In the American Society for Music Theory, Music in the Moment was the most-discussed book of 1998. The book was the topic of a three-hour session at the Society's national conference, and its author, the philosopher Jerrold Levinson, was mentioned 90 times on the Society's Internet discussion group—more often than any other person except Schenker, Beethoven, and Schoenberg.
Levinson's sudden notoriety arises from his provocative claim that large-scale form is almost irrelevant to his own experience of music—a type of experience that he suspects is shared by 'a vast, silent legion of architectonically unconcerned listeners' (p. xi). As a rough way of introducing this claim, Levinson reviews a position taken by Edmund Gurney (1880) in The Power of Sound—a position that Levinson calls concatenationism. In its starkest form, concatenationism asserts that musical understanding, enjoyment, and value arise entirely from the 'impressiveness' of a composition's individual parts and from the 'cogency of succession' between adjacent parts. By contrast, large-scale relationships between 'parts widely separated in time' have no influence at all on the listener's experience.
Some of the furor surrounding Levinson's book results from a mistaken belief that concatenationism, in the strict form outlined above, represents Levinson's final, considered opinion on musical experience. To the contrary, most of Levinson's book is devoted to criticising and modifying the concatenationist viewpoint. By his penultimate chapter, Levinson has arrived at a far more nuanced position that he labels 'qualified concatenationism'—which gives an important, though subsidiary, role to large-scale connections.
Levinson's qualified position rests in large part on a nice distinction he draws between three ways that listeners might experience large-scale form. The first possibility is that people hear form in music much as they see form in architecture. That is, listeners somehow take in the broad sweep of a piece in a unified perceptual act. This idea is reminiscent of what Schenker calls Fernhören, or long-distance hearing.
A second possibility is that listeners do not so much perceive form as think about it—labeling themes, sections, and so on, and taking note of the relationships between them. As an example of this view, Levinson cites Hindemith's (1952) idea of 'co-construction', in which auditors build up a kind of intellectual diagram as they listen. For Levinson, the distinctive quality of this idea is that listeners' recognition of large-scale connections is assumed to be precise and conscious.
Unfortunately, Levinson muddies the foregoing distinction by frequently lumping the two types of listening together under the rubric of 'architectonicism'. In the interests of clarity, I will call the first type of listening synoptic (a term sometimes used by Levinson). The second type of listening I will call intellectualised.
Levinson's own view of large-scale form represents an important part of the qualified concatenationist position. According to Levinson, the listener's consciousness generally remains in the present—in the bit of music now playing and in its immediate connection to the bits before and after. However, the listener's experience of the musical present may be shaded or colored by unconscious memories of the musical past, or by unconscious expectations for the musical future. These memories and expectations can be quite vague and subliminal: for example, listeners need not recognise that they are hearing the first theme of a sonata reprise; instead, they may simply feel that the theme is familiar and open-ended.
For reasons I will discuss below, Levinson claims that synoptic and intellectualised listening play at best a peripheral role in the listening experience. The central role is played by moment-to-moment listening, faintly tinted by memory and expectation.
The persuasiveness of Levinson's stance rests largely on its phenomenological appeal. For the most part, you either feel that Levinson is describing your experience or that he is not. In this respect, Levinson's mode of presentation resembles a great deal of writing in music theory—and shares with that writing the danger of over-generalising from personal experience. For the committed skeptic, Levinson does offer some observations and arguments that seem, at face value, to favor qualified concatenationism. Yet, on closer inspection, a great deal of this evidence simply evaporates.
The most fundamental argument that can be offered for concatenationism is that it acknowledges the temporal nature of musical experience. As Gurney observed, a work of music, unlike a work of architecture, is simply not available all at once; at any moment, that is, only a very small part of a musical work is aurally present. This observation seems to argue against the possibility synoptic listening, although it leaves the possibility of intellectualised listening untouched.
As Levinson points out, however, Gurney's line of reasoning appears vulnerable to a reductio ad absurdum. The trouble with Gurney's argument is that it applies to anything experienced in temporal sequence. Following Gurney's logic, that is, one could adopt a concatenationist view of 'philosophical arguments, mathematical proofs, whodunits, treatises, and so on' (p. 169). Yet the understanding of any of these objects requires, at least in Levinson's view, some grasp of their overall form.
Levinson answers this objection by pointing out that there are, after all, important differences between rational structure and musical structure. The parts of a musical structure are valuable for their own sake, while the value of a rational structure depends on the attainment of a larger objective. The tacit assumption here is that global objectives are irrelevant to musical value. Only a hardened concatenationist, of course, would accept this assumption; a Schenkerian would assert that large-scale goals are crucial. Thus Levinson's attempted defense of Gurney comes dangerously close to begging the question.
Levinson also considers another type of reductio ad absurdum. This one begins by observing that, if no global structure is evident in a musical moment, neither is any local connection or flow between successive moments. Pushed to its logical conclusion, that is, Gurney's argument reduces music to a succession of disjointed sensations that could be reordered to no loss of effect. Such a conclusion would be as damaging to a concatenationist as it is to a partisan of large-scale form.
To escape from this dilemma, Levinson proposes a mental process that he calls 'quasi-hearing'. This refers to a listener's ability to keep up to a minute or so of continuous music in 'vivid apprehension'. Music that is quasi-heard is larger in scope than what can actually be heard at a given instant. Unlike intellectualised hearing, however, quasi-hearing does not depend on a conscious act of recollection.
Given the possibility of quasi-hearing, Gurney's position regains its earlier force. In the terms proposed by Levinson, what is important to musical experience corresponds to what can be quasi-heard. But this compatibility should not surprise us, since the notion of quasi-hearing was introduced expressly to shore up Gurney's position. A Schenkerian arguing against Levinson could just as conveniently propose a type of hearing that applies to more extended structures; indeed, this is exactly the type of hearing designated by the concept of Fernhören.
Some of Levinson's simpler observations are more robust. For example, he points out that listeners frequently sing along with music, at least under their breath. He also points out that when music is interrupted, listeners have more or less definite ideas about how it will continue. (These expectations were first measured experimentally by James Carlsen in 1981.) Both 'inward seconding' and 'continuational ability' do seem to indicate that listeners are closely following the musical flow. This does not, of course, rule out the possibility that listeners are also involved with higher levels of structure.
To summarise, the essence of qualified concatenationism is this: large-scale form does affect listeners, but it does so by unconsciously enhancing the experience of musical flow. This, however, is not the only qualification in qualified concatenationism. Through the course of his book, Levinson makes a number of further allowances for large-scale form.
First, Levinson concedes that his ideas are based principally on introspection. Although he suspects that many listeners experience music more or less as he does, he acknowledges the possibility of listeners for whom large-scale relationships play a more direct and prominent role.
Second, Levinson allows that, even for himself, conscious awareness of a large-scale relationship can sometimes make the musical surface seem more impressive or cogent. He hastens to assure us, though, that such enhancements are both rare and feeble. For the most part, Levinson feels that awareness of global relationships has little power to change musical experience. 'Most often, I suggest, intellectual apprehension of large-scale form has no noticeable influence on the rate at which music gels for a listener on the plane of basic musical understanding' (p. 128). For Levinson, the proper role for large-scale analysis is not to improve experience, but to explain it. Levinson's prescription for enhancing musical experience is simply to listen to a piece repeatedly.
Levinson's third concession is that awareness of formal artifice—such as palindrome or cyclic structure—does afford a certain intellectual pleasure. He maintains, however, that this pleasure is distinct from the pleasure of involvement in the musical surface. Putting this distinction another way, Levinson maintains that the pleasures partake of two distinct objects:
[L]arge-scale formal awareness in music listening...could reasonably be maintained to have essentially a different object—the formal structure of a piece, rather than the piece, a hearable stretch of music, itself.
(p. 69)
What is questionable here is the covert assumption that 'a hearable stretch of music' can be treated as a synonym for 'the piece...itself'. If we discount this expedient definition, Levinson's argument reduces to the tautology that 'large-scale formal awareness' partakes of 'formal structure', while moment-to-moment listening partakes of 'a hearable stretch of music'. This merely reiterates the distinction between intellectualised and qualified-concatenationist listening; by itself, this distinction provides no evidence that one type of listening is more important than the other.
Levinson's final, somewhat limited concession relates to 'higher-order aesthetic properties'. He maintains that some such properties, such as unity, can be communicated simply by cogent succession, while other properties, such as grandeur, can be communicated by a cumulative effect in which each successive passage is more and more enhanced by unconscious memories of what came before. Even so, Levinson argues, there remain certain effects—such as triumph, regret, nostalgia, consolation, and ironic humor—that simply cannot be achieved unless the listener draws a conscious connection between two remote sections.
When all of these concessions have been made, does anything remain to distinguish the concatenationist position? According to Levinson, the answer is an emphatic yes. First, although there are valuable experiences to be had through the contemplation of large-scale form, Levinson finds that those experiences are relatively few in number and minor in intensity. It remains the case, according to Levinson, that the bulk of listening enjoyment is had in the moment.
In support of this claim, Levinson points to the appeal of musical excerpts:
Concatenationism can explain why the phonographic offerings of thirty years ago titled 'Fifty Great Moments in Music' and the like, involving excerpts averaging three minutes in length, while not representing perhaps the ideal approach to the classical repertoire, yet provided something of significant, self-standing values, whereas a visual analog of such offerings, say 'Ninety Greatest Square Inches of Abstract Painting,' complete on one sturdy cardboard-backed poster, would be little more than a joke.
(p. 165)
The implication is that no one would have bought such recordings unless the excerpts had retained much of the value. In a similar vein, David Huron (personal communication) points out that he can start his car in the morning, turn on a classical radio station, and listen with enjoyment to some masterpiece or other—until he pulls into his parking space a few minutes later and turns the radio off. This experience is quite telling: for unlike the musical excerpts in an anthology, the snippet heard while driving is selected at random, with no respect for the formal divisions within the work.
Levinson argues not only that large-scale form is of negligible value, but that its contemplation affords only a secondary or 'parasitic' kind of enjoyment. That is, the contemplation of large-scale form is nearly worthless unless the musical surface is enjoyable from moment to moment. As evidence, Levinson proposes a trenchant thought experiment.
Consider a piece of music with unarresting themes, uninteresting harmonic development, incoherent instrumentation, transitions of minimal cogency, and in general progression of a sort that either totally stymies quasi-hearing or else bores it into submission—but suppose that this piece displayed some intellectually apprehendable patterning of the [large-scale] sort we have discussed. Who would care?
(p. 156)
A more complete version of this experiment is carried out several pages later.
Consider a movement M1, displaying all sorts of 'rational' relationships—symmetries, balances, repetitions over large spans—but little cogency of succession of part on part. Consider a movement M2, with cogency of succession at every point but no discoverable, contemplatable large-span relationships of a 'rational' sort. Which is better music? Without question, M2. Cogency of succession is clearly more important to musical worth than demonstrable large-scale balances, symmetries, or repetitions. But consider now movement M3, which has the virtues of both M1 and M2. It seems unclear, at the least, how much better M3 is musically than M2.
(pp. 163-164)
Like any thought experiment, this one will only persuade readers who can replicate the results in their own minds. For such readers, however, the results do seem to show that only cogency of succession is both necessary and sufficient for musical value. Large-scale relationships are only interesting when cogency of succession is in place—and even then they offer a meager return on investment. In short, the contemplation of large-scale relationships
is essentially icing—though sometimes of a most impressive sort—on the musical cake, and not the cake itself. The basic point is that if the cake is not in place, there's nowhere for the icing to go.
(p. 157)
Curiously, Levinson makes little attempt to review the experimental literatures on the perception of time in general and musical form in particular.[1] This could have been a profitable undertaking, because Levinson's idea of quasi-hearing seems quite similar to what psychologists call working memory (Baddleley 1986). The one experiment that Levinson cites is Cook's (1987) finding that trained listeners seem aesthetically indifferent to large-scale tonal closure. Taken at face value, Cook's result suggests that even Levinson's few concessions to global form may be unnecessary—but Levinson chooses not to explore these implications.
A more surprising omission is Levinson's failure to explore the critical writings associated with intellectualised and synoptic form. Aside from passing references to Schenker and Hindemith, Levinson makes no attempt to document the use of synoptic and intellectualised assumptions in the literature of music theory. Instead, he takes it for granted that most theorists are blind adherents of 'architectonicism'. His implicit supposition is that theorists hew to architectonicism simply because they don't know any better—that is, because they have not been properly introduced to the qualified-concatenationist position.
Had he taken architectonic ideas more seriously, Levinson might have placed them in their philosophical context. Given his extensive philosophical background, Levinson could have tried to relate synoptic and intellectualised listening to idealist or formalist aesthetics. In doing so, he would have had to contend with authors whose ideas about listening include moral claims that, if accepted, make Levinson's advocacy of concatenationism seem irresponsible. Hanslick (1854), for example, would have agreed with Levinson that most listeners can and do derive the bulk of their enjoyment from unreflective immersion in the musical moment. For Hanslick, however, such enjoyment was 'pathological' rather than 'aesthetic': unreflective listening, Hanslick felt, was animalistic and degrading, the musical equivalent of drunkenness. To Hanslick, Levinson would seem like a man passing a hip flask at a meeting of Alcoholics Anonymous.
In order to come to terms with theorists' claims about large-scale form, it seems important to establish whether those claims are prescriptive or descriptive (an issue on which many theorists are unclear[2]). To the degree that claims about form are prescriptive, moral opinions like Hanslick's may well have influenced the prominence that theorists assign to different types of listening. Levinson does not attempt to trace the influence of such moral views, nor does he attempt to answer them explicitly.
To be fair to Levinson, a moral position could be inferred from his claim that analysis has little power to change the listening experience. Unfortunately, Levinson makes this claim only in passing, without offering any real evidence for it. If the claim is accepted, however, it renders moral judgements moot: for if listening is unchangeable, there can be no moral compulsion to change it.
I cannot review this book properly without commenting on the uproar it has caused among music theorists. While some of the commotion can be traced to uninformed opinions briefly circulated on the Internet, much of it can be traced to the book itself. Levinson presents the strict form of concatenationism quite early, insists that it is basically correct, and then spends the next hundred pages adding one qualification after another. This rhetorical strategy keeps the reader off guard, wondering throughout where the author is headed and what makes him so sure of himself. Levinson should not be surprised if some of his readers have reacted defensively.
Levinson clearly feels that music theorists have exaggerated the importance of large-scale form, and he does not hesitate to dress us down for it. 'It is dispiriting', he writes,
to think of the many persons fully capable of appreciating the glories of classical music, to speak of no other kind—such as jazz—who have turned away without even venturing to cross the threshold, disheartened by the mistaken belief, which music theorists and commentators often do little to dispel, when they are not actively promoting it, that elaborate apprehensions of the form and technique of music are necessary to understanding it, and thus to reaping its proper rewards.
(p. 174)
The rhetoric here is clearly overwrought. Among the droves of listeners who stay away from classical music, how many of them are actually frightened off by music theorists? The usual complaint among theorists is quite the opposite: that no one much cares what we have to say.
It is difficult to know why Levinson chose to present his careful observations in such a strident manner. At times I found myself wondering if it was all a kind of publicity stunt. Did Levinson strike a combative pose in order to attract the notice of music theorists? If that was his aim, he succeeded.
Despite its problems, Music in the Moment remains a brave and insightful book. You do not have to agree with Levinson's position to recognise that he has pointed to some vital and underserved questions in music theory.
An issue that Levinson points to directly is the nearly-forgotten importance of musical flow. Until about forty years ago, music theorists spent much of their energy describing chord progression, voice-leading, and other phenomena related to 'cogency of succession'. This tradition entered a new era in the 1950s, when a number of scholars began to characterise musical succession more precisely using the tools of information theory.[3] With the Schenkerian revolution of the 1960s, however, these promising developments were simply tossed out. If Levinson is even partially right, there may have been quite a baby in that bathwater.
A second lesson emerging from Levinson's work is that we have often been remiss in specifying how large-scale relationships are experienced by listeners. Levinson's distinction of synoptic, intellectualised, and qualified-concatenationist listening provides useful categories. In our writings, we should indicate more precisely which category we mean to invoke.
A final issue, though not raised explicitly by Levinson, certainly nagged at me while I read his book. The issue is that music scholarship lacks an explicit theory of memory. Nearly all of the concepts discussed by Levinson seem to cry out for cognitive elaboration. As I remarked earlier, the notion of quasi-hearing might be fleshed out in terms of working memory. Similarly, the proposal that memories colors the musical present raises questions regarding the way that such memories might be formed or evoked. What makes a tune or chord memorable, and how much correspondence is required for the memory of one passage to color another? The precise characterisation of musical memory is one of the truly great problems in our field.[4]
Baddleley, Alan, 1986: Working Memory. (Oxford, U.K.: Clarendon Press).
Carlsen, James C., 1981: 'Some Factors Which Influence Melodic Expectancy', Psychomusicology, 1, pp. 12-29.
Clarke, Eric F., 1999: 'Rhythm and Timing in Music', in D. Deutsch (ed.), The Psychology of Music (San Diego, CA: Academic Press).
Cook, Nicholas, 1987: 'The Perception of Large-Scale Tonal Closure', Music Perception, 5, pp. 197-205.
Cook, Nicholas, 1990: Music, Imagination, and Culture. (Oxford, U.K.: Clarendon Press).
Coons, Edgar, and David Kraehenbuehl, 1958: 'Information as a Measure of Structure in Music', Journal of Music Theory, 2, pp. 127-161.
Gurney, Edmund, 1880 [1966]: The Power of Sound. (New York: Basic Books).
Hanslick, Eduard, 1854 [1986]: On the Musically Beautiful: A Contribution Towards the Revision of the Aesthics of Music, trans. Geoffrey Payzant. (Indianapolis, IN: Hackett).
Hindemith, Paul, 1952: A Composer's World. (Cambridge, MA: Harvard University Press).
Meyer, Leonard B., 1957: 'Meaning in Music and Information Theory', Journal of Aesthetics and Art Criticism, 15/4, pp. 412-424.
Temperley, David, forthcoming: 'The Question of Purpose in Music Theory: Description, Suggestion, and Explanation', Current Musicology.
Youngblood, Joseph E., 1958: 'Style as Information', Journal of Music Theory, 2, pp. 24-35.
1. The experimental literature on musical form is reviewed in Cook (1990) and Clarke (1999).
2. For a discussion of the prescriptive-descriptive confusion in modern theory, see Temperley (forthcoming).
3. Among the pioneers in this area were Youngblood (1958), Coons and Kraehenbuehl (1958), and Meyer (1957). Meyer has continued to write speculatively in this vein for the last forty years.
4. I am grateful to David Temperley, David Huron, and Beth Harrison for thoughtful comments on an earlier draft of this review.