In Language and Silence and In Bluebeard's Castle Dr Steiner spoke of the 'retreat from the word', the distrust of language that has made us increasingly ill at ease in the interpreted world. This is still a major theme of [Extraterritorial], which takes Borges, Nabokov and Beckett as representative men, unhouseled and unhoused, warming themselves by rubbing two languages together. Their 'extraterritoriality', however, has compensatory graces: for the loss of confidence in language, the inability to feel truly 'at home' in it, is in substantial measure a reluctance to take it for granted, a painful awareness of its importance, which can lead to elegant, self-conscious artifice as well as to silence. Indeed, the retreat from the word has been rather noisy; and now, with much rattling of sabres, the forces are being regrouped under the banner of linguistics for a new campaign. That, roughly is the picture. There occurred in the first part of the century a crisis of language which led to a reassessment of our relation to it. The change in our awareness of language has accorded it so central a role that man can now be defined as 'the language animal'….

[Concern] with language in its various guises has never been so wide-spread or fashionable. In literary criticism the 'ultimate meaning' of a work is likely to be what it implies about the relation of the subject to his language…. Most disciplines will soon be reorganised as studies of 'languages' which transmit 'information': 'I have no doubt that theories of coding and of fields will soon be prevalent in the study of art, of music, of social institutions. Already the biological disciplines, linguistics and anthropology are working in close mutual awareness and with an often shared vocabulary. This is a revolution of perspective that concerns us all.'

It does indeed, and one object of concern might be the treacherous lure of 'shared vocabulary'. If, as Dr Steiner tells us, we have indeed 'lost confidence in the act of communication itself', then we might at least be wary of transferring terms or arguments from one context to another. The dangers are particularly evident in his essay on Chomsky, where haste to pose large questions about the nature of man leads to substantial misinterpretation of the theory and misuse of its key terms. Dr Steiner speaks, for example, of 'innate deep structures' and argues that 'a postulate of deep-structured linguistic universality must entail a reasonable procedure for translation between different languages.' Here he is confusing the deep structures of sentences of a language—the underlying logical relations which give 'John bought the car' and 'The car was bought by John' the same meaning—with the innate organisation of the mind which enables a child to acquire the complex grammar of a natural language when presented with relatively little linguistic data…. To suggest that Chomsky 'would regard as "merely of the surface" questions which seem to me primary and ontological' is perhaps to fall foul of the connotations of 'deep' and 'surface' and elevate misunderstanding to fundamental disagreement. From a phenomenological point of view the differences between languages are certainly more important than their similarities, but this does not mean that the differences are linguistically 'deeper'.

One must, however, admire Dr Steiner's honesty in printing Chomsky's comments on his essay as footnotes. These will dispel misunderstanding and enable one to concentrate on the major question: what are the advantages of adopting a linguistic perspective in other disciplines? What benefits does the revolution promise?

In biology and genetics, information theory displaces a mechanical causality and introduces the concept of a 'very generalised form of learning into the fundamental mechanics of evolution'. Dr Steiner's argument is promising but as yet rather sketchy.

In the sphere of literary criticism, where one might expect him to speak with some authority, his claims are surprisingly modest. Though he cites with approval the work of the Russian Formalists, the Prague Aestheticians and a few contemporary critics, he admits that this deployment of linguistics 'has not really contributed all that much to our reading of a poem' and that the complexity of literary effects is such as to defy formal linguistic explanation. The principal benefit is of another kind: we are now more aware of the 'orders of difficulty which are involved' and can pose in more precise terms questions about the interaction of metric, phonological and syntactic patterns. But one suspects that for Dr Steiner the real pleasure of the revolution lies in the larger questions about literary language with which he can challenge linguists and critics: 'Are certain languages more apt to literature than others?' Do great languages 'run down' to the point where they inhibit creation? How will changes in our awareness of language affect literature?

The implied answers are pessimistic—Dr Steiner has not abandoned his customary role. But against his strictures on contemporary culture and its threat to literature one might set his account of the other forms in which the best imaginative work of our time finds expression: from the inviolate, austerely frivolous structures of Borges's 'fictions' and Nabokov's Pale Fire, through Lévi-Strauss's Mythologiques, to the interaction between imagination and exact sciences which will reorganise our intellectual world. Whether or not, as he suggests, Needham's Science and Civilisation in China is the contemporary equivalent of Proust, it does not seem true that because of the decline of literature the vivacity of contemporary intellectual life is 'spurious'. Neither Dr Steiner's Spenglerian vision nor his occasional misunderstandings should blind us to the importance of the works he discusses and the questions he asks. Caution, after all, is a pale virtue, and a man's reach should exceed his grasp.