5.3.3 Behaviorism
（1）Introduction
Behaviorism is associated with the experiments on conditioning animals'behavior，which were based on stimulus-responsereward.They were carried out by Pavlov in Russia towards the end of the nineteenth century and improved upon later by Watson in America in the twenties.The latter believed that learning processes were based on complex chains of stimuli and responses—the more complex the stimuli and responses one gets，the more complex learning becomes（Brown，1980）.Skinner was the influential psychologist who took the lead in neo-behaviorism from the thirties till the sixties.He claimedthat S-R theory could be transferred to the study of human behavior including speech（see Skinner，1957）with minor adjustments，thus reviving the Lockians'“tabula rasa”-based principle in learning theory，and rejecting the study of the mind；i.e.the unobservable.
In the Skinnerian framework language acquisition was seen in the following way：the child produces a combination of sounds that resembles a meaningful word，the parents reward it by a reinforcing smile，a kiss，or by producing the object referred to operant or instrumental conditioning（Rivers，1964，174-75）.From the one-word utterance stage，the child moves to the twoword utterance，till he reaches the sentence/utterance stage—“a building up”habit formation that reminds us of Skinner's experiments on pigeons.As the child acquires more syntactic and morphological habits，he proceeds by generalizations and substitutions through trial and error while relying on his parents'secondary reinforcements or reinforcing himself.Deviant responses are not rewarded and are consequently subject to extinction.According to the behaviorists，these stages were not predetermined；and since the one-word utterance functions like a sentence-utterance，there is little to be said about stages of cognitive development.Skinner's theory faced much criticism，the harshest of which was by Chomsky（1959）.Skinner's emphasis on the observable made him concentrate on the response to infer the stimulus.One human being's responses could be different from another's.Suppose that three or thirty people were shown a picture；their reactions would range from“Remember our holiday in Italy”to“Gosh？”or to silent admiration，surprise or horror.The incapacity to predict the responses or to identify the stimuli without the responses，made the whole theory look unscientific.However，the avoidance of dealing with abstraction，including displacement—“the not here and not now”—and the view of man as“a bundle of twitches”encouraged the mediation theoreticians to propose some ideas which were still within the stimulus-response frame.
Although he was not the originator，the birth of behaviorism is most commonly associated with John B.Watson.John Watson was a colorful character who argued the behaviorist viewpoint from an extremist position.Essentially，he put forward that psychology should not be based on the accumulation of subjective impressions about mental concepts.Watson's view was that psychology should be based solely on observable events and that it should divorce itself from mentalistic conceptssuch asimagery and consciousness because these were unobservable（Watson，1913）.In developing behaviorism，Watson was undoubtedly influenced by two major scientific figures，Pavlov and Darwin.
Pavlov had worked extensively on understanding the conditioned reflex and had shown in a variety of experiments that naturally-occurring responses such as salivation in the presence of food could become associated with previously neutral stimuli such as a bell—a process known as classical conditioning.For Watson，this was an ideal situation：here was a learning mechanism which apparently could be fully understood simply in terms of the external events experienced by the animal.Pavlov worked on basic reflexes in animals and it might be argued that there are fundamental differences between these phenomena and those that a student of human psychology must study to account for all human behavior.At this point Watson turned to Darwin and his theory of evolution.In his writings，Darwin had stressed the continuity of species including the then controversial idea that humans evolved from apes.Crucial to Darwin's thinking was that the continuity between species was not only biological but also behavioral.Thus，in a memorable quote Darwin asserted that：“Psychology will be based on a new foundation，that of the acquirement of each mental power and capacity by gradation.Light will be thrown on the origin of man and his history.”
Watson used Darwin's evolutionary theory to argue that psychology could be studied meaningfully by systematic investigation of how a simpler animals learned，that all animals，including humans，learned in the same way and that humans were quantitatively，not qualitatively，different from other animals.In his writings，Watson not only sought to dismiss mental concepts from the realm of scientific psychology，he also argued that they could never play any causative role in the explanation of behavior；they were epiphenomena.An epiphenomenon is something that happens as the end result of a process—rather like the exhaust fumes of an engine.Consciousness，for example，was attributed to feedback when the throat muscles initiated speech responses.As for images，Watson denied that they existed at all and challenged anyone to prove him wrong.
Watson's account of learning was，in essence，extremely simple.All behavior arose as a result of conditioned relationships between stimuli and responses.In order to understand why an individual behaved in a certain way，one need look no further than the observable events surrounding the development of that individual.To make this point forcibly，Watson carried out the now famous experiment on“Little Albert”in which a previously neutral stimulus，a white rat，was paired with a noxious stimulus，a loud bang，until a point was reached where Little Albert cried at the appearance of the rat on its own.Later，Watson suggested that Albert might seek psychoanalytic help to understand his fear of rats when all that was needed was knowledge about how he had been conditioned.
Watson bowed out of academic life prematurely as a result of a domestic scandal，but by then the behaviorist movement had an enormous momentum with new figures adopting the same stimulus response framework.These figures included Edwin Guthrie and dark Hull but the most famous was B.F.Skinner.The major achievement of Skinner was to establish that conditioning took two forms.We have already considered classical conditioning in which a pre-existing，innate response comes to be associated with a new stimulus.Skinner pointed out that there was another form of learning that he termed operant conditioning.According to this idea，an animal may produce a response and depending on the outcome，that response may or may not be repeated when the same situation is encountered again.
If Skinner had limited himself to the explanation of animal behavior，then it is arguable that the whole course of psychology might have been different.However，Skinner was an extremist and sought to apply his approach as widely as possible，and this took him to the difficult issue of human language.In his book Verbal Behavior（Skinner，1957），Skinner argued that speech was not the expression of ideas，rather it was the emission of“verbal responses”.According to Skinner，all language is somehow under the control of some stimulus so that any utterance is considered to have an identifiable external cause.
Skinner's ideas about language came under attack from the linguist Noam Chomsky who，in a famous book review（Chomsky，1959），showed just how inadequate a behaviorist account of language was.At one point Chomsky examines the notion of stimulus control of language when considering someone looking at a painting，A person might say；“Dutch”or instead say“Clashes with the wallpaper”，“I thought you liked abstract work？”，“Never saw it before”.“Tilted”，“Hanging too low”，“Beautiful”，“Hideous”，“Remember our camping trip last summer？”To account for this variety of responses，Skinner is forced to argue that the picture itself comprises many smaller scale stimuli and that each can evoke a different response.As there is no way of restricting the number of stimuli that might potentially constitute a complex stimulus，Skinner's concept of stimulus control became valueless.
Another problem for Skinner is that much of what we hear and say is novel.According to Skinner，our ability to respond to a sentence we have never heard before arises because there is some similarity between that sentence and one we have heard before.Chomsky viewed this idea as“pointless and empty”and that it was much easier to explain our comprehension of novel sentences by asserting that we have internalized the grammar of our language.The need for this idea is neatly illustrated by the phenomenon of over-generalization that can be observed in children's linguistic development.An example of this concerns irregular plurals，e.g.mouse-mice.At an early stage of development，children might use the correct plural but，later on，they start to use an overregularized form（e.g.“mouses”）.How could reinforcement explain this？The child has never heard it before so it could not have met with reinforcement.The only way to explain this is to assert that the child has acquired a rule which it is misapplying—in other words it has formed an internal representation.
（2）The major features of behaviorism can be summarized in the following
（a）The need to develop theories that explain what goes on in our environment is as old as the human race.Although human behavior（including learning）was long considered unresearchable，behavioral scientists today are making discoveries that are useful to teachers in the classroom.
（b）Almost always，behavior is more complicated than it seems at first glance.Theories of learning with a scientific basis help us to view learning more realistically and enable teachers to teach more effectively.
（c）Learning refers to relatively permanent changes in behavior that occur as a result of experience.These include changes in the probability，latency，rate，and magnitude of behaviors.
（d）The behavioristic approach to studying learning focuses on environmental conditions that stimulate learning.The emphasis for the behaviorist is specific stimuli that produce the responses we call“learning.”
（e）One of the earliest contributions to understanding learning in terms of stimulus-response connections was made by Ivan P.Pavlov who discovered that a previously neutral stimulus（such as a bell）that produces an orienting response is capable of eliciting a conditioned response（such as salivation）when presented by itself，after being paired with an unconditioned stimulus（such as meat）.This simple alteration of behavior as the result of experience became known as“classical conditioning.”
（f）Pavlov's work was brought to the attention of American psychologists by John B.Watson，a leading figure in behaviorism who was able to demonstrate that emotional reactions of humans could be modified by classical conditioning procedures.Classical conditioning principles have also been useful in studying behavioral phenomena such as prejudice and interpersonal attraction.
（g）Classical conditioning appears to be one form of Edwin Guthrie's principle of contiguity，which holds that once a stimulus and response have become associated，a recurrence of that stimulus will tend to be followed by the same response.Guthrie believed that any form of complex learning could be explained by variations on his simple law of learning.His principle of contiguity proved to be especially useful in breaking learned habits.
（h）Another behavioristic explanation of learning was offered by B.F.Skinner，who did much work on operant conditioning.To produce operant conditioning，a stimulus is introduced after an organism makes a particular response（or an operant）.If the response tends to increase，once it is made contingent on that stimulus，then it is said that the stimulus is reinforcing.
（i）Reinforced responding may occur without learners'awareness.Different schedules of reinforcement have been shown to affect response acquisition，rates of responding，and performance under extinction.Principles of operant conditioning are often applied in the classroom，and performance plotted as a function of time yields useful information in the form of a learning curve.
（j）Learning theorists who study operant conditioning distinguish among positive reinforcement（a stimulus that increases the probability of a response when added to a situation），negative reinforcement（a stimulus that increases the probability of a response when removed from the situation），and punishment（an aversive stimulus that is aimed at decreasing the probability of a response when added to the situation）.Punishment produces unpredictable and often undesirable effects.
（k）Teachers often face the problem of determining what types of reinforcement are most effective for changing behavior.The age of children as well as their home background influences whether symbolic or concrete reinforcement will be most effective.
（l）Behavior modification is an application of operant learning principles aimed at changing undesirable behavior in positive ways.Teachers who use“behavior mod ification”are told to ignore students who are misbehaving and give them attention only when they are behaving in satisfactory ways.
（m）Much research shows that behavior modification，when properly applied，can be an effective classroom technique for teachers intent on increasing positive behaviors and reducing or eliminating undesirable behaviors.Teachers who use behavior modification techniques may set up token economies，which can be integrated with contingency contracts and used with the Premack principle：reinforcing students'performance by permitting them to engage in the more interesting activities once they have completed the less interesting tasks.
（n）One of the most effective reinforcers that a teacher can use is praise which is also important for teachers to set standards for reinforcement and arrange for self-reinforcement，whereby students select criteria for rewards and reinforce themselves.
（o）Skinner's answer to the practical problems many teachers face in using operant conditioning in the classroom was the“teaching machine.”Its operation was based on reinforcement principles but did not live up to its early promise.The teaching machine did stimulate programmed learning，which was the basis of computer-assisted instruction（CAI）as well as other personalized systems of instruction.
（p）Learning theories based on conditioning and reinforcement principles undoubtedly extend our understanding of some，but not all，learning processes.They are particularly deficient in illuminating students'thinking processes.For this reason，educators turn to cognitive views of learning，which focus on the ways internal mental processes can be enhanced through instruction.
（3）Early doubts about behaviorism
From the outset，not everyone shared Watson's view that behavior could be explained fully without the help of unobservable constructs.This was true even for those who worked on learning in animals.Edward C.Tolman was strongly opposed to the idea that all behavior could be understood solely in terms of the relationship between stimulus and response and，along with his students，he designed ingenious experiments that challenged the extreme behaviorist view.Tolman argued that most behavior was determined by a general goal rather than a set of inflexible links between stimuli and responses.For example，a bird has the goal of building a nest but the resultant behavior is flexible，taking account of variations in the availability of nest building material and other events that might intervene during the process.
Linked to this idea of a goal，Tolman argued that when animals were learning，they developed an internal representation of the problem they were trying to deal with—something termed a cognitive map.In one experiment，rats were taught to swim through a maze to obtain food from a goal box.The water was then drained from the maze and the rats placed in the maze again.Almost immediately they were able to run through the maze to obtain more food.According to strict behaviorist ideas，this transfer of learning should have been a very slow process because of the massive change in responses required（swimming to walking）and stimuli（wet to dry）.Tolman argued instead that the rats had developed some internal representation of the maze so that they“knew”where the food was independently of the specific physical conditions under which they had learnt about the maze.Perhaps Tolman's most famous experiment is one in which rats were first trained to run through a runway to obtain food from a goal box.Once this response was established，the apparatus was changed by the addition of 18 new pathways and the blocking of the pathway the rats had learned to run down.The interesting question was，where would the rats go now？According to a simple stimulus-response account，the rats would either have no idea where to go or，if some degree of generalization had occurred，they might preferentially choose paths 9 or 10 because these are most similar in position to the trained pathway.In fact，the rats greatly preferred pathway 5—a finding Tolman attributed to the rats having a map of where the food was.
（4）Limitations of the behavioral view of learning
The proposition that people learn because their responses are conditioned or reinforced is one that is acceptable to teachers，psychologists，and non-professionals alike.It is an idea that helps explain the learning that takes place in a classroom—but only up to a point.It helps explain why Jean learned her multiplication tables faster than anyone else in the class-namely，because her performance was reinforced by the teacher's encouragement.But it does not explain why Sara，who could have learned the tables as fast as Jean and was encouraged as often as Jean was，never learned them at all.We can say，of course，that the teacher's words of encouragement had a reinforcing effect on Jean's behavior but not on Sara's，but such an explanation leaves us at loose ends.Why did the teacher's attempts at encouragement not reinforce Sara's behavior？Nor does reinforcement theory give us any clues to what kinds of reinforcements we should use to deal with learners more effectively.
One of the dangers in a conditioning or a reinforcement approach to learning is that one may become overly concerned with techniques.Thus it is easy to believe that one can promote learning by the proper technique or combination of techniques.It is this belief in a mechanical approach to learning that leads one to ask the“experts”：“How can I get my Sunday school class to learn the books of the Bible？”or“How can I get my child to stop sucking her thumb？”or“How can I teach students the importance of good table manners？”
The psychology of instruction has，in recent years，shifted from the behavioristic orientation exemplified by the work of B.F.Skinner to a cognitive and affective（that is，thinking and feeling）model.In such a model，the learner is viewed as one who selects，transforms，and translates the data of experience into representations or concept.They are both meaningful and useful.The idea that learning is“primarily a change in behavior due to reinforced practice”seems too narrow today，and psychologists now prefer cognitive approaches that emphasize that one can learn by observing others，by watching a model，by viewing a demonstration，by listening to a lecture，by being told，by reading a book，by constructing images，and by elaborating words into sentences.In brief，cognitive approaches emphasize that one can learn without practice or reinforcement of overt behavior，and that one may learn by actively changing perceptions of experience，by constructing new meanings and interpretations of events.
In a review of research in the psychology of instruction，Lauren B.Resnick（1981）also concluded that most research on teaching processes is now focused on cognitive behavior and is concerned with the ways in which internal mental processes can be enhanced through instruction.
If one put theories based on conditioning and reinforcement to the test of an adequate theory of learning，it appears that they undoubtedly extend and sharpen our understanding of some learning processes but not all.They are most useful in explaining the learning that occurs incidentally—even accidentally—in the classroom，as well as in everyday life outside its walls.They extend our understanding of a limited number of the conditions and forces that affect learning，and within the confines of the laboratory situation they permit reasonably accurate predictions about the behavior of learners.But they are most deficient when it comes to clues to why students learn or do not learn.They tell us nothing about students'thinking processes or the motivation that leads them to seek or avoid learning experiences.