Chomskyans chase down a blind alley

It is futile to try to reduce language to a set of scientific principles, says Michael Bulley. Was it Noam Chomsky's politics or his linguistics that got him voted the world's top intellectual last year? This year is an appropriate time to evaluate the linguistic Chomsky; it is the 50th anniversary of Syntactic Structures , the work that launched him as the leading light of a new way of looking at language.

That work had great influence, causing many people to devote their careers to pursuing its line of thought. While I admire Chomsky's academic seriousness and am sympathetic to his political opinions, I think his view of language is not only wrong but harmful. Misunderstanding the nature of language can also give you a false view of humanity. Here, then, is a criticism, for the non-specialist, of Chomsky's view of language and so, I suppose, of his view of humans.

When we speak or write, it seems we try to think of the right words and then produce them. About 50 years ago Chomsky, followed by others, decided it was not that simple. They called perceptible words "surface structure" and developed concepts such as deep structure and transformations, on the supposition that there were genetically based linguistic forms. A central concept was "Universal Grammar", existing in newborn children, enabling them to become language users. A huge body of writing was produced of procedures to relate any sentence to general patterns. The purpose was to account for language scientifically.

After a while, Chomsky and others decided the transformational-generative project was too big. They seemed not to have considered that it had not helped to explain how any real sentence had expressed its meaning. The theory needed to be put differently. The levels and rules between the principles of language itself and the surface structures of different languages could mostly be dispensed with as they could not be formulated completely. This new theory has the appropriate title of the "minimalist project". Where is it going? Will it end up saying that, when we speak, we just think of the words and say them?

In his latest writings, Chomsky has speculated on the origins of language, imagining it beginning with some genetic quirk, though he now says genetics cannot explain everything. There seems little connection between these new ideas and real words. Chomsky's earlier writing was based on made-up sentences, nearly always in English. Those examples were not only stilted but some, offered as normal, went against conventions. People in the movement seemed untroubled, having the attraction of accounting for language by rules.

The most telling argument against the Chomskyan scientific approach involves meaning. Few areas of linguistics study words only as objects. Science studies only real things, but the only reality of the meanings of sentences lies in electrochemical events in our brains. No picture, film or objective description of those events could be related to our experience of understanding sentences. Academically usable analyses of the meanings of sentences are bound to be crude and will anyway refer to concepts - things that do not exist.

The Chomskyans, asserting that sentences contained their meanings, tried to objectify concepts. This resulted in an attempt at "generative semantics". I agree with Wittgenstein that sentences do not contain meanings but produce them. Take a written sentence: each time it is read, it means something different, since we are all different and everyone is always changing. Of course, if the sentence refers to something of practical importance, you hope there will be enough similarity among people's understanding. Since meaningful language depends on reactions that cannot be objectively related to its sounds or signs, a scientific approach to grammar seems wrong.

The main Chomskyan error is the belief that language is something in itself, encompassing all linguistic possibilities. As for its genetic basis, our genes, certainly, made language possible, but what humans have done with that possibility has been up to them. The Chomskyan view has language as a system with permutations. In it, grammar is a biological reality, rather than a set of abstract terms. For me the word language is a shorthand for certain conventional activities for which we should not seek principles. The Chomskyan aim of discovering grammar is futile as, like language, it does not exist.

The notion of grammar in our genes has encouraged a neutral view of language and a fatalistic view of humans. While linguistics should be done objectively, the wrong conclusion that language is a biological phenomenon has caused some people to claim that judgments are irrelevant to it. I say, however, that to care about language is not to be romantic or arrogant but to care about something we have created. Its future health is up to us.

Michael Bulley does freelance language work in France and writes on language and philosophy.