This festschrift for Michael Dummett consists of 11 essays by contemporary professional philosophers. Exactly who they are the reader is presumably expected to know from their names, since the editor, who clearly does not believe in providing either the usual index or the usual bibliography, is no less parsimonious in providing information about the contributors. The topics they have chosen all relate to areas in which Dummett has specialised, notably philosophy of language and philosophy of mathematics. Some contributors are evidently out to give Dummett a rough ride. One thing this demonstrates, albeit unwittingly, is that none of them can match the ponderous panache of Dummett's own writing on his favourite themes.

The old chestnut about whether language is prior to thought or vice-versa is re-roasted by Christopher Peacocke until charred to a cinder. Alexander George struggles tortuously to convict Dummett of misrepresenting Gottlob Frege's view of the communicability of thoughts. Sanford Shieh queries Dummett's rejection of Quine's holism. John McDowell takes up another "topic that Michael Dummett thinks he has disposed of"; in this instance, so-called "modest" approaches to semantics. Jason Stanley discusses the semantics of "rigid" designators. John Campbell reconsiders Dummett's view of the reality of the past. Barry Taylor brings up the same issue in connection with Dummett's defence of McTaggart's "proof" of the unreality of time. The remaining four papers, by Crispin Wright, George Boolos, Charles Parsons and the editor, all deal with exegesis of points in Frege's mathematics, and in particular with issues connected with Hume's Principle.

Frege, as might be expected, is the name that crops up more frequently than any other in these pages, apart from Dummett's own. It is an interesting question whether Dummett's obsessive preoccupation with the work of Frege is a reflection of the interest of a generation of philosophers or, in part, a cause of it. Perhaps both. Certainly it seems to relate to the intellectual blinkers that many philosophers nowadays wear when approaching questions of meaning. This often alienates those temperamentally inclined to wider and more humane perspectives. What we say, what we write and what we think are three important components of our mental lives, although philosophers commonly conflate the first two. What is far from clear is why one should turn to Frege, of all people, for the key to enlightenment in such matters. It is rather like seeking guidance from a lock-keeper about the role of waterways in an integrated transport system.

Mathematics, contrary to what one might expect from the importance accorded to it in the western intellectual tradition, plays a minor or even negligible part in many cultures. And even where it is rather central (as in cultures based on reverence for calendars and time-keeping) the "philosophy of mathematics", if any, is often simple in the extreme. More importantly, mathematics is seldom if ever treated as the paradigm of thinking or of language. In order even to begin to understand the role of Frege in modern philosophy, one has to realise how desperate by the 20th century European philosophers had become to establish an academic corner in studies of language and mind that was safe from the empirical encroachments of linguistics and psychology.

̽»¨ÊÓƵ

Along came mathematician Frege - a man, as Dummett says, "not celebrated in his day" - looking for unassailable foundations for mathematical discourse. For this purpose he put forward what - according to taste - is an extremely profound or extremely naive theory of meaning. Professional philosophers like(d) it because Frege's account was locked into the philosophical distinction between "true" and "false". Hence also locked into the assumptions underlying Aristotelian logic. This was retrospectively seized upon by Frege's successors as providing a philosophical basis for the semantics of language in general. But in all probability Frege would have been no less surprised to find himself hailed posthumously as the founder of modern philosophy of language than Sir William Jones would have been to find himself credited with the foundation of comparative philology. In both cases a later generation has rewritten history to suit itself.

Dummett's Frege is a figure reconstructed explicitly to play this historical role. The subtext for much of this book is the question of how far Dummett's Frege is credible. Frege's great merit, in Dummett's eyes, is not so much to have tried to establish arithmetic as a continuation of logic, but to have done this on the basis of a semantics that rejects "psychologism". Frege's initial "insight", according to Dummett, was that "a sentence is the smallest linguistic complex which one can use to say anything: hence the meaning of a word is to be given in terms of the contribution it makes to determining what may be said by means of a sentence containing it."

̽»¨ÊÓƵ

The snag is that, for many of us, Dummett's claim renders Frege's qualifications as a semantician problematic, to say the least. For while this view of the sentence might conceivably do for a restricted notational code in which the simplest "sentences" are of the form "2 + 2 = 4", when generalised to languages like English or French it goes manifestly awry. And no amount of quibbling about what it is to "say" something will put it straight. Frege's "insight", in fact, is just a resuscitation of one of the stalest platitudes of traditional grammar, according to which a sentence is, or should be, the "expression of a complete thought". But this notion, patently, was nothing if not "psychologistic". Frege rescued it for his purposes by drawing an obscure and contentious metaphysical distinction between "ideas" and "thoughts", the latter being sempiternal items which somehow exist independently of individual thinkers. But this manoeuvre in turn brought trouble in its wake. Since "thoughts" are presumably no use to anyone unless they can be "grasped", it seems that in the end Frege hoisted himself with his own petard. For any serious account of "grasping" a thought will readmit by the back door most of the "psychologism" that had been so ostentatiously kicked out by the front door. (Unless one is prepared to adopt a strictly behaviourist theory of how language works.) Whether Frege can be posthumously rescued from his own semantic muddles is one of those questions that could only become important as the result of generations of philosophical in-breeding. Reading arguments about such issues as whether Frege believed thoughts were communicable makes one wonder why Frege bothered to publish anything at all. He could, after all, have remained silent like Cratylus and just wagged his finger. Somehow or other, Dummett and his disciples seem to have got themselves stuck in a time warp where the Fregean mathematical cart leads the linguistic horse. This will not do, if only for the obvious reason that without language there would be no mathematics. But "there is no absurdity too great for some philosophers to espouse it" - a view Dummett himself holds, as one contributor reminds us. There seems to be a moral here about those who live in glass houses being careful when it comes to throwing stones.

Roy Harris is editor, Language and Communication.