Dismissing Chomsky's political writing as ephemeral, albeit effective, journalism, Pieter Seuren begins an assessment of his work in linguistics:
He dominated linguistics for four full decades, from 1960 till 2000 and since 2000 his influence has still been reverberating in many ways... [But] it looks very much as if his influence is rapidly declining, and with good reason, because his theoretical approach has been shown to be flawed in too many ways by a large variety of critics. For all we know, his reputation may turn out, sub specie humanitatis, to have been a momentary flare.
And this, more or less, is the Chomsky I read during my undergraduate years in the mid-60s and had abandoned by the time I entered SUNY Buffalo in 1973 for my Ph. D. work:
Chomsky subsequently tried to reinterpret the notion of an algorithmic generative grammar in realist terms, that is, as a theory of how the human mind deals with language, and no longer in the purely instrumentalist terms current in the brand of structuralism that he had been taught and according to which all that counts is to provide a precise and concise statement of the facts of each language, regardless of how they are implemented in human brains. This move from instrumentalism to realism was brought about by his contacts with a small group of young Harvard psychologists, headed by Jerome Bruner and George Miller, who were in the process of creating the new cognitive science in opposition to behaviourism. The incorporation of the concept of a generative grammar into the new paradigm of cognitive science provided him with the opportunity of being part of a much more general and more profound revolution affecting the whole of society, the passing from ‘inhuman’ behaviourism to ‘human’ cognitive science and thus the reinstatement of human values in the modern world, giving impetus to a new and wide-ranging emancipatory movement.
This is the Chomsky of Aspects of the Theory of Syntax (1965).