BOOKS: FICTION: Drafted into a battle for the brain

KINDS OF MINDS by Daniel Dennett Weidenfeld pounds 11.99

In 1991 Daniel Dennett published a book called Consciousness Explained. Its readers duly searched for Dennett's explanation of consciousness, but it was hard to find. In fact it seemed to consist of the denial that consciousness existed. Consciousness Explained came to be known as Consciousness Ignored.

Faced with this reaction, Dennett told a story about himself (or so the story goes): Dan Dennett goes out to take the air and meets a group of people. He tells them what he has done. He has taken on our theories of mind and swept them clean. He has abolished a whole zoo, a vast menagerie, of misunderstandings, fictions, dubious and unwarranted posits - the ordinary conception of consciousness among them.

His interlocutors are very impressed. They're very interested. But they're also very interested by something else. Why is there a very large elephant standing behind Dennett? Why is it following him wherever he goes? Dennett has no idea what they're talking about, and swivels round. There is the elephant. It is the elephant Consciousness - huge, faithful, unshakable, unexplained.

If Dennett told this story, he didn't learn from it, for the same thing happens in his new book. But although Kinds of Minds is skewed and partial as an introduction to the study of the nature of mind, because of its problems with consciousness, it's still useful. A lot of what it says is true, and Dennett's emphasis on the continuities between different kinds of mind, as one moves up the evolutionary scale, is as valuable as his description of the discontinuities. The book makes good use of biological data and of work in artificial intelligence - among the minds (or "minds") that Dennett considers are those of ants, dogs, robots, antelopes, people, pigeons and plovers - and it's much less bossy and self-concerned than Darwin's Dangerous Idea, which he published last year (a couple of the chapters in Kinds of Minds are directly adapted from Darwin's Dangerous Idea).

The problem of consciousness remains, however. Dennett reaffirms the model of consciousness which he put forward in Consciousness Explained - the "Multiple Drafts" model, according to which "mental contents become conscious ... by winning the competitions against other mental contents for domination in the control of behaviour".

Even if this were true, however, it wouldn't give us any sort of explanation of consciousness, or of how it arises. Dennett tries to anticipate this objection: "Suppose all these strange competitive processes are going on in my brain", and that the conscious ones "are simply those that win the competitions. How does that make them conscious? What happens next to them that makes it true that I know about them?" Dennett replies that such questions betray a deep confusion: "what you are ... just is this organization of all the competitive activity." But this just ignores the problem of consciousness.

He has another go: "What makes a mind conscious" is "not what it is made of, or how big it is, but what it can do". But this is no better. It's tautologous if you include having conscious experience among the things the mind can do, and it's false if you don't.

"Can it concentrate? Can it be distracted? Can it recall earlier events? Can it keep track of several different things at once? Which features of its own current activities can it notice or monitor?" These are the sorts of thing Dennett cites as proofs of consciousness, but they get us nowhere. It's either true by definition that things like concentration and distraction involve consciousness - in which case Dennett's case amounts to the claim that the presence of consciousness is proof of the presence of consciousness. Or it's not true by definition: in which case we can imagine a computer that can be said to do all these things, although it has no conscious experience at all.

Consider the simplest of cases: the case of detecting the colour red. Taken in one way, this obviously requires having conscious experience. Taken in another way, it implies no such thing: it is easy to build robots that can distinguish effortlessly between colours, and pick out the one red thing in the room, without having any conscious experience.

Dennett cannot help us on consciousness; he has a quasi-religious blindness on the topic. He also has some very strange views on ethics and epistemology. All these things severely limit his discussion of mind - and one blushes for him when he tries to be "fair" to "poor old Descartes", who has no need of his sympathy. Still, Kinds of Minds is worth reading. It has some attractive passages, and the ways in which it is wrong are as instructive as the ways in which it is right.