Peter Kivy, “The Corded Shell” (1980)

Derek Matravers (The Open University)
MPSG Retrospective Review No. 2020:4
Published 26. March 2020

In a 2017 paper Jerrold Levinson wrote of Peter Kivy’s 1980 book, The Corded Shell (Kivy 1980):

In the beginning – or, more exactly, the seventies, when I was in graduate school at the University of Michigan – was the void, and darkness was upon the face of the waters. Philosophical reflection on the experience, meaning, and powers of music by analytic philosophers was almost non-existent. And then, as the 1980s dawned, came Peter Kivy. Suddenly there was light, and analytic philosophy of music was born.

Levinson 2017: 269

The Corded Shell was re-published in 1989, with four extra essays, under the title Sound Sentiment (Kivy 1989)[1]. In the book, Kivy discusses descriptions of music in emotive terms – for example, ‘that music is sad’. This example is not intended by Kivy (or indeed by anyone else) as a serious contribution to music criticism. It is a placeholder sentence, designed to raise philosophical issues and Kivy rightly lambasts anyone who took him to think otherwise (181). What could such a sentence mean? Clearly, it does not mean what it would mean if I were to say ‘Peter is sad’. That would mean that Peter felt sad, and music, being insensate, does not feel anything. Hence, we could explore two further options: that it means something like ‘the music makes me sad’ or ‘the music sounds like a sad person’.

The first of these options is the arousal theory. In its crude form it holds that, for some emotion E, music expresses E if music arouses E in some suitably qualified observer.[2] Kivy brought forth an array of objections to this view, concluding that ‘sadness is a quality of the music, not a power of the music to do things to the listener’ (23). Instead, he held that ‘expressive qualities are genuine “objective” qualities of music’ (60). This view was not particularly original, as Kivy was happy to acknowledge. It had been put forward by O.K. Bouwsma (Bouwsma 1954) and Monroe Beardsley (Beardsley 1981) in the fifties and sixties. However, what was original, and compelling, was the way in which Kivy filled out the account.

Kivy drew on the work of the Florentine Camerata who, he claimed, developed a style of music that resembled the rise and fall of the speaking voice and hence, by this resemblance, expressed whatever emotion would have been expressed by that voice. Kivy generalised from this, claiming:

We hear sadness in this complex musical line, we hear it as expressive of sadness, because we hear it as the musical resemblance of the gesture and carriage appropriate to the expression of our sadness. It is a ‘sound map’ of the human body under the influence of a particular emotion. (53).

However, as Kivy realised, the account is incomplete. Why should the major mode and diatonic character of music contribute to its positive character? It is not through resemblance. Kivy’s solution was to appeal to convention: ‘they are merely customary accompaniments of happy emotions and are recognised as such when they occur’ (80). Thus we have the theory: expression of an emotion in music is a matter either of contour (resemblance) or of convention.

The theory explained enough, and was sufficiently grounded in argument and evidence, for Kivy to have faith in it. Nonetheless, he was honest enough to foreground, and attempt to deal with, two apparent problems. The first was that expression seems a unitary (or vaguely unitary) phenomenon; what I mean by ‘the music is sad’ does not vary depending on the piece of music I am talking about, yet Kivy’s account (either contour or convention) suggests it should have at least these two different ways of meaning. Kivy struggles with this, attempting to get to a unity by suggesting that what is now convention had its origins in contour (82). Even if this succeeded it is not obvious it would solve the problem. The second issue was that, even if it is true that some music resembles the human body in the throes of expressing emotion, it resembles a number of other things as well; in Kivy’s inimitable list, ‘the waves of the ocean, and, for all I know, the rise and fall of the stock market or the spirit of capitalism’. Why does it not express those? Kivy’s reply to this draws on an earlier discussion to the effect that we cannot help but see the face of a St. Bernard dog as expressing sadness: ‘we, for whatever reason, tend to animate our perceptions, and cannot but see expressiveness in them, any more than we can help seeing expressiveness in the St. Bernard’s face’ (62).

The fundamental flaw in Kivy’s theory, of which these two problems are merely a part, is that he failed to distinguish two quite different questions. The first, the causal question, is ‘what are the properties of music that cause the experience of expression?’ The second, the constitutive question, is ‘what constitutes expression?’ The first is an empirical matter – the kind of question that is settled by scientific enquiry. The second illuminates the content of the claim: what is meant by ‘the music is sad’? In answer to the second question people have traditionally attempted to throw light on the nature of the experience; to provide a re-description of the experience of expressive music which makes clear the relation between music and emotion. The causal and constitutive questions are independent of one another – the answer to one does not determine the answer to the other.

Inasmuch as The Corded Shell focusses on one of these questions, it is on the causal question. Hence, Kivy speculates that what causes us to hear music as sad is that the music ‘is a “sound map” of the human body under the influence of a particular emotion’. Of course, if one is trying to determine what causes some type of thing (the experience of expression), there is no reason to think that there will be only one type of cause. Hence, Kivy is quite right to identify as many causes as he needs to – hence, a conventional link can also be a cause. In answer to the causal question, then, there is no problem with having both contour and convention.

However, when Kivy takes this to be an answer to the constitutive question there is a problem with having both. The answer to the question ‘what is meant by “the music is sad”’ cannot be ‘the music has the right contour’ or ‘the music has the right conventional link’. As I said above, expression is a unitary (or vaguely unitary) phenomenon: the meaning of ‘the music is sad’ does not change depending on the cause. My saying ‘It is raining’ might be caused by my seeing puddles or it might be caused by drops falling on my face, but the meaning of ‘it is raining’ does not change between these instances.

The great irony of The Corded Shell is that its lasting theoretical contribution is not its official theory – that expression is a matter of contour or convention. Rather, it is in the very brief replies Kivy gave to cover the pickle he had got himself into by not using that official theory to answer both the causal and the constitutive question. Those replies are too brief for us to formulate a definitive view on Kivy’s behalf, but we can extrapolate from them to two currently influential theories.

The first picks up on the claim that we cannot help ‘seeing expressiveness in the St. Bernard’s face’, and construes expression as a matter of experienced resemblance. In the same way as I can hear a foghorn as resembling the sound of a whale, I can hear music as resembling the human expression of emotion (well, alright, it will have to be cross-modal – hearing a sound as resembling an appearance – but that can be sorted). This is what it is to experience music as expressive – however caused. There is an odd historical detail here, which is that Kivy’s book appeared at the same time as a paper by the New Zealand philosopher, Stephen Davies in which Davies, independently, put forward an expressive resemblance account (Davies 1980). The oddity is that while Davies was content to acknowledge Kivy’s work, Kivy, as far as I am aware, never acknowledged Davies. Davies went on to develop the expressive resemblance account in considerable detail (see in particular, (Davies 1994)).

The second picks up on the claim that we ‘tend to animate our perceptions’. This has affinities with accounts which construe expression in term of our disposition to hear sounds as the expression of something – call it a persona. Jerrold Levinson has developed this with considerable force. What it is for music to be expressive of some emotion is for the right kind of listener to be disposed to imagine, of the music, that it is the expression of that emotion by a persona (Levinson 1996).

The Corded Shell was undoubtedly significant in putting analytic philosophy of music on the map. It seems to have spurred Levinson to abandon his studies of the metaphysics of properties (the subject of his PhD) and turn his attention to the philosophy of music – a sphere in which he has made a distinguished contribution. Stephen Davies, as we have seen, was already writing in the field without Kivy’s help. Roger Scruton and Malcolm Budd both started writing on music in the early 1980s, although the extent to which they were influenced by The Corded Shell is unclear: Kivy gets only one footnote in Scruton’s eight early essays (reprinted in (Scruton 1983)) and no mention at all in Budd’s magisterial Music and the Emotions (Budd 1985). The Corded Shell, in sum, is a slightly strange beast. As we have seen, what it got right, it got right somewhat by accident. However, it is in the personal history of a lot of us; it was a book we read and (in my case at least) spent some time trying to refute. It also gave at least some of us entrée into Kivy’s company (he sadly died in 2017). To end on a personal note, it was that introduction, rather than The Corded Shell, that enriched my life philosophically and in so many other ways.

  • Beardsley, Monroe C. 1981. Aesthetics: Problems in the Philosophy of Criticism (Hackett: Indianapolis).
  • Bouwsma, O.K. 1954. ‘The Expression Theory of Art.’ in William Elton (ed.), Aesthetics and Language (Blackwell: Oxford).
  • Budd, Malcolm. 1985. Music and the Emotions (Routledge and Kegan Paul: London).
  • Davies, Stephen. 1980. ‘The Expression of Emotion in Music’, Mind, 89: 67-86.
  • Davies, Stephen. 1994. Musical Meaning and Expression (Cornell University Press: Ithaca).
  • Kivy, Peter. 1980. The Corded Shell: Reflections on Musical Expression (Princeton University Press: Princeton, NJ).
  • Kivy, Peter. 1989. Sound Sentiment: An Essay on the Musical Emotions (Temple University Press: Philadelphia).
  • Levinson, Jerrold. 1996. ‘Musical Expressiveness.’ in, The Pleasures of Aesthetics (Cornell University Press: Ithaca).
  • Levinson, Jerrold. 2017. ‘Peter Kivy and the Philosophy of Music (1980-2002)’, The British Journal of Aesthetics, 57: 269-82.
  • Matravers, Derek. 1998. Art and Emotion (Oxford University Press: Oxford).
  • Scruton, Roger. 1983. The Aesthetic Understanding (Methuen: London).

[1] All further page references are to this edition.
[2] I tried, unsuccessfully, to defend a less crude version in (Matravers 1998).

Sign up for periodic announcements via our Newsletter!