ABSTRACT
Generative grammar This article is about the body of work which owes its original inspiration to the insights of Noam Chomsky in the mid-1950s and has been continually revivified by his insight up to the present. It has become one of the most influential syntactic theories of the twentieth century and, although by no means all practising linguists adhere to its principles and results, none can ignore them. Since its inception there have been huge developments in the theory and reactions to it have often been violent. In the mid-1960s work on the developing theory of ‘transformational generative grammar’ (TG) was perhaps coherent enough for one to be able to talk of a school of ‘transformational’ linguistics. This has not been possible for many years. Many who grew up within the model have gone on to develop theories of their own, often in reaction to the current work of Chomsky, and even among those who would describe themselves as generative linguists there is considerable divergence. That having been said, many linguists adhere to some version of a grammar that owes its intellectual genesis to one or other of the continually developing models offered by Chomsky. This entry is organised into four sections, based loosely around some of his more influential publications: Syntactic Structures (1957); ‘Standard Theory’, developing from Aspects of the Theory of Syntax (1965); ‘Principles and Parameters’, the theory developing out of Lectures on Government and Binding (1981) and Barriers (1986a); and some of Chomsky’s most recent ideas, stimulated by The Minimalist
Syntactic Structures
When Syntactic Structures was published in 1957, the position it took on the nature of linguistic activity was sufficiently at odds with that of the prevailing orthodoxy that it was appropriate to refer to it as revolutionary. The first chapter declared that grammar was an autonomous system, independent of the study of the use of language in situations, and of semantics, and furthermore that it should be formalised as a system of rules that generates an infinite set of sentences. This approach contrasted sharply with the
(then) fashionable orthodoxy that believed that the application of appropriate procedures to a corpus of data would yield a grammatical description. Chomsky rejected the use of a corpus, proposing instead that the empirical adequacy of a grammar should not be judged by whether it accounted for some finite body of observable data but by whether it could generate an infinite number of grammatical sentences and in doing so account for certain types of intuitive judgements that native speakers have about their language. Among these judgements are grammaticality judgements: that is, that a string of words, particularly a novel string, is or is not a well-formed sentence; that certain sentences are ambiguous, i.e. that a single sentence can have more than one interpretation; that distinct sentences can paraphrase each other, i.e. that distinct sentences can, in particular respects, have identical interpretations; that certain sentence types (affirmative and negative, declarative and interrogative, etc.) can be systematically
of this kind, it is claimed, constitute what speakers know about their language, and in addition to being able to generate all the grammatical sentences of the language a grammar should also account for this knowledge. It was mentioned above that Chomsky pro-
posed that grammar should be considered as an autonomous system, independent of semantic or phonological systems, though, of course, bearing a relation to them. Furthermore, he proposed that the syntax itself should consist of a number of distinct but related levels, each of which is characterised by distinct rule types and bears a particular part of the descriptive burden. We shall look briefly at the two most important components in a syntactic structures model: the phrase-structure (PS) component and the transformational component. The PS component consists of a set of PS rules
which formalise some of the traditional insights of constituent-structure analysis. Consider, for example, the following set of rules, adapted from Chomsky (1957: 26 and 111; items in curly brackets, { }, are alternatives, e.g., number is either sing[ular] or pl[ural]).
