Grammar and language in Syntactic Structures

In his review of Chomsky’s Syntactic Structures (1957), R.B. Lees praised the work as a pioneering effort in constructing a truly scientific theory of language. Very soon, Chomsky’s 1957 book was regarded as inaugurating the ‘generative revolution’ in linguistics. The present study focuses on the use made by Chomsky, in his 1957 book, of key concepts such as ‘structure’, ‘level’, ‘grammar’ and ‘language’, while relating his argumentation to theory formation in linguistics in the 1950s. It is shown that, from the point of view of theory construction, Syntactic Structures has an ambivalent status: the work goes beyond distributionalism in showing the acceptability of level mixing, and in extending syntactic analysis to the relationships between utterances, but at the time of addressing the issue as to what it is to ‘understand’ a sentence, the work relapses into American structuralist orthodoxy. Also, in Syntactic Structures, Chomsky reverts to the Bloomfieldian skeptical attitude concerning the use of meaning in linguistic description, and no reference is to be found here to the speaker’s ‘competence’. As a result, the book appears to be rooted in the structuralist tradition, reflecting the structuralist concern with rigorous methodology; its distinctive quality was to deal with linguistic methodology as constituting the core of a general theory of grammar. “Whether the change that actually took place—the advent of and eager reception of the approach called transformation theory—should be described as internal or external, as a revision and rehabilitation of Descriptive Linguistics, or as a displacement of it, is no simple one, for which reason I save it for another day” (Wells 1963: 48).