Taking folk psychology seriously

In Charles Dickens's Hard Times, Mrs Gradgrind remarks from her sickbed: "I think there's a pain somewhere in the room, but I couldn't positively say that I have got it." Despite its manifest absurdity, such a remark would not be entirely out of place with current thinking in the cognitive sciences, for, notwithstanding the recent interest in consciousness studies, theories of cognition still tend to deal with abstract disembodied processes. Even among those who take consciousness seriously (unlike Daniel Dennett and his endlessly recycled Rylian metaphors), there is a tendency to explain perception and other cognitive processes in terms of computational functions and then we are left to puzzle over what on earth the experience is for.

E. J. Lowe, professor of philosophy at Durham University, has little time for such orthodoxies - to him cognition cannot be explained without recourse to subjective experience. The book is both an outspoken attack on physicalism and a delightfully old-fashioned plea to take our commonsense intuitions of personhood seriously - the word "confabulation" does not merit a single entry in the index, and "folk psychology" is not treated with derision. Despite its brevity and relaxed accessible style, it offers a convincing demolition of a range of physicalist dogmas, including Jerry Fodor's modular theory of perception and the Frege-Wittgenstein theory of thought and language. Lowe's arguments are varied but often based on pointing out the artificial intelligence-based agenda of functionalism. He is particularly critical of David Marr's computational theory of perception, quoting Steven Pinker: "Subjective experience is noncontroversially epiphenomenal if one subscribes to the computational theory of the mind in general. In no computational theory of a mental process does subjective experience per se play a causal role I subjective experience, if it is considered at all, is assumed to be a correlate of I processing."

Lowe argues that this sort of theory may be useful for designing artificial intelligence (AI) devices, but tells us very little about human perception, which is based on conscious experience and general intelligence. He considers the evidence from neuropsychology but concludes that blindsight and other visual agnosias fail to overturn an intelligence-experience-based theory. The reason AI theorists give so little emphasis to this is because general intelligence and conscious experience are so poorly understood and almost impossible to model.

So much for the hatchet job, but what about Lowe's own theory? He would describe himself as a neo-Lockian empiricist, and is happy to defend Locke's volitional theory of action against its many critics. Consciousness and the self - the subject of experience - are strongly emergent properties of mind, language and culture. But Lowe would also describe his theory of selfhood, as "non-Cartesian substance dualism", at which point I began to part company. His dualism is based on the simple observation that a volitional act, such as the raising of an arm, is a discrete event, and the only discrete "cause" that we are aware of is the act of willing itself. If we look into the brain all we find is a continuous, chaotic branching neural hotchpotch, with nothing that could be seen as a discrete cause. However, he refuses to accept Descartes' doctrine of essences, whereby something has to be either extended or immaterial. But I do not see how this fits with the idea that the self is a simple substance. He is clear that the self is not the body or part of the body, but never quite tells us what physical properties it does have.

The biggest problem is with the idea of mental causation. Lowe is fully conversant with the physicalist objections such as conservation of momentum and causal redundancy, but concludes (rightly) that such critiques are basically circular. The "physical" is defined in terms of causal closure etc. The last thing I want to do is defend physicalist dogmas, but this was the weakest part of the book. And yet all Lowe's observations of our sense of self, volition and agency are true, and should not be written off as "folk psychology". I was left feeling that "non-Cartesian dualism" was not quite radical enough and that the answer might lie in the infinitely wilder theories of Spinoza or Schrodinger.

Keith Sutherland is publisher, Journal of Consciousness Studies.