It is a matter of fact that behaviorism has influenced foreign language teaching.Whether or not mentalist attitudes will be equally influential is a matter for conjecture.While criticizing behavioristic practices，Chomsky has said that he cannot see any application of his theory of language in teaching and indeed does not see why anyone should think that it ought to have applications.In developing his views，little could be further from his mind than the interests of language teachers.A language teaching methodology is not something to be derived directly from a linguistic theory.

However，it is possible to accept this opinion while at the same time arguing that nonetheless language teaching ignores linguistic theory at its peril，and that a methodology that runs counter to what linguists believe to be the nature of language needs careful investigation.Attempts will be made therefore to discover the possible implications of the mentalist account of language acquisition.

If people take this point of view，it should not surprise us to discover that it provides us principally with arguments against certain practices rather than a comprehensive and constructive picture of the way things ought to be done.In the first place，there is the notion that the learning process is essentially uninfluenced by external factors.Carefully planned schedules of reinforcement are unnecessary，since learning will take place whether or not the individual is reinforced.There is no need for the amount of active participation in language production that the behaviorists and others require，since learning can take place without repetitious active responding.Learning involves the internal mastery of the rule，and making the response is not so much a step in the acquisition of the rule as a sign that the rule has or has not been learned.

The learning mechanism operates through its capacity to formulate rules about the language once the individual has been exposed to it.The essential condition is exposure to the language，and as long as this exposure continues，the learning mechanism will operate.What is needed in language teaching，therefore，is adequate exposure to the target language.The greater the exposure to meaningful language，the more effectively the learner can formulate and revise his hypotheses about the structure of the language.As can be seen that he induces rules from the data and then attempts to use these rulesin producing and interpreting more language.Since external conditions are irrelevant，the whole panoply of principles that have been enunciated for the selection and sequencing of language for teaching purposes are quite superfluous.The material of language teaching should be extensive samples of natural language，not language specially produced and limited to suit the apparent needs of learners.Up to a point there is implicit approval here of the literary tradition in language teaching，since this did expose the child to language produced by native speakers for consumption by native speakers.The linguistic range that the child would meet would be enormous.Its weakness lies in its incompleteness as a sample of language use.Much that a child learning his mother-tongue would encounter is not to be found in the literature.

All of this casts doubt on the careful linguistic control that is characteristic of modern teaching materials.It also raises the question of the validity of language learning through drilling.Drilling is a technique for maximizing active language production by every pupil in a way that allows him to be reinforced immediately.But it is hard to see the value of this if there is doubt about the notions‘active response’，‘repetition’and‘reinforcement’.If the pupil is not to learn the structure of the language by hearing and constructing large numbers of analogous sentences，how is he to do it？The temptation is clear to fall back on the traditional procedure of expounding a rule and then testing the learning of that rule through exercises.Foreign language learning would return to the realm of problem-solving subjects.The word‘rule’should have to serve to refer both to，the small child's hypotheses about the structure of his language and to the generalizations that the language teacher has often found it desirable to make to his pupils and elicit from them.In the former case，it refers to a totally unconscious process；in the latter it is quite explicit.The author of the present study knows of no clear evidence that the best way to develop unconscious mastery of the rules of a language is through the initial conscious learning of those rules.It is the informally reported experience of many learners that rules of language learned in this way are rarely fully assimilated and continue to interfere with the spontaneous use of language.Linguistic theory cannot possibly help here.Even if one believes that every language learner has to acquire the rules of the target language，the question of how they are best acquired is still left open.

Another feature of the behaviorist approach，and again this view that is shared by many who would not subscribe to behaviorist theory in detail，is the insistence on the advantages of ensuring that pupil's utterances are always correctly formed.If what they say is correct，what they learn will be correct.In contrast to this，it must be now considered that the possibility that a correct response does not necessarily indicate that the person is learning，and that an incorrect response may be a satisfactory sign that learning is proceeding normally.A pupil might be able to construct English or French forms correctly under tightly controlled conditions where what is required may be little more than imitation，but as soon as he is put in a freer situation，people may discover that he has not in fact mastered the rule.Alternatively there are some kinds of mistake resulting from overgeneralization，as in the example of the child learning English as a mother-tongue，which could be considered to show a reasonable sensitivity to the rule-structure of the language.The correction of such errors may assist the learner to accurate use of the rule more rapidly than a system which expects him to repeat only correct forms.Whichever approach was adopted，one could be satisfied that a rule had been fully mastered only when the pupil was using it correctly under free conditions.Practice in the choice of forms is an essential stage in language learning and cannot take place under fully controlled conditions.

There is little to be said from the psychological point of view about how a child acquires the meaning of a language.The difference from the behaviorists lies not only in a rejection of the stimulus-response analysis，but in the importance that is attached to meaning.Not only is meaning the whole point of language；the structure of a language itself cannot be properly learned unless it is fully meaningful.Purely mechanical drilling，such as it is possible to find in many courses，would be rejected because it would not even be an effective way of learning the structural elements of the language.The learning of a foreign language should therefore be a meaningful activity throughout.There seem to be no implications as to how meaning should be taught，only the conclusion that it is a most important aspect of language teaching.