Chomsky in retrospect — 3

by pieterseuren

When one takes a cool and dispassionate look at Chomsky’s linguistic work, disregarding the mythology around it, what one finds is a conspicuous lack of actual empirical or analytical work and a dogmatic, non-dynamic and non-ecological view of language. He never ever engaged in actual syntactic research, leaving it to his underlings to do so. His were only the ‘grand ideas’, which, however, turned out not to be so grand after all. His own linguistic observations were trite and repetitive in the extreme (“John is easy to please”), showing no real interest in language as a human phenomenon beyond the possibilities it affords for grandiose general vistas and empty promises. What is known as ‘Chomskyan grammar’, in its various guises over the years, is largely the product of his followers, who did the hard work—often with too much subservience to the master and too cavalier an attitude with regard to data and to general considerations that would contradict or nullify their analyses.

Meanwhile, it was claimed that the resulting theory of grammar was universally applicable to all languages with only a small number of parameters leaving a choice of alternative options, which would account for the diversity of languages. While this is in itself a laudable programme (for which the precept was already formulated by the French linguist Nicolas Beauzée in the 18th century), its full realization, given what we know today, requires a long-term research programme, spanning generations. Yet the Chomskyans thought it could be done overnight—a sad case of intellectual hubris. The final result has been a highly speculative, etiolated notion of grammar according to which sentences are formed by algorithmically combining morphemes into larger wholes following a small set of uniform formal procedures lightheartedly taken to be of a universal nature, but without any regard for what these randomly produced sentences mean. The absence of a principled link with meaning made this theory absurd in a realist, and perverse in an instrumentalist sense (see my Chomsky’s Minimalism, OUP, 2004, p. 70).

Chomsky is still often hailed as the inventor of transformational generative grammar. In actual fact, however, he used the ideas for transformational generative grammar developed by his teacher Zellig Harris, marketing them under his own label. This fact is well-known, though Chomsky, followed, as always, by his faithful, defends himself by saying that his notion of transformation differs essentially from that developed by Harris, referring to Harris’s earliest notion. But Harris did not stop there, nor did Chomsky stop at Harris’s notion he took over. The fact is, quite simply, that Chomsky took Harris’s notion, did some tinkering and sold it as his own.

He opposed behaviourism in all sorts of ways, both in the human sciences and in politics, applauding the advent of the new paradigm of cognitive science and the concomitant recognition of human values and rights, but he has been unable to accept the fact that the recognition of meaning as the central factor in language and language use is part of that new paradigm. He never could depart from the old behaviourist frame of mind, which saw meaning as a ‘mentalist’ phenomenon, outside the sphere of science, just like emotions or morality or beauty. For him, grammar has always been “autonomous and independent of meaning” (Syntactic Structures, Mouton, 1957, p. 17), which means that sentence structure is to be defined in terms of linguistic form only, in the shape of a generative algorithm producing sentences at random without any regard for their meaning. This view has stayed with him ever since 1957, but for a short intermezzo during the mid-1960s, when he, in Aspects of 1965, briefly went along with the young generative semanticists, who saw a grammar not as a generative but as a mediational algorithm, turning semantic representations (deep structures) into corresponding surface structures.

In general, his notion of human language is dogmatic, rigid, mechanical and extremely restricted. He only has an eye for the mechanics of monolithic speaker’s grammars. The obvious fact that language is used for communication between speakers/writers and listeners/readers does not interest him and is even denied by him or reduced to an unimportant epiphenomenon. Instead of giving an analysis of the concept of communication, incorporating the essential notion of socially committing speech act, he sees language as primarily an instrument for silent soliloquy. Moreover, the undeniable fact that each speaker’s linguistic competence covers a great deal of active and passive internal variation as regards grammar, lexicon and phonology, variants being marked for specific values on geographical, social and interactional parameters, has always been lost on him. For him, dialects and sociolects are simply different languages (“ideal systems”) with different monolithic grammars (see Chomsky, On Language, The New Press, New York, 2007, pp. 53–57), which shows him ignorant of the fact that every competent speaker of any language has at his or her command a vast number of internal variants of the same language. Chomsky’s dismissive comment in this regard was (ib. p. 54): “It has been suggested that the language system of an individual does not consist in the interaction of ideal systems, but in a single system with some margin of variation. If that is it, then it’s not very interesting.” Who made the suggestion that internal variation in competence is only marginal is not specified. In fact, what has been suggested by many for decades is that internal variation in linguistic competence is all-pervading at all levels, that competence is by definition internally variable, each variant being marked for its geographical, social and interactional value, which raises the question of how all those socially marked variants can be integrated into one single competence system. This daunting—and extremely interesting—question has so far not been answered satisfactorily.

Disappointingly but not unexpectedly, sociolinguistic variation within a single language is called into play only as a means of escaping from counterexamples, often, and often incorrectly, said to be “dialectal variants”, meaning that they do not have to be taken seriously. This attitude is hard to reconcile with what is said a few pages earlier in the same On Language (p. 48): “if the linguist is interested in the real nature of human beings—which is what I suppose—then he will seek to discover the system that is really utilized,” professing a realism that is belied by many other statements to the effect that grammars are not meant to describe the system used but only to characterize data—his old instrumentalism.

So lacking in insight into the social nature of natural language is Chomsky that he maintains in earnest that the primary purpose and raison d’être of natural language is silent soliloquy. This in the face of the well-known fact that no language allows for the expression of pure propositions and that all languages require a socially binding speech act operator over any proposition expressed for there to be a well-formed sentence. It must be said that even his staunchest followers tend to be reluctant to quote their master on this issue.

In a lecture entitled “Language and Freedom” given in Chicago in 1970 (Chomsky, For Reasons of State, New York, 1970, pp. 387–408), Chomsky admitted to being unable to connect the two topics of language and freedom in any other way than by an appeal to the fairly abstract notion that both freedom and language are innate—the only reason why language is ever mentioned in his political writings. A real linguist with his political leanings would have grabbed the chance to elaborate on language as the expression of social and cultural identity, on the widespread linguistic oppression in totalitarian states, the many struggles for official recognition of local languages, etc. But we see none of that. Nowhere in his incredibly numerous political writings, speeches, pamphlets, interviews, what have you, is there any mention of the role of language in the political life of nations and societies. In one who calls himself a linguist and appears at the same time to be so preoccupied with the well-being of people on earth, one would have expected some openness towards the social aspects of language. But no, not even the old Saussurean insight that a language constitutes social reality is ever mentioned in this man’s writings. So much for the linguist he claims, or is claimed, to be. … to be continued