Tip! Click on musical notes to start player. 

JMM 1, Fall 2003, section 2

Jerrold Levinson


It has sometimes been remarked that making music - that is, composing, performing or improvising it - involves thought or is a form of thought. If so, what is the nature of the thinking that goes on in making music? And what of listening to music? Is the experience of the comprehending listener also a kind of thinking? How does musical thinking differ from the paradigm of thinking, that is, the formulation and manipulation of thoughts in words? Can musical sequence itself, rather than the activity of producing or auditing it, be regarded as a kind of thinking? In short, is music thought?

In the course of trying to shed light on these issues I will take as a springboard various remarks of Wittgenstein on music that are to be found here and there in his writings. I will also yield to the temptation to emulate, in a small degree, Wittgenstein's elliptical, oracular manner, a manner particularly apt to the exploratory stages of a philosophical investigation, which is certainly the case here. Whether what results should be considered an homage, a parody, or some mixture of the two, I leave to my audience to decide.


It seems clear from a number of Wittgenstein's remarks, especially ones directed to particular composers, that he was indeed inclined to regard music as thinking.[1] In one place we find the following invocation: "The strength of the thoughts in Brahms's music." (CV 23). In another place we are told that one "...can point to particular places in a tune by Schubert and say: look, that is the point of the tune, this is where the thought comes to a head." (CV 47)

What is most striking about these observations is how natural it seems for Wittgenstein to think of music as a kind of thinking, how little in need of defense he appears to take that to be. What if one invoked, by contrast, "The strength of the thoughts in the cuisine of les freres Troisgros", or "The strength of the thoughts in Michael Jordan's basketball playing"? Would this seem as natural? Could we easily speak of a moment in Jordan's progress to the basket, or of a dish in a ten-course meal at Troisgros, where "the thought comes to a head"? I suggest not.


In the Investigations and elsewhere, Wittgenstein remarks that one might describe the effect of a passage of music by saying 'Here it is as if a conclusion were being drawn'. (PI 182)

There are a couple of things to note about this. First, Wittgenstein does not say that, in such a passage, a conclusion is being drawn; rather, it is as if a conclusion were being drawn. So far, then, we are in the realm of analogy or metaphor, or perhaps of the dawning of an aspect. Second, the character of some passages of music to which Wittgenstein is calling attention is specifically that of seeming to draw or reach a conclusion, as after a period of reflection; it is not the idea of merely concluding, in the sense of stopping or terminating. Compare the endings of Beethoven's Piano Sonata op. 110 or Dvorak's 7th Symphony, which seem to sum up and crystallize what has gone before, with the endings of, say, minuet movements from symphonies of the Classical period, even great ones such as Mozart's 40th or 41st. The former have this special rhetorical character of concluding, whereas the latter have only the mundane character of coming to a close - however satisfyingly.


Wittgenstein treats the phenomenon further in another place:

If I say, for instance: here it's as though a conclusion were being drawn, here as though someone were expressing agreement, or as though this were a reply to what came before - my understanding of it presupposes my familiarity with conclusions, expressions of agreement, replies. (CV 52)

What Wittgenstein is underscoring here about the appreciation of music is this. Music is not understood in a vacuum, as a pure structure of sounds fallen from the stars, one which we receive via some pure faculty of musical perception. Music is rather inextricably embedded in our form of life, a form of life that is, as it happens, essentially linguistic. Thus music is necessarily apprehended, at least in part, in terms of the language and linguistic practices that define us and our world.

But by the same token, should we not expect that our understanding of linguistic phenomena will sometimes be inflected by our musical understanding, especially in light of the fact that our musical capacities are awakened at least as early as our linguistic ones? For example, we may describe certain speech as "sing-songy", a conversation as not having the right "rhythm", and the papers at a conference as not "harmonizing". Furthermore, in tonal languages, such as Japanese or Indonesian, the distinction between speaking and singing is to some extent effaced. Though language may be essential to the human form of life - whereas music, though universal, arguably is not, since we can presumably imagine human life without music, but not without language - once both are present their interpenetration is assured, and we cannot help interpreting the one in terms that are rooted in the other.


It is true that the question 'what are you thinking?' most often elicits a verbal answer, such as 'It's going to rain', or 'I need to buy milk soon', or 'She is very attractive'. But why not, on some occasion, a musical phrase, or even a particular phrasing of a musical phrase? If someone asks me what I was thinking, can I not sometimes truthfully say 'the opening of Mendelssohn's Violin Concerto'? Could I not, in response, even whistle that opening, and in a particular way? Note that the former response would not be the same as saying 'I was thinking of the opening of Mendelssohn's Violin Concerto'. For of course anything might be an object of thought. But that doesn't make it into an example of thought. No, the Mendelssohn opening is what I was thinking, not what I was thinking of...


We say of some music that 'thought went into it', or 'there was thought behind it', and mean to contrast that with cases we might rather describe as 'thoughtless note-spinning'. Is the distinction between 'thoughtful' and 'thoughtless' music - or 'thought-filled' and 'thought-free' music - coincident with that between good and bad music? If not, it is probably not too far removed from it.

When we estimate the quality of music we often refer to the mind that is revealed in it, the mind one comes in contact with in listening to it, the mind that is reflected in it, and so on. Granted that there is more to mind than thinking, can there be less to mind than that? If not, then can we easily deny the label of 'thinking' to music of any worth, given the mind that stands before us in sound when such music is played?


Wittgenstein remarks in several places that it is common to experience a musical phrase as a question. It is also not uncommon to experience another phrase as an answer. And experiencing music in such ways seems part of what it is to understand music. Note that this is a matter of phrases striking us as questions and answers on more than the purely musical plane; that is, we are here speaking of more than the sense in which one phrase can serve as musically the answer to another, in terms of completing its melodic arch or balancing its harmonic movement.

Now, if two phrases of music strike us as having a more-than-musical question-and-answer relation, must there be a content to the question the first phrase seems to embody? In other words, must it be possible to say what exactly the first phrase is asking? If not, then what does the claim that the phrases have a more-than-musical question-and-answer relation amount to? Perhaps just that they convey the character or physiognomy of questioning-and-answering, though without constituting a specific question-and-answer.


In The Brown Book Wittgenstein observes:

...if repeating a tune to ourselves and letting it make its full impression on us we say, 'This tune says something', and it is as though I had to find what it says. And yet I know that it doesn't say anything such that I might express in words or pictures what it says. (BBB 166)

As we have already noted, very often music makes on us the impression of a communicative act, and more specifically, a speech act or utterance. There is nothing more common than the sense that expressive music is speaking to us, and though the embedded claim is perhaps not to be taken literally, neither is it merely a weak metaphor whose cash value would be simply that the music seems meaningful, or that one gets something out of it. As Wittgenstein says, the impression of speech from music is so strong that we often feel impelled, however misguidedly, to try to ascertain exactly what is being said. And we are not satisfied, it seems, unless we can exhibit what is said in other than musical terms, and preferably verbal ones.

But should we be thus dissatisfied? Is there a communicative medium that should be privileged above all others which help to constitute the lived world? One is reminded of the anecdote in which Beethoven, having played for some visitors his latest piano sonata, was asked, 'But what does it mean, Herr Beethoven?", to which his response was just to play the sonata over again.

To require that musical thought, if it is to truly deserve that label, must be such that it can be rendered articulate or verbally paraphrased, would seem to smack of a double-standard. It would not impugn the claim to being thought of a stretch of discourse to note that what it conveyed could not, so far as we could see, be put into music. Why, then, should it be held to impugn the claim to being thought of a stretch of music that what it conveys cannot, in general, be put into words?


It is instructive to draw up an illustrated catalog of "thoughtful" actions that we can hear in musical passages, or that we can hear musical passages as instantiating: asserting (e.g. the opening of Schubert's Piano Trio No. 2, Op. 100),[] questioning (e.g. the opening phrases of Beethoven's Piano Sonata No. 18, op. 31 no. 3),[] musing (e.g. Schumann's 'Des Abends'),[] imploring (e.g. the flute introduction to Bellini's aria 'Casta Diva'),[] angrily despairing (e.g. the opening of Mahler's 2nd Symphony),[] menacing (e.g. the opening of Stravinsky's Symphony in Three Movements),[] defying (e.g. the opening of Beethoven's Fifth Symphony),[] cajoling (e.g. the sixth part of Vaughan William's Job: A Masque for Dancing),[] comforting (e.g. the first section in moderate tempo near the beginning of Fauré's Requiem),[] disapproving (e.g. the orchestral interjections in the first part of the finale of Beethoven's Ninth Symphony),[] and even nose-thumbing (e.g. the opening of the finale of Beethoven's Second Symphony).[] And there are passages in which can be heard meditating, applauding, bemoaning, heaven-storming, and so on. Can a medium capable of summoning up such a range of mindful actions be a domain in which thought is absent?


So one kind of musical thought is this: musical passages wearing an appearance of thoughtful acts, such as questioning, concluding, searching, and the like. But another kind, surely, is this: musical passages giving evidence of thought processes in their creator.

Let us bring the contrast between these two senses of musical thinking into clearer relief. One is thinking that seems as if embodied in musical process, that is, thinking that the music itself strikes us as being engaged in, or perhaps, that we are induced to imagine that the music is engaged in. Another is thinking in the composer that we take to be implied by musical process, that is, thinking that the music betokens on the composer's part. We might even go so far as to say, as Wittgenstein would urge us to do, that we directly hear the composer's thought in the musical process. For we are confronted with compositional choices at every turn that we cannot but regard as manifestations of mind.

Some examples of the first kind of musical thinking, of which we had a number of illustrations earlier, would be where music seems to be embarked on reflection, or to be lost in wonder, or where one musical phrase seems to answer the question posed by a preceding one. Examples of the second kind of musical thinking would be the assessment we infer Bach must have made in devising a fugue theme combinable with itself in counterpoint, or the judgment we suppose Mozart to have exercised, in composing a piano sonata, in designing a second theme whose character would contrast suitably with that of the first theme, or the vision we understand Beethoven to have displayed in opting for a C-sharp rather than a C-natural in the fourth bar of the opening theme of the 'Eroica' Symphony, setting up a tension exploited significantly later in the movement.

Again: embodied thinking in music is thinking we ascribe to the music, as something it appears to be doing, and has no identifiable object, whereas implied thinking in regard to music is thinking we ascribe to the composer, and has a quite definite object, namely the evolving composition itself.


Yet possibly the most important way in which music is a kind of thought does not reside in either music's frequent suggestions of thoughtful actions, or its implications of thoughtful fashioning on the composer's part. It may reside instead in the mere succession from chord to chord, motive to motive, or phrase to phrase at every point in any intelligible piece of music, whether or not there is any suggestion of recognizable extramusical action, or any implication of specific compositional deliberation. Call such musical thinking intrinsic musical thinking.

But why call such succession thinking? Obviously this is not enjoined by any rule of language. Still, musical succession has features that set it apart from succession in general. It is a purposive-seeming, goal-directed temporal process, an intelligent form of continuation in time, and one naturally subject to assessment in cognitive terms, such as 'coherence' or 'logicality' or 'making sense'. In addition, it is succession that we know emanated from a human mind, and that we hear under the influence of that postulate. If one insists that that is not enough for thinking, is one not just assuming that thinking is necessarily in words? And why should one assume that? Of course, if music be admitted to be thinking on the grounds just offered, the door is also open for dance, mime, and abstract film to be considered thinking as well. But such an implication is not, I think, to be feared.


Let me attempt to trace the process of embodied thought in the first movement of Beethoven's 'Tempest' Sonata, Op. 31 No. 2, one of the most rhetorical pieces of music in all Beethoven's oeuvre.[2]

The movement's opening gesture,[] a four-note rising motif in Largo tempo beginning with an arpeggiated A major chord, has about it a pronounced air of uncertainty and wonder (ms 1-2). It is followed by a descending Allegro motif in d minor which anxiously frets, ending in an Adagio turn of questioning character (ms. 2-6). Next the Largo motif returns, to be followed by a more excited variant of the earlier Allegro d minor music (ms.7-16), whereupon the music gathers resolve in a passage in octaves (ms. 17-18), before issuing in a full cadence on the tonic and the first episode of pure affirmation, a declaration in the bass and in Allegro tempo of the opening four-note rising motif, now wearing a minatory cast, but rounded off in the treble by a new motif, plaintive and supplicating.

Move now to the beginning of the development, where the rhetorical character of the movement becomes even more pronounced. The opening four-note Largo motif returns, with its initial arpeggiated chord extended, and is heard three times, each time outlining a different chord, ever more removed from the tonic (ms. 93-98). It is hard not to hear this as a deepening of the uncertainty and wonder expressed by the Largo motif on its first appearance, and the agitated minor key music that succeeds it as an exacerbation of the minatory proclamations of the exposition (ms. 99-118). Consider, finally, an episode at the end of the development prior to the recapitulation, where the music takes on even more unmistakably than before the appearance of a mindful agent. After six measures of sustained chords and a descending passage in bare octaves (ms. 133-142), which strike one as clearing a space for reflection, there follows a recitative with all the earmarks of a soliloquy, in part due to the thinning out of the musical texture and the starkness of the melodic line that remains (ms. 143-158). This utterance is at turns meditative, questioning, and anxious - the last of these due to an eruption halfway through of the fretting motif from the sonata's beginning, though in muted form.[3]

As one attends closely to this movement, one cannot fail to be struck by the mind manifested in its progression, and more specifically, by the series of communicative acts incarnated in the music itself. Yet whose mind is so manifested, one may ask? In one sense, it is the mind of the imagined agent of those acts, what one may call the persona of the music. In another sense, it is the mind of the composer, who has in effect constructed, or caused to emerge, the persona to whom the communicative acts heard can be directly attributed.

But at this point an objection may be raised. Let us grant that the music of the first movement of the 'Tempest' Sonata exhibits a series of images of thought-filled actions. Does that show that the movement literally constitutes thought, or is literally a thought process? Here we must add that this movement of the 'Tempest' Sonata is, of course, not a random or accidental concatenation of such images. It is, rather, a meaningful concatenation of them, one that makes sense to us, and one that induces us to imagine a mindful agent of those acts, of whose mental life the music then appears as the narrative. But, the objector continues, all that shows is that music, or at least some music, is the narrative of an imaginary thought process, not a thought process itself. But how much of a difference is that?


Consider now, in a change of gears, how the sorts of thinking involved in the activities of composing, performing, improvising, and listening music saliently differ. One way they differ is this: following is a key idea in the last of these -listening - but not the others. To understand music to which one is listening, at bottom, to follow it, that is, to experience its evolution in an involved way, exercising certain perceptual abilities and emotional sympathies, anticipating and projecting that evolution, responding appropriately in the moment to each twist and turn. That following music - as opposed to mere listening, or half-listening - is a form of thinking is evidenced by the near impossibility of doing any other thinking, of an unequivocal sort, at the same time. Musical process absorbs and effectively fills the mind that attends to it with any seriousness.[4]

But following is not, it seems, of the essence of composing, performing or improvising. Rather, determining - that is, the determining of notes as constitutive of a work - would seem to be the essential activity of composing, interpreting - that is, the interpreting in concrete sound of notes already given - the essential activity of performing, and generating - that is, the creating of music on the spot, subject only to relatively loose constraints, the essential activity of improvising.

Now, on the one hand, these activities of determining, interpreting, and generating music might all be classed as productive, whereas that of following music might by contrast be classed as receptive, though that should not make one lose sight of the anticipatory and constructive element in the activity of following music by ear. On the other hand, the activities classed as productive in a sense also involve actions of following: the composer follows one measure with another as he composes, the performer follows his reading of one phrase with his reading of the next, the improviser follows what he has just played by playing something else. But those sorts of following are manifestly not the same as that involved in listening. In the one case what is central is the tracking of what already exists, whereas in the other case what is central is a bringing into being at each step.


As an illustration of the thinking involved in improvising, I turn to Stan Getz's solo on his famous recording of Antonio Carlos Jobim's 'The Girl from Ipanema'.[]

Knowing Jobim's basic tune, and hearing Getz's treatment of it, we marvel at the melodic, rhythmic, harmonic and timbral possibilities that Getz brings out that we didn't suspect the tune possessed. I would single out for remark just the high melodic leap Getz takes about midway through the repeat of the first strain - where the lyric, significantly, has the words 'sways so gentle' - and the playful one-long three-short rhythm he introduces in the refrain in place of the original dotted one. If this is not thinking in sound, then what is it? Surely music that in the span of a mere 40 bars manages to suggest a whole way of being - for my part, I have often wished to live some of the time as that solo sounds - cannot be music in which thought fails to be present.

Note that with improvised music a distinction invoked earlier - between embodied thinking in music, wherein music presents us with images of thoughtful actions of an imaginary persona, and implied thinking in music, whereby musical process betokens or signifies thought in the compose - has almost no purchase. Is it Getz or is it the music's persona who exults in that high-flying turn mentioned a moment ago? Is it Getz or is it the music's persona who "gets down" in that one-long three-short rhythm? And is there much importance in deciding?

Music, we may remark, and improvised music especially, stands as one of the supreme exemplars of the fusion of inner and outer in mental life that Wittgenstein was at pains to underline, a fusion that in the last analysis invites us to transcend the opposition between inner and outer, a transcendence that was perhaps the ultimate goal of Wittgenstein's philosophy of mind.


If music is in some sense thinking, then, as noted earlier, bad music should tend to have the character of bad thinking. One of the ways we mark out bad thinking is by the epithets 'dumb' and 'stupid'. Hence, if music is thinking, at least some bad music should wear the appearance of dumbness or stupidity. One clear example, I would say, from an early work of a great composer, Franz Schubert, is the beginning of the finale of his String Quartet No. 7 in D.[]

Now why is this music so stupid, and thus bad? The problem, zeroing in on just the opening six bars, which are even repeated, is that it consists in two largely unrelated ideas, and more specifically, of a first idea of utter banality followed by a second idea which is an emphatic closing gesture, one entirely unjustified by the meager four bar ditty that precedes it.

Compare this, though, to a little known piano sonata by Beethoven, the Sonata No. 16, op. 31, no. 1, in G.[] This music once also struck me as somewhat stupid, because of its quirky premature use of a closing gesture similar to that featured in the Schubert, and the similarly unpromising character of its melodic materials. On longer acquaintance, though, the music seems anything but stupid. The Beethoven is, despite its modest materials, miles beyond the Schubert in development, flow, and organicness.

Of course, not all bad music is stupid, or bad because it is stupid. Some music is bad because it is bland, or bombastic, or bathetic, or unbalanced - just to stick with 'b' words. This is to be expected, since even if music is thinking, and thus sometimes bad in the way thinking is generally, music is other things as well - movement, gesture, pattern, expression, narration, depiction - and can thus exhibit failures in those respects, and not just fallings-off from the cardinal virtues of thought, such as cleverness or cogency.


Finally, what is the connection between Wittgenstein's views on understanding music - that it is manifested by a complex of behaviors, such as illustrative gestures, apt comparisons, suitable hummings, and appropriate movings to music, that its criteria are neither inner acts of comprehension nor articulate paraphrases of musical content but a range of outwardly demonstrable responses and capacities - and the claim that Wittgenstein also endorsed, which has been my focus here, that music, no less than language, incarnates thought?

Here is one way to articulate the connection or connect the dots: Both music and language are forms of thought. Understanding music should therefore be analogous to understanding language. The former, like the latter, is a matter of use, that is, of knowing how to operate with the medium in question in particular communicative games, in particular contexts.[5] But knowing how in regard to music, as with knowing how generally, does not consist in propositional knowledge but rather in behavioral and experiential abilities and dispositions. Hence if music is thought we should naturally come to understand it as we come to understand thought in words; not by learning how to decode or decipher it, but by learning how to respond to it appropriately and how to connect it to and ground it in our lives.


Parallel to the question at the heart of this essay – Is music thought? – would be the question, Is speech thought? In other words, one might wonder whether a stretch of intelligible verbal discourse was literally thinking, or was instead only the expression of literal thinking, that is, certain occurrences or processes in the mind of the speaker. Wittgenstein, of course, argued that there is no reason to think of thinking as a purely inner process, of which our observable behavior, however intelligent, can be no more than the outer shell, and thus no reason not to recognize as thinking the normal deployment of language. But for those who balk at the idea even that intelligible verbal discourse is thought, that is, that thought has outer as well as inner forms, and who claim that such discourse only manifests thought, the central claim of this paper can be suitably recast. It becomes this: Intelligible music stands to literal thinking in precisely the same relation as does intelligible verbal discourse. If that relation be not exemplification but instead, say, expression, then music and language are, at any rate, in the same, and quite comfortable, boat.





To refer to this article: click in the target section