In what is styled by its publisher a ‘major new work by the pioneer of philosophy of music’, Peter Kivy argues that absolute music has no representational or narrative content. (Peter Kivy, Antithetical Arts, On the Ancient Quarrel Between Literature and Music, Clarendon Press, Oxford, 2009.) He says that we cannot as yet say philosophically what its human significance is. Anything of quasi-moral force is ruled out (so much for Plato, Rousseau and Nietzsche). And while Kivy has some sympathy for the view that music can produce a quasi-religious ecstacy, as a secularist he has little time for there being anything objective about that. In the end music produces pleasure, Kivy says, and pleasure is good: not the only good, but a good nonetheless.

Kivy has no answer to the Benthamite who prefers push-pin to the late quartets, but then ‘one need not defend absolute music against push-pin, or the narrative and representational arts, for that matter. For those who have experienced its ecstatic effects, that is all the defense it needs, or can ever have, push-pin notwithstanding.’ But, what of the explanation of this power music has? It cannot rely in music's narrative, dramatic or literary content, for (according to Kivy) it has none. ‘For me, its power over us remains a divine mystery. Or, in other words, I haven't the foggiest.’

As this is more or less the end of Kivy's book, readers may be forgiven for feeling a little disappointed. During the course of his 261 pages a lot of time is spent criticising specific attempts to supply absolute music with representational or narrative schemata, and to detach examples from any specific emotional meaning. A whole chapter is spent on Shostakovich, specifically on an attempt to interpret his Tenth Symphony as having to do with the mood in the USSR following the death of Stalin, in the course of which the by now somewhat familiar objections to Volkov's Testimony are trotted out. But, however tenuous a particular interpretation of Shostakovich might be – and to see his music in such historically specific terms might well be unjustifiable – this is not enough to see absolute music as always and irremediably abstract. We do not need to see every piece of absolute music as if it were Til Eulenspiegel or Hamlet in sound, to take two other examples Kivy discusses.

Actually the case of a Strauss tone poem is interesting. It is true that his Don Quixote does take us through various incidents from Cervantes, marked in the score. But that is not why one continues to listen to it, or what its power consists in. The programme can be seen as just the scaffolding which holds the structure together, much like sonata form might have done for Haydn and Mozart. Strauss's Don Quixote is a powerful piece of music because, among other things, it evokes the poignancy and pathos of the Don, it conveys a sense of his blundering and his muddle-headedness, and (in a way Cervantes only hints at, and which might be too sentimental for some), of his ultimate nobility. At least, that is what one might say if one were not Kivy, or, to put Kivy in context, Stravinsky in his purist mood, claiming that music expresses nothing but itself. (Rich that from the composer of The Rite of Spring, one might think.)

The weakness of Kivy's position is that it is belied not just by the experience of absolute music – which is often that of a dramatic essence without the narrative content or detail – but also by what composers have said about their compositions. Think, for example, of Beethoven's Eroica or of his Heilige Dankgesang; of Berlioz's Harold in Italy, of Smetana's Ma Vlast or of Debussy's La Mer; or of the programme for Mahler's Third Symphony. The fact that Mahler withdrew this programme does not mean that it cannot help us to articulate what we hear, that it does not make sense of what we hear when we hear the music. Are we to suppose that when composers have not themselves written titles to their works, very similar in feel and tone to those for which they have, we suddenly move from powerful and concrete drama to something undramatic and purely pleasurable? If so, we would have to conclude that while there is something highly illuminating in Proust's matchless account of the first movement of La Mer (which does not rely on any literal reading of sounds into things or vice versa), there is absolutely nothing in his equally matchless accounts of Cesar Franck's violin sonata and string quartet, even though they masquerade as descriptions of works by the fictional Vinteuil.

And are we to suppose that there is no connexion between the meaning of the music of a song and the same or similar music when it appears without words (e.g. between Schubert's use of the same melody in his song Einsamkeit and in the second movement of the String Quartet in G, D 887, or in Bach's use of similar music in his Passions and in purely instrumental pieces)? Kivy speaks of the ancient quarrel between literature and music. But why should there be a quarrel? Might there not be something right in the claim of those neo-Platonic thinkers of the Italian renaissance (to whom Monteverdi was deeply indebted), who affirmed that music was ‘nothing other than speech, then rhythm, and lastly sound’ (Giulio Caccini, preface to Nuove musiche, Florence, 1602). In other words, far from there being a quarrel between music and literature, musical sound is itself grounded in human logos.

Deryck Cooke once based a whole ‘language of music’ on what he took to be the emotional content of particular musical phrases when used in the context of Lieder or other vocal works. And while this may have been too mechanistic an approach, Cooke was surely nearer the truth than our pioneer philosopher of music, as was Mahler who, with forgivable exaggeration, said that ‘the symphony is the world, it must embrace everything’. Maybe not quite the whole world, even in Mahler 3, but the human world and bits of the world. If the philosophy of music is unable make sense of any of this, and for all its pioneering spirit, for music lovers the words ‘must try harder’ will inevitably come to mind.