Emotion picture show

This book presents a battle. It is a battle between the orthodoxy of the cognitivist theory of expression in art and the arousal theory, which is generally thought to be false. At the outset it is refreshing to find a philosopher writing on aesthetics extensively illustrating his arguments with our experience of music in a field usually dominated by the visual arts.

Many see the arousal theory as tempting but discredited. Derek Matravers claims to offer a defensible and subtle version of the arousal theory able to answer and indeed surpass in explanatory power the views of its cognitivist critics. The route to this conclusion is pretty long and winding; but it is worth the effort, especially as the book is well written. The distinctive feature of the cognitivist account is that the emotional expression of artworks is describable in terms of properties we come to recognise the work as having without reference to a person's experience of it. Against this view Matravers argues that we can make sense of a work being expressive only by reference to the phenomenology of the experience aroused in - caused in - the experiencer: some non-cognitive, non-representational state of mind. So Matravers's view can be called a "causal arousal theory". The core of the theory can be stated thus: "A piece of music expresses an emotion if it causes a listener to experience a feeling (that is a different state), and this feeling will cause him to believe that the work expresses sadness." This view has a lot to commend it once it is properly understood. At the outset it fits with our intuitions about our relationship to artworks we find expressive. Surely the fact that a piece of music is sad has some intimate connection with the effect it has on the listener.

One of the most serious objections to the notion that reference to the mental state of the experiencer is required in talking of a musical work's expressive nature is that it points dangerously in the direction of saying that the work itself is unimportant to the point of elimination. If the expressive value in a work is its ability to cause a certain mental state, then why bother with the work, why not just use drugs, say, that produce the same effect? In principle one could ingest Bruckner's Eighth without having to bother listening to music.

Matravers's answer is cunning. Part of it is that in the case of music causing a feeling within us, our attention is directed to the music; we have a mental representation of the music in a way we do not of the drug.

A major component of Matravers's theory is an analogy with colour and other secondary properties. While it is the case that colours are caused in us by certain determinable primary properties (so correctly we say the object is red) it is also the case that seeing something as red is not co-extensive with any single determinate set of physical properties. There are all sorts of physical reasons why an object might look red. This means that in order to say which objects are red we are bound to refer to what people see, not just to physical properties. Such a property is dispositional: it is present if in appropriate circumstances it can cause a certain experience. Not to confuse matters, music is made up of secondary qualities (sounds), which unlike the primary properties that cause our experience of colour we do notice when we have the emotional experience they cause. But this does not undermine the analogy. If we say a piece of music is sad we are quite correct to say the music is sad insofar as we mean that it possesses certain basic properties that cause in us certain experiences. Moreover we need not be aware of the exact nature of these basic properties any more than we need be aware of the primary physical qualities of an object in order to be in a position authoritatively to assert that it is red.

It may look like a serious criticism of Matravers's theory that if it were true then music could be built up formulaically just as a blind chemist might produce red paint. He firmly grasps this and agrees that to a degree this is possible - and surely he is right - although at the outset the blind chemist would need to know which formula produces red paint. But for the results to be sure and satisfactory someone with the appropriate powers of discrimination is required on the receiving end. This is partly because basic properties operate in groups.

This is a complex book, but it is also an important and convincing one and should be read by anyone interested in the relation between art and the emotions.

John Shand is associate lecturer in philosophy, Open University.