They have asserted that behaviorist explanations of language acquisition cannot account for it. According to the behaviorists, to learn language is to learn a sequence of stimulus-response links. The child's internalized "rules" (the sneer quotes are the behaviorist's, who does not deign to use such language) are similar to the "rules" involved in motor sequences like brushing one's teeth and tying shoe laces, or in any other well-learned motor activity. Against this, Chomsky and his followers have argued that the child cannot be seriously maintained to have learned a different set of stimulus-response links for each utterance he makes (Chomsky, 1965). Life is too short for learning all the word strings we use.
According to the semantic approach the child learns how different meanings are expressed by different sentence structures ( Quine, 1972). One might have expected such an approach to be formulated very soon as a reaction against behaviorist explanations, with their complete neglect of meaning. But such was the stranglehold of behaviorism on theory construction that the semantic approach was not formulated for a long time. The behaviorist edifice succumbed only to the truculent attacks of Noam Chomsky. Chomsky's linguistic theory, transformational grammar, gave rise to an alternative approach to language (Chomsky, 1986).
Chomsky as a b
Chomsky as a behaviorist conceptualizes discrimination learning in language Discrimination learning ensues when adult use of a word conflicts with that of the child. The process will be somewhat as follows (Baker, and McCarthy, 1981): (1) the child encounters something that reminds him of a paired referent, whether because it resembles it or because it was previously experienced in contiguity with it; (2) the adult uses for this new instance a word which differs from that learned for the paired referent; and subsequently (3) the child notices certain salient attributes in which the new instance differs from the paired referent. For instance, (1) the child sees a horse that reminds him of the referent of the previously learned word "doggie"; (2) the adult calls it "horse"; and (3) the child notices that the horse, unlike doggie, has a mane. The latter property may henceforward operate as a discriminating cue: It will be a NEGATIVE CUE for the word "doggie", and a POSITIVE CUE for the word "horse".
To forestall a possible misunderstanding, I want to point out that this earlier discussion is intended to explain how the child delimits the use of words, and not how he acquires distinctions between things. That is, the previously discussed process is not claimed to lead to his distinguishing between, for example, dogs and horses. On the contrary, the ability to make such a distinction--on the basis of differentiating properties, such as the horse's mane--is presupposed here (for, otherwise, how could he ever find out when to use "doggie" and when to use "horse"). The child may become aware of the difference between a horse and a dog--or between two different dogs, for that matter--without adult prompting. The issue here, however, is the child's use of words: To learn the correct use of a word it is not sufficient just to perceive differences between referents, but the child must also observe how these differences correlate with the applicability and nonapplicability of the word ( Wexler & Culicover, 1980). The child is innately not acquiring the correct grammar;