.18. Georges Rey, review of Language, Music, and Mind,

in The Philosophical Review 106 (1997): 641–645.19. Rey, review of Language, Music, and Mind, p. 643.20. The psychologist Eric F. Clarke writes, “It is an as-

sumption of most performance research that expression isprimarily used to convey musical structure to listeners.” EricF. Clarke, “The Perception of Expressive Timing in Music,”Psychological Research 51 (1989): 2–9, quote on p. 3. (By‘expression,’ Clarke means expressive properties such asnuances.)

21. Raffman writes, “The structural description is socalled because it represents the signal as instantiating thebasic structural elements of the music—namely its pitchesand rhythms—as well as the more abstract structural fea-tures inferred therefrom. In calling these the structural fea-tures, I mean inter alia that they constitute a system of el-ements whose tokens are (1) discrete, (2) type-identifiableby some finite mechanical procedure, and (3) combinablein certain role-governed ways. . . . The structural descriptiondoes not, at least not explicitly, capture what I shall callthe nonstructural features of the music. Dynamics (loudnesslevels), tempi (perceived speeds), and timbres (tone ‘col-ors’) come immediately to mind. . . . For the time being Ishall characterize the nonstructural features, unhelpfully, asthose audible features of the music that fail to satisfy the trioof conditions just cited” (LMM, p. 25). The more thoroughcharacterization of nonstructural features Raffman signalshere turns out to be her account of nuances. Unfortunately,she understands these to be discrete properties that are char-acterizable in terms of direct description.

22. Some minute variations in pitch and timing have tim-bral effects (objectives), and some timbral adjustments con-tribute to more global timbral effects. These sound qual-ities are central in rock, as Gracyk emphasizes: “Rock isa music of very specific sound qualities and their texturalcombination. Specific sounds are as central to the music asare specific colors in painting” (Gracyk, Rhythm and Noise,p. 61).

23. This picture emerges from my own experience; thisis (at least) a way that some rock musicians deal with musi-cal nuances. I certainly do not mean to imply that classicalmusicians are less conscious of nuances.

24. Something similar may occur in classical composi-tion, but the nuance considerations will be less specific, Ibelieve, limited to types of nuances (cashed out in terms ofperformance practices and, perhaps, stipulated in a score’ssupplementary instructions). See note 5.

25. Notice that we may take the comparisons used bygood critics to have a similar function.

26. Paradigm examples of this kind of timbral adjust-ment, and the resulting blend, are found in the singing ofThe Jordinaires, and in Marty Robbins’s backup singers, es-pecially the tenor.

27. Raffman’s nuances are perceptual properties, notacoustic events; in this regard, see especially her criticismof Nelson Goodman (LMM, pp. 115–116, 119).

28. Even though many psychologists work with a def-inition of nuances (“expressive variations”) identical toRaffman’s, some have noted the distinction between nu-ances thus defined and something like what I have been call-ing nonstructural objectives. Consider these two examples.John Sloboda writes, “We can hear rhythmic imprecisionand rubato with appropriate training, but fine differences intiming are more often experienced not as such, but as differ-ences in the quality (the ‘life’ or ‘swing’) of a performance.”John A. Sloboda, The Musical Mind: The Cognitive Psychol-ogy of Music (Oxford University Press, 1985), p. 30. Eric F.Clarke writes, “Auditory events, or more specifically musicalevents, are inherently multi-dimensional, and . . . althoughthey may be theoretically, and even empirically, decompos-able into unitary components, this may destroy or concealemergent properties of the whole event.” Eric F. Clarke,“The Perception of Expressive Timing in Music,” p. 3.

10 The Journal of Aesthetics and Art Criticism

29. In seeking to understand the way subtle perceptualproperties operate in other kinds of artworks, it seems to methat we may similarly benefit from orienting toward ends.

First and foremost, I owe thanks to Lydia Goehr andMichalle Gal. Related papers were presented at The Amer-ican Society for Aesthetics, Eastern Division Meeting (2008)and at the second Aesthetics Anarchy (2008); I benefit-ted from the comments of those present. Thanking ev-eryone who has helped me to think about musical nu-ances would require naming many former bandmates—first

on that list would be my father, Les Roholt. For com-ments on related papers, and support, I would like tothank Hanne Appelqvist, Christopher Bartel, Taylor Car-man, David Clowney, Joseph Dubiel, Ted Gracyk, GarryHagberg, Sean D. Kelly, Hagi Kenaan, Margaret Moore,Jonathan Neufeld, Christopher Peacocke, Eric Rubenstein,Sirine Shebaya, and Brian Soucek. Of course, I am responsi-ble for any errors that remain. Finally, I would like to thankMichael Root for helping me to begin thinking about theseissues as an undergraduate in his metaphysics course.