The title, the cover illustration and the jaunty photo of the author all conspire to suggest that this book is aimed at the non-specialist. The book's scope - "the history of philosophy, the theory of knowledge and philosophy of science, of logic and language, of space, time and causality, of society, ethics, politics and aesthetics . . . dialectics and the sociology and philosophy of philosophy" - supports this view. The fact that the whole lot is trussed up in 267 pages appears to confirm it.

However, the epigrammatic references to philosophers and their views that perforate the text can only make sense to a sophisticated reader. But such a reader will not expect to be illuminated on so many topics in such a short space and will be irritated by the imprecise off-hand references. Roy Bhaskar claims that "the most influential correspondence theories in this century have been Wittgenstein's picture theory, Tarski's semantic theory and Popper's theory of verisimilitude." The last of these is not a theory of truth; the second is a theory devoid of metaphysical pretensions; and the interpretation of Wittgenstein's early theory of language long ago moved on from the crude empiricist reading that Bhaskar favours. What then, is Bhaskar doing?

He is advertising a form of transcendental realism that he calls critical realism. It is, however, unclear who is supposed to be taken in by the advert. Those who know Kripke's work will be irritated to find his concept of rigid designation misreported. Those familiar with the problem of personal identity will hardly be moved by a one paragraph resolution that recommends we take into account that our "geo-historical constitution . . . is an ongoing affair". That will not provide a definition of identity through time, for as time passes so do the geo-historical conditions. Those who have studied ethics will be unimpressed by the bravado of the chapter on "Living Well" that welds together consequentialist, deontological, feminist theory, virtue theory, Habermasian, communitarian and universalist approaches.

The range of issues is appropriate to an accessible introduction to philosophy, but it is done in a way that renders the activity of philosophy opaque. Bhaskar notes that he has often been charged with employing neologisms. Fair criticism, to which he fairly responds that an abstract terminology is often a prerequisite for intellectual progress. But Bhaskar's love affair with the proliferation of terminology and taxonomies is more than a stylistic irritation.

ÁñÁ«ÊÓƵ

For a book on the problems of philosophy there is precious little argument within its pages. Bhaskar's instinct when confronting a problem is not to argue through it, to examine its presuppositions and to explore argumentative avenues of rival theories. His instinct is to name and classify, to produce abstruse diagrams that are intended to plot the conceptual geography of the philosophical terrain. The book's 64 diagrams are scattered into the text with minimal explanation. Even when bearing helpful titles like "Four-Planar Social Being Encompassing the Social Cube" the relation between picture and text is not spelt out. Figure A.12 covers page 234 with 42 separate linking devices of lines, broken lines, arrows, broken arrows and double-headed arrows connecting text in three different fonts. The significance of the different kinds of linkages and fonts is nowhere explained.

Bhaskar's unchecked impulse to classify and offer diagrams is symptomatic of a problematic conception of philosophy. Philosophy, for Bhaskar, never really goes deeper than providing a taxonomy.

ÁñÁ«ÊÓƵ

A key idea is what Bhaskar calls "referential detachment'. This is meant to ground his defence of a reality independent of the conceptual structures of language and thought. Bhaskar claims that we can find a place for ontology - an account of the world - by referential detachment, the "detachment of the act by which we refer to something from that to which it refers". But it is a platitude to note that the idea of reference involves the idea of the thing referred to and that this is separable from the act of reference. Platitudes about reference may support a minimal form of realism; they do not support the variety that Bhaskar needs in order to ground a transcendental ontology.

Unfortunately, it is impossible to tell whether Bhaskar's " referential detachment" is more than platitudinous. The concept appears on a number of occasions, but it receives no sustained elucidation. Given the voluminous work on realism in recent decades this is simply not good enough.

Much of the relevant work on realism comes from Donald Davidson. Bhaskar calls Davidson a verificationist. This has been said before, but usually on the basis of an argument. On another reading, Davidson's denial of a scheme/content distinction means that our ontology consists in an account of the world as present to us in our thought and talk. This is a kind of realism. It may even be enough to save science from the loose talk of Kuhnian relativisms. Unfortunately, such possibilities drop through the swiftly sketched taxonomy that Bhaskar offers.

Bhaskar does not argue with philosophers, he mentions them. It is a scandal for Verso to claim that amongst other things, the book discusses Heidegger's "scandal of philosophy". It does not. There is mention of it and, in a short paragraph on page 23, the problem of the existence of the external world along with the problem of induction is solved by appeal to the concept of ontological stratification. Given the absence of a general defence of what Bhaskar means by ontology this is worthless.

ÁñÁ«ÊÓƵ

Philosophy books that cover a broad front only work when they provide the scaffolding of argumentation and theory on which the uninitiated can exercise their intellectual muscles. Swinging through the hoops of Bhaskar's diagrammatic taxonomies is simply disorientating.

Michael Luntley is a senior lecturer in philosophy, University of Warwick.