5.2.1　Behaviorism and language acquisition

（1）Learning as acquired behavior

Learning is not easy to define formally because there are many different perspectives，each emphasizing a different facet of this complex process.A definition of learning could refer simply to the overt behavior.For instance，the fact that someone successfully drives an automobile suggests that the person has learned to drive.A definition of learning also could refer to an internal state of knowledge.Many dictionaries define learning in this fashion as“knowledge acquired by study.”From an evolutionary perspective，the idea of learning as overt behavior is important.Genes build an organism that has many characteristics.Some are morphological traits，such aseye color or limb structure，and some are internal neurological mechanisms that control，or provide a capacity for，behavior.Natural selection operates on both types of traits.Thus，animals are at a selective advantage if they possess certain beneficial morphological characteristics or if they possess certain neurological structures that allow for adaptive behavior.Stated differently，having morphological traits that are advantageous，or behaving in a manner that is advantageous for survival，puts the individual at a selective advantage.Over the many millions of years of biological evolution，such animals were more likely，on the average，to rear their offspring successfully and thus to perpetuate their genes.The point is that the behavioral manifestation of learning is important for survival and adaptation.What the animal does is of the utmost importance because，from the point of view of evolution，it is the animal's behavior，not its knowledge，that constitutes its coping strategy.

（2）Learning as knowledge

Learning may also be viewed as a transition from the state of ignorance to a state of knowledge.It is easy for humans to appreciate this point because we have been taught facts，values，and general knowledge our entire lives.Other creatures also possess knowledge，however.For example，animals know their territory；they know how to find food or water by using various landmarks；and they know where to expect to meet predators，even though they do not always behave in a manner that reveals such knowledge to a human observer.The challenge posed by animal learning research，then，is to know precisely what knowledge an animal has acquired，and the rules by which knowledge is obtained.

If learning means knowledge acquisition，then what do animals learn about？This is a difficult question to answer，even in the case of human beings，with whom we can communicate easily.We may，however，make an educated guess.First，an animal learns about stimuli in its environment because many stimuli serve as a signal for some important outcome.The sight of a distinctive tree or a certain path，for instance，may cause an animal to expect to“find”a water source nearby（the animal has learnt to associate the tree or path with the water source）.A robin searches for worms in a backyard after a summer rain because the water glistening on the blades of grass signals that worms may be near the surface.

Second，an animal learns about its own behavior.It learns that if it performs a given action，then a certain outcome will be forthcoming.A pet dog，for example，knows that it will receive food if it“shakes hands”with its owner from a vending machine.You recognize that the machine（the stimulus）contains cans of soda.The stimulus has significance or meaning in that respect.You also recognize that your own behavior isimportant in securing a soda.Putting coins in the slot leads to the delivery of the can.Simply knowing that soda is associated with，or predicted by，the machine does not，by itself，produce the soda.You must also have a separate bit of knowledge—namely，that performing the correct sequence of actions is necessary as well.

（3）Behaviorism

For behaviorists there is not a theory of language learning as such，but merely the application to language of general principles of learning.In general terms there is no difference between the way one learns a language and the way one learns to do anything else.Beyond recognizing that only human beings possess the capacity for language there is no need to postulate any complex internal endowment that allows us to learn and use language.The behaviorist is committed to admitting as evidence only that which he can observe，so that his data are the utterances that people make and the conditions under which they are made.In fact he would say that these limitations，which he feels to be necessary if he is to be properly scientific in his approach，do not prevent him arriving at an adequate description of language behavior.It follows from this that learning is controlled by the conditions under which it takes place and that，as long as individuals are subjected to the same conditions，they will learn in the same way.What appears to be variation in learning ability is really no more than different learning experience.This would include previous as well as actual experience，but in practice much less emphasis is placed on the influences of past learning than on the factors operating directly in the situation in which the learning is going on.

Every utterance and every part of an utterance is produced as the result of the presence of some kind of‘stimulus’.The stimulus，to which the utterance forms a‘response’，may be physically present in the situation；it may be verbal，since language can be produced as a response to other language，or it may be internal，in that a state of thirst may provide the stimulus for an utterance like‘I would like a glass of water’.For a child to learn to make such a response in the first place，his first attempts at producing the piece of language will have to be‘reinforced’.A child can be reinforced in a variety of ways.Parental approval acts as a powerful reinforcer.A physical need in the child may be met as a result of his utterance.His own language may act as a stimulus to action or language on the part of someone else.If such reinforcement does not take place，as when a child sayssomething that is not understood by those around it，the piece of language，the response，is not learned.In this way，correct pieces of language are acquired，but incorrect utterances are not.

A single emission of a response，even if it is reinforced，is by no means enough for learning to take place.Only if a response is repeated can it be fully learned.Indeed strength of learning is measured in terms of the number of times that a response has been made and reinforced.A word that has been uttered thirty times is better learned than one that has been said twenty times.The notion of repetition is therefore extremely important.More important still is the fact that a response that is not made cannot be repeated and reinforced and therefore cannot be learned.It is the making of the response that is the learning process.If there is no reinforcement，the learning is then extinguished.If an individual learns by making a response，it follows that he cannot learn from situations that do not demand a response from him.He does not learn to produce language by watching and hearing other people use it.Nor does he learn by having language use described or explained to him.If it is active language use that he is to learn，his responses must be active language responses.Hearing may be a response，but through repeated hearing one learns only to hear and not to speak.

In short，one learns only what one practices doing.As long as a child is making correct responses，he can be reinforced and will learn.If he made only correct responses，he would learn that much more rapidly.In practice he may well utter incorrect forms.He may not hear accurately a form that he is expected to imitate，or，more significantly，he may create an incorrect form by analogy with another form which he has learned correctly.Having learned that filled is the past tense of fill，he may overgeneralize this response and apply it to sell‘hereby producing the form selled’.The ability to make analogies is taken as given by behaviorists and it applies to non-linguistic forms of behavior too.If incorrect responses were never reinforced this might not be an important consideration.However，an incorrect form can easily appear in a generally correct utterance and if the whole thing is reinforced，the incorrect is being learned as thoroughly as the correct.In any case，as learning advances，the absence of an adverse reaction on the part of hearers may be sufficient to provide reinforcement and the learner then becomes the judge of his own speech.Error may then pass unnoticed，be repeated and thereby be learned.

Since in his description of language behavior，the psychologist is concerned with the mother-tongue，like the linguist he is bound to assign primacy to speech.In all but pathological cases，language appears first in the form of speech.It is inevitable that speech should be far more significant for the psychologist than writing which only develops later as a secondary form，very largely derived from speech.Writing is not a necessary stage in the developmental process and indeed it is still absent from the behavior of many human beings.The approach to writing，therefore，can only be through speech.In this summary of behaviorist views，there is one further，most important point to be mentioned and this concerns meaning.So far correct and incorrect responses have been talked of as if this was entirely a matter of correctness of form.But a response does not only have to be formally correct，it has to be the appropriate response to the stimulus.In learning a language one has to acquire both a‘formal’and a‘thematic’repertoire of responses.Most behaviorists eschew use of the word‘meaning’altogether.To think of the meaning of a word as something built up and stored inside the individual is to imply a mental structure of some kind that the behaviorist finds objectionable and unnecessary.What we might traditionally call meaning is simply the ability to produce an appropriate response to a stimulus.One does not need to use such terms as‘understanding’.As he learns his mother-tongue，the child comes to recognize not just individual stimuli but whole classes of stimuli to which a word or utterance is an appropriate response.He learns to draw the boundary between those situations in which the word may be properly used and those to which it would be an incorrect response.The individual does not control his use of words.Through his learning experience，the words are automatically elicited from him by the presence of the appropriate stimulus.People can be satisfied with any language performance where the utterance of the child constitutes'an adequate response to the stimuli that are present.We do not need to enquire into the‘meaning’of what has been said，or whether the child has‘understood’or‘meant’what he has said.