||The study of transformational grammar is a linguistic enterprise which revolutionized the study of syntax when it first appeared in the 1950s. Noam Chomsky proposed that grammar is an independent system, quite separate from other components of language such as phonology or semantics. The grammatical component was held to comprise various sub-components, each interrelated and each characterized by the operation of rule systems which conjoined to determine those sentences which were permissible, and those sentences which were not allowed in the language (ungrammatical).
A central aim in transformational grammar was to capture the insight that certain sentence types are clearly related to one another. Consider the case of active and passive sentences:
Â a. Matthew drank a lot of wine. (active)Â b. A lot of wine was drunk by Matthew. (passive)Theoretically, it would be desirable to acknowledge this relationship within the grammar, rather than leaving it as an arbitrary, ungoverned linguistic artefact. This aim can be achieved by first recognizing that the sentences we hear or read are merely the end product of more abstract, underlying syntactic processes. In transformational grammar, each sentence is first characterized as an abstract deep structure, which is then passed on to the transformational component. Among other rule systems, the transformational component contains rules which deal with related sentence types. The relationship between active and passive sentences can therefore be explained in terms of their originating from a single, underlying, deep structure.
In effect, a transformational rule converts one syntactic form into another, and consequently there must be an input to the rule (known technically as a structural analysis) which contrasts with a (transformed) output (known as a structural change). Transformational rules include such operations as the movement, deletion, copying and adjoining of elements within the sentence as appropriate. Once all the relevant transformational rules have applied, the result is the so-called syntactic surface structure, which still requires the operation of the phonological component to realize the sentence as we know it in its written or spoken form. In the most recent version of Chomsky\'s syntactic theory, the concepts of deep and surface structure have been redefined and rechristened D- and S-structure respectively.
Throughout the 1960s, there was an explosion of research in the field of transformational grammar. Eventually, it became apparent that there was, in fact, an unhealthy proliferation of transformational rules, which raised the question of what the fundamental rationale for certain rules should be, especially if they were to retain any truly explanatory (as opposed to merely descriptive) power. There was, too, an excessive concentration on the vagaries of a single language, English, rather than on the stated ambition to discover and explain aspects of syntax common to language per se. Currently, the concept of transformation is still an axiomatic feature of the Chomskyan approach to syntax, but it is now much more heavily constrained and also more relevant to the concerns of universal grammar. MS
Further reading A. Radford, Transformational Syntax.