Stream of the self-conscious

Only in the last quarter of this century have psychologists come stumbling out of behaviourism's "long night". That they have been able to do so is down to people like Rom Harre, who warned (long before it became fashionable) that the habits of this problem child would end in a kind of desertification in psychology. This book opens with Harre exploring further his interest in the methods of psychology and he advances once more his taste for a "discursive psychology".

This preference for discourse as data directs and shapes the book. Harre says that his aim is to make a case for "the methodological thesis that the empirical study of grammar is the route by which the relevant forms of human experience can best be revealed". He tackles this by showing that "the grammatical function of the first person is indexical and not referential". Most of the book opens up and explicates this theme in relation to the self.

In doing this he is refreshingly dismissive of some of the nonsense of popular and professional psychology. The current rash of consciousness-explainers such as Francis Crick get short shrift, since their neuronal explanations are of the same order as Gilbert Ryle's ghost in the machine, where "both cause and effect are of the same category, material entities". He warns also against looking for something else - like "self-esteem" - "in me". Self-esteem is not a property, he reminds us, let alone a causal one. Psychologists must be wary of creating nonsense phenomena, snarks and boojums, with nonsense properties.

Brilliant as elements of it are, I had a number of problems with this work. The first is that there seems to be something contradictory about a book which repeatedly draws on the later Wittgenstein, but then proceeds to analyse doggedly and atomistically the contribution of language to an understanding of self. Language (particularly the word "I") is proffered as the key. A good example is in Harre's examination of the sentence "I am not here". He says that it seems like nonsense. Its putative nonsense is, according to him, down to the fact that the "I", being a space-and-time referent, does some of the work of "here". So it is nonsense, apparently, to say "I am not here". But the problem is that reality confounds and supersedes the linguistic analysis - and this, surely, was Wittgenstein's point in his later work. Anybody who lives in a family with a 16-year-old daughter and a ringing phone has heard the sentence, "I am not here". I know exactly what it means, and it makes perfect sense.

This leads to my second problem: reading too much Wittgenstein can be bad for one's writing style. Just when you think you have understood something by Harre, you come up against: "should we use these concepts for understanding habitual actions in preference to the causal concepts that would recommend themselves if we assimilated the habitual fully to the causal?" Or you meet a neologism like "individuatable". In short, it is not an easy read. Do not be lured into a false sense of security by the word "introduction" in the title. It is more like a stream of consciousness - albeit a sometimes compelling one.

The stream style leads to an idiosyncratic itinerary, which takes detours around some seemingly obvious stopping-off points. Why, for instance, given the heavy emphasis on "I", is there no reference to developmental psychology's fascinating analysis of young children's developing employment of "I" as a deictic shifter - a word that changes its meaning depending on who is uttering and who is listening.

If you are prepared to snatch at the brilliant bits and pass over the gnomic stuff in silence, you might agree with some of the hyperbole of the blurb: "this landmark workI profound reappraisalI masterly analysisI tour de forceI" It is certainly the last two of these, but possibly not the first two.

Gary Thomas is chair in education, Oxford Brookes University.