Missing the writing on the wall

To write a whole book about linguistic signs without ever discussing the profound semiological differences between speech and writing is nowadays a remarkable feat. This is what Rudi Keller has succeeded in doing. The achievement, unfortunately, bears depressing witness to the severe myopia concerning literacy that still prevails in some areas of modern linguistics.

The result is not so much a theory of linguistic signs as parts of several different theories, from Plato to Peirce, cobbled unhappily together. The explanation seems to be that Keller is at heart an old-fashioned historical linguist who has rashly ventured into the semiotic jungle and lost his bearings. So he looks for "theoretical" arguments that can be used to re-identify the familiar landmarks of 19th-century language studies, including the belief that linguistic communication is telementational and that languages are not only essential for rationality but are rationally used by their users for telementational ends.

This redemptive enterprise, in the absence of any language-independent account of rationality, involves circularities and inconsistencies at every turn, especially noticeable in Keller's question-begging treatment of "meaning". His basic strategy seems to be that semantic problems can be resolved by stipulative fiat. He declares "I am here" to be an "analytically true" sentence, but gives no account of analytic truth or of sentences. He "decides" to call meaning "that which makes the interpretation of signs possible" without realising that this is just a roundabout way of stating the platitude that treating something as meaningful involves finding meaning in it.

For Keller, linguistic signs "have meaning" even in the absence of interpreters: that is, when not in use. What kind of unused existence they enjoy he does not divulge. Nor does he ever distinguish between "having meaning" and "having a meaning". Into this semantic pot-pourri he throws the thesis that what makes a linguistic sign interpretable is "its rule of use in the language".

Hence from the Scylla of meaning to the Charybdis of rules. Keller wants to claim both that the meaning of a word is nothing other than its rule of use (a proposition he evidently believes to be Wittgensteinian) and that it is misleading to suppose that there are rules of use that both you and I know, because rules do not exist "outside and independently of those who follow them". According to Keller we cannot sensibly ask the question: "Why is an expression used as it is?" because the only available answer is: "That is just the way it is." Hardly a satisfactory basis for any kind of semantics or any theory of rules.

Between meaning and successful communication, in Keller's view, lies much complicated reasoning based on selected non-linguistic information. Just how we pick out the right information and the right inferences would be a miracle, were we not rational creatures. But since rationality is equated with language-using, this closes another circle. Which semiological criteria distinguish a linguistic sign from any other rule-based symbol, or even verbal from non-verbal communication, remain mysteries that Keller's theoretical investigation never gets round to elucidating.

Roy Harris is editor, Language and Communication.