Learning How To Do Things With Words
Before I begin, we require a necrology in order to set the background. It is for LAD, Chomsky's (1962) Language Acquisition Device, a veritable child prodigy that, for its 10 years of sway, helped produce a new way of seeing what is involved in acquiring language - and thereby dug its own grave. LAD, let us recall, was what linguists refer to as a discovery procedure, that is to say, a means of discovering the rules by which acceptable sentences in a language are put together. Its input was a sample of the language, however encountered; its output was the set of syntactical rules that would generate all the well-formed sentences possible in the language and none that were illformed. The base of this recognition or discovery program was presumed to be the language-learner's innate grasp of the universals oflanguage. The local language being learned, according to this view, was merely a realization in local form of the syntactic universals of language. The innate grasp of these linguistic universals of language was assumed to be independent of any knowledge of the nonlinguistic world. Nor, indeed, did the recognition program require anything more than that the learner (or discoverer) of the language be a bystander: the spoken corpus of speech flowed round and into him, and the rules came out the other end. It did not require, for example, that he already know what the language referred to - that he have concepts about the real world being referred to - nor that the learner had to enter into particular kinds of dialogue with the speakers of the language. As an enthusiastic David McNeill (1970) put it, "The facts of language acquisition could not be as they are unless the concept of a sentence is available to children at the start of their learning [po 2]."