BOOK REVIEW / Taking the linguistic turn: 'Origins of Analytical Philosophy' - Michael Dummett; Duckworth, 25 pounds

MUCH HAS been made in recent years - especially in the light of last year's lamentable 'Derrida Affair' at Cambridge - of the gulf between 'Anglo-American' and 'Continental' philosophy. Some zealots on this side of the channel would have us believe that this gulf is unbridgeable. But why? What, exactly, is supposed to be the irreconcilable difference between the two traditions? And what do the terms 'Anglo-American' and 'Continental' mean in this context?

One fairly common answer (in England and America, anyway) is that 'Anglo-American' means 'analytical', while 'Continental' means 'phenomenological', the former being a tradition of hard-headed, rigorously logical thought based on the work of Frege, Russell and Wittgenstein, the latter a tradition of woolly- minded nonsense founded by Husserl and Heidegger. But, as one can see, this notion has only to be stated for its central absurdity to be clearly manifest. For two of the acknowledged founders of the analytical tradition, Frege and Wittgenstein, were quite emphatically not English or American. The analytical tradition, it appears, is Continental.

Such reflections lie at the heart of this entirely admirable new book by Michael Dummett, the recently retired Wykeham Professor of Logic at Oxford. Attention to the history of analytical philosophy, would, Dummett thinks, 'contribute to closing the absurd gulf . . . between 'Anglo-

American' and 'Continental' philosophy by showing that, at bottom, the origins of analytical philosophy are the same as those of the phenomenological school.' Frege and Husserl, the founders of the two traditions, were contemporaries of each other, working on the same problems and influenced by the same body of work. In a brilliantly apt analogy, Dummett compares them to the Rhine and the Danube, 'which rise quite close to one another and for a time pursue roughly parallel courses, only to diverge in utterly different directions and flow into different seas'.

The book has an appropriately Continental provenance. It began as a series of lectures at the University of Bologna in 1987 and was first published (in English) in the Italian journal, Lingua e Stile. Since then, it has been published as a book in both German and Italian. This new version differs substantially from those already in print, especially in its treatment of Husserl, which is both more extensive and more sophisticated than in the original lectures. This is partly in response to Barry Smith's review of those lectures, and partly also in response to David Bell's superb study of Husserl, published by Routledge in 1990, in which Bell presents an ingenious, stimulating and persuasive defence of Husserl against both Frege and Dummett - a defence which Dummett only partially answers in this book.

Dummett's aim in tracing the analytical and the phenomenological traditions back to their common roots is not only to show what they have in common, but also - by showing how Frege's work improves Husserl's - to illustrate what makes the analytical tradition distinctive from and superior to its rival.

According to Dummett, the crucial step taken by Frege but not by Husserl is the so-called 'Linguistic Turn'. In its broadest sense, this phrase is used to describe the development of philosophy in this century away from questions of knowledge and towards questions of meaning. Dummett - as he expounds in his book on Frege's philosophy of mathematics published last year - has a startlingly specific answer to the question of when and by whom the Linguistic Turn was taken. It was, he believes, made by Frege in paragraph 62 of Foundations of Arithmetic. There Frege asks how numbers are given to us and then immediately substitutes for this question another one about the meaning of sentences containing words for numbers. In that substitution, Dummett claims, the Linguistic Turn, the decisive step of philosophy this century, was taken.

It is possible, I think, that Dummett's emphasis on the 'Linguistic Turn' - and, indeed, on Frege's work generally - in defining and shaping the analytical tradition leads him to present a rather partial and distorted account of what analytical philosophy is. 'What distinguishes analytical philosophy', he writes unequivocally, 'is the belief, first, that a philosophical account of thought can be attained through a philosophical account of language, and, secondly, that a comprehensible account can only be so attained'. Elsewhere, he has expressed much the same idea by saying that the two 'axioms' of analytical philosophy, are one, that the task of philosophy is to analyse thought, and two, that the only effective way of analysing thought is to analyse language.

That this, as it stands, can't be quite right is shown, I think, by applying it to two widely different thinkers: Jacques Lacan and Bertrand Russell. Lacan has argued explicitly for both Dummett's 'axioms', and therefore - if Dummett's definition is right - ought to be regarded as an analytical philosopher. Russell, however, who, at various times, argued against both 'axioms' would fall, on this characterisation, outside the analytical tradition. And yet, while Bertrand Russell is universally acknowledged as a founder of analytical philosophy, Lacan is firmly established as the favourite bete noire of contemporary analytical philosophers.

With regard to Russell, Dummett excuses himself from discussing his role in shaping the analytical tradition on the perfectly reasonable grounds that his interest is in tracing the origins of the tradition back to its Continental roots. He might still, however, have acknowledged that Russell's conception of analytical philosophy (or 'the philosophy of logical analysis' as Russell calls it in The History of Western Philosophy) differs fundamentally from his own. For Russell, the task of philosophy was not to analyse thought, but to understand the world, and, for him, the analysis of language was only one among many ways of attempting that task.

The fact that Lacan satisfies Dummett's definition of an analytical philosopher points to a wider problem with Dummett's whole discussion of the two traditions. By concentrating on the 'Linguistic Turn' as the defining characterisfic that distinguishes analytical and phenomenological philosophy (and, by implication, 'Anglo-American' and 'Continental' philosophy) Dummett seems to leave no place for contemporary non-phenomenological Continental thought. For the work of, for example, Foucault and Derrida - not to mention Lacan - is explicitly a reaction against Husserlian phenomenology. And yet analytical philosophers notoriously find it entirely alien.

To overcome this sense of alienation, we need a historical map that shows not just the divergent paths taken by Fregean and Husserlian philosophy, but also the course taken by Foucault, Derrida et al away from their phenomenological origins. As many people over the last few years have argued, this course follows a strangely parallel route to that described by Dummett as defining analytical philosophy (it even has its own 'Linguistic Turn') which is why some have come to expect from the two traditions either side of the Channel not only a better understanding of each other, but also a fruitful interaction.