Behaviorism (John B. Watson – 1913)

Thorndike and Pavlov provided important contributions to behavioral psychology, but it was John B. Watson (1878-1958) who championed the popular behaviorist movement. Pavlov’s contribution was made from the discipline of physiology and was somewhat indirect. His connection with American behavioral psychology was initially made by Watson, who felt that Pavlov’s experiments provided a good example of a sound experimental method used to observe the conditioning process of the secretory reflex, by monitoring the flow of saliva (Watson, 1916, p. 92; 1928, p. 35; 1930, p. 50). As for Thorndike, it is unlikely that he would have labeled himself a ‘behaviorist’, since it wasn’t until 1913 that the term began to come into vogue. This new term, and the perspective on the study of psychology to which it referred, quickly became the dominating school of psychology in American universities. It was in his article entitled, Psychology as the Behaviorist Views It, that Watson (1913) positioned behavioral psychology as “a purely objective experimental branch of natural science” with a “theoretical goal” of “prediction and control of behavior” (p. 158). Watson (1928) more plainly defined behaviorism by saying that,

Behaviorism is the scientific study of human behavior. Its real goal is to provide the basis for prediction and control of human beings: Given the situation, to tell what the human being will do; given the man in action, to be able to say why he is reacting in that way. (p. 2)

Later, in reflecting on the behaviorist movement, he wrote,

Behaviorism, as I tried to develop it in my lectures at Columbia in 1912 and in my earliest writings, was an attempt to do one thing—to apply to the experimental study of man the same kind of procedure and the same language of description that many research men had found useful for so many years in the study of animals lower than man. (Watson, 1930, p. v)

Watson’s initial research focused on animal subjects such as rats (1903), rabbits (Watson & Watson, 1913), birds (e.g., 1907; 1908a; 1910), and monkeys (1908b; 1909). But by the year 1919 he had been able to apply the same experimental procedures to the study of man—the goal he had established for himself in his 1913 article. This article has come to be referred to as the Behaviorist Manifesto.

Through his own efforts and through the reports of other researchers working in the same field, Watson collected data through “daily observation of several hundred infants from birth, through the first thirty days of infancy and of a smaller number through the first years of childhood” (Watson, 1930, p. 118). From this data he concluded that “young children taken at random from homes of both the poor and of the well-to-do do not make good subjects” (p. 149) because their behavior was too complex. His solution to this problem was to study hospital-reared children belonging to wet nurses. Perhaps his most famous experiments were those conducted to establish conditioned emotional responses in “Little Albert” by exposing him to various small animals and simultaneously sounding a loud noise that had been found to elicit crying. Through repeated pairing of the animals with the noise, the animals themselves came to elicit responses of fear, crying, and avoidance behavior—where previously they had not (Watson & Rayner, 1920). Several other experiments conducted with children are accounted in Watson’s 1930 publication entitled, Behaviorism.

Watson’s perspective on learning—i.e., his theory of habit formation—is illustrated in the following example generalized from his observations of several children in similar situations:

To make the whole process a little more concrete, let us put in front of the three-year-old child, whose habits of manipulation are well established, a problem box—a box that can be opened only after a certain thing has been done; for example, he has to press inward a small wooden button. Before we hand it to him, we show him the open box containing several small pieces of candy and then we close it and tell him that if he opens it he may have a piece of candy. This situation is new to him. None of his previously learned formed manipulation habits will completely and instantly work in this situation. None of his unlearned reactions will help him very much. What does he do? That depends upon his previous organization. If well organized by previous handling of toys, he goes at the problem at once—(1) he picks the box up, (2) he pounds it on the floor, (3) he drags it round and round, (4) he pushes it up against the base-board, (5) he turns it over, (6) he strikes it with his fist. In other words, he does everything he has learned to do in the past in similar situations. He displays his whole repertoire of acts—brings all of his previously acquired organization to bear upon the new problem. Let us suppose that he has 50 learned and unlearned separate responses at his command. At one time or another during his first attempt to open the box, let us assume that he displays, as he will, nearly all of them before he pushes the button hard enough to release the catch. The time the whole process takes, we will say, is about twenty minutes. When he opens it, we give him his bit of candy, close up the box and hand it to him again. The next time he makes fewer movements; the third time fewer still. In 10 trials or less he can open the box without making a useless movement and he can open it in two seconds. (Watson, 1930, p. 204)

Watson explained this instance of learning—the ability to open the box with increasing speed and with fewer and fewer useless movements—as a function of frequency and recency. The act that is performed most frequently persists while the rest die away. The act that has been performed most recently is more likely to appear sooner in the next succeeding trial. Watson’s explanation of recency and frequency as the basis for habit formation was criticized by some writers, and specific experiments were performed to demonstrate the inadequacy of these two factors alone to account for learning (Gengerelli, 1928).  However, these factors do not form Watson’s complete picture of learning. In his introduction to a republication of Watson’s Behaviorism (Watson & Kimble, 2002, p. xii) Kimble lists nine hypothetical laws of learning identified by Watson.[1] The first two are frequency and recency. The remaining seven are

3. Conditioning is a process of stimulus substitution: “The [conditioned stimulus] now becomes a substitute stimulus—it will call out the [response] whenever it stimulates the subject” (p. 21)

4. The process of conditioning is ubiquitous, “So far as we know we can substitute another stimulus for any stimulus calling out a standard reaction” (p. 22). Thus, learning never produces truly new responses. “The organism starts out life with more unit responses than it needs” (p. 24). The process that appears to establish new responses “concerns itself really with stimulus substitutions and not reaction substitutions (pp. 25-26).
Laws 5-9 came from Pavlov, by way of G. V. Anrep (Watson does not give a reference).
5. “Conditioned responses [may be] temporary and unstable. After periods of no practice they cease to work [but they can] be quickly reestablished.”
6. “The substituted stimulus can be made [so specific that no] other stimulus of its class will then call out the reflex.” But, in apparent contradiction to this idea, Watson also noted that conditioned responses generalize (transfer) to similar conditioned stimuli.
7. “The magnitude of the response is dependent upon the strength of the [conditioned] stimulus”.
8. “There is a marked summation effect. If a dog is conditioned separately to [two stimuli], there is a marked increase in the [strength of the response] if the stimuli are given simultaneously.”
9. “Conditioned responses can be ‘extinguished’” (pp. 28-29).

Though Watson’s role as the recognized founder of behaviorism as a school of psychology is clear (Morris & Todd, 1999), his impact on educational learning theory is limited, as evidenced by the (at best) tangential coverage he is given in comprehensive books on learning theory (e.g., Bohlin et al., 2009; Bower & Hilgard, 1981; Driscoll, 2000; Eggen & Kauchak, 1999; Hilgard, 1948; O’Donnell et al., 2007; Olson & Hergenhahn, 2009; Ormrod, 2003; Sternberg & Williams, 2010; Woolfolk, 2010). Perhaps this is because his explanation of frequency and recency was never fully accepted as sufficient to account for learning, and because his other laws—as summarized by Kimble—weren’t really unique, with most of them having been adopted without change from Pavlov.


[1] The page number references used by Kimble are based on the 1925 printing of Watson’s Behaviorism

 

Classical Conditioning (Ivan Petrovich Pavlov – 1928)Operant Conditioning (Burrhus Fredric Skinner – 1938) >

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s