Before he invented behaviorism, John B. Watson considered learning one of the most important topics in psychology. Watson conducted excellent empirical research on animal learning. 

 

 

Before he invented behaviorism, John B. Watson considered learning one of the most important topics in psychology. Watson conducted excellent empirical research on animal learning. He developed behaviorism in part to promote research and elevate the status of learning in psychology. Watson was much less successful in the adequacy and originality of the mechanisms he proposed to explain learning. By assimilating the method of classical conditioning and adopting Pavlov’s theory of stimulus substitution, Watson linked behaviorism with a new method that could compete with both Titchener’s method of introspection and Freud’s methods of psychoanalysis. Watson’s interest in explaining psychopathology led to the discovery of conditioned emotional responses and a behavioristic explanation for the learning of phobic behavior. Watson established learning as a central topic for basic research and application in American psychology. 

Learning in animals is probably the most important topic in the whole study of behavior (Watson, 1914, p. 45). 

No experimenter has yet set his experimental problems in such a way as to construct from his data a guiding theory of habit formation (Watson, 1925, p. 25). 

In 1930, when he revised Behaviorism for the last time, Watson expressed satisfaction that behaviorism was strongly entrenched as a point of view in American psychology. He took the occasion to repeat a major theme of his career, the contrast between the old psychology of James’s and Titchener’s introspection and what he called the new psychology of behaviorism. From his point of view, Watson had achieved great success in his primary goal of changing the subject matter of psychology from consciousness to behavior. After telling his readers that consciousness was the subject matter of the old psychology, Watson repeated the main point of his manifesto on behaviorism (Watson, 1913a), that “behaviorism, on the contrary, holds that the subject matter of human psychology is the behavior of the human being” (1930, p. 2). 

What historical factors led Watson to propose a change in the subject matter of psychology from consciousness to behavior? Although many factors, both in American culture at large and in American psychology, contributed to the origins of behaviorism (Burnham, 1968; Mills, 1998; O’Donnell, 1985), the challenge of explaining learning was central to the origins and subsequent development of John B. Watson’s behaviorism. By 1930 Watson could also take satisfaction that not only behaviorism, but also the topic of learning, was strongly entrenched in American psychology. 

For Watson, behaviorism represented much more than a change in the subject matter of psychology. Many topics in psychology could be approached objectively. Indeed, in one of the first of a genre of critiques of behaviorism, Titchener (1914, p. 5) pointed out that calls for objective psychology were common in the history of psychology, so “Watson’s behaviorism is neither so revolutionary nor so modern as a reader unversed in history might be led to imagine.” The target of Watson’s behaviorism was all introspectionists, especially Titchener. In order to replace the method of introspection, Watson issued a call for the adoption of new methods in psychology for the study of behavior. A corollary of behaviorism was Watson’s call for the adoption of new methods for the study of Learning. 

Behaviorism elevated the status of learning as a topic of research by psychologists. Behaviorism became Watson’s chosen vehicle for calling attention to research on learning. Watson’s (1925) popular book, Behaviorism, was a showcase for research on learning, especially Watson’s own research on emotional learning in children. For Titchener, the most important topics in psychology were sensation and perception. (See Tweney, 1987, for the details of Titchener’s program of research in experimental psychology.) For Watson, behaviorism was a platform from which he attempted to persuade his colleagues and the public that learning was a more important topic for research than were sensation and perception. 

In contrast with Watson’s success in changing the subject matter of psychology from consciousness to behavior, and as Boakes ( 1994) pointed out, Watson was much less successful as a learning theorist. The quotation from Watson with which this article begins indicates that he eventually recognized that his efforts to develop a behavioristic theory of habit formation were not successful. Watson failed to find a behavioristic substitute for Thorndike’s law of effect. He also failed to develop a heuristic explanation for maze learning. After writing an editorial calling for research in comparative psychology on imitation (Watson, 1904), Watson (1908, p. 172) may have been slightly embarrassed when he reported that with respect to his own research on imitation in monkeys, “I unhesitatingly affirm that there was never the slightest evidence of inferential imitation manifest in the actions of any of these animals.” Thus, imitation represented a failure in Watson’s efforts to advance comparative psychology by investigating learning. Watson (1925) simply adopted Pavlov’s theory of stimulus substitution as an explanation for Pavlovian conditioning. Thus Watson does not rank with Pavlov and Thorndike as a learning theorist. Watson’s major contribution to learning theory was the prediction and discovery that emotional responses could be conditioned (Watson & Rayner, 1920). With the exception of the discovery of conditioned emotional responses, Watson’s enduring achievement for learning was an increase in the status of the topic. By identifying the mechanisms of trial and error learning and maze learning as unsolved problems in experimental psychology, Watson paved the way for the neobehaviorism of Hull, Tolman, and Skinner in the 1930s. 

The topic of learning has been neglected in prior scholarship on Watson (for an exception, see Malone, 1990a). Todd (1994) identified Watson’s extension of the province of learning as one of the topics requiring additional historical analysis. He asked, “Could behaviorism have become an important philosophy without specific empirical principles?” (1994, p. 165). The answer is “yes.” Even with very few wellestablished principles of learning, Watson’s behaviorism achieved its collateral objective of increasing the attention paid to the topic of learning. Therefore, an exploration of the role of learning in John B. Watson’s published papers provides an additional key for placing classic behaviorism in a proper historical context. 

How Watson spotlighted the topic of learning 

Learning is mentioned only obliquely in Watson’s ( 1913a) famous manifesto for behaviorism, “Psychology as the Behaviorist Views It.” The purpose of this paper was to change the subject matter of psychology from consciousness to behavior, so it represented more of an attack on Titchener’s structuralism and the functional psychology Watson had been a part of at Chicago than a vehicle for presenting fresh data on learning. Watson mentioned an eclectic melange of several topics for research that could be approached behavioristically. The list included the study of learning and habit formation, instincts, the comparative study of sensory processes including vision in both animals and humans, and a set of applied problems including psychopathology. In his manifesto of 1913, Watson’s characterization of the areas of basic research in psychology was a lament about what he believed to be experimental psychology’s low status among the natural sciences, which he put as follows: 

I do not wish unduly to criticize psychology. It has failed signally, I believe, during the fifty-odd years of its existence as an experimental discipline to make its place in the world as an undisputed natural science. Psychology, as it is generally thought of, has something esoteric in its methods. ( 1913a, p. 163) 

The key to understanding this quotation is Watson’s use of the word methods. Watson’s goal was to replace the method of introspection with “objective” methods. Consider the topics of sensation and perception, which were an important area of research for Titchener and his students (Tweney, 1987). Watson could have simply called for a behaviorism that represented a return to Fechner’s objective methods of psychophysics, but he did not. In his manifesto, Watson did not single out any particular topic in experimental psychology for special favor. 

After the manifesto for behaviorism, Watson’s next major opportunity to advance the agenda of behaviorism was his book on comparative psychology. He used this book as a platform from which to elevate the status of learning as a topic for research by psychologists. Consider Watson’s comments about learning in his treatise on comparative psychology, Behavior: An Introduction to Comparative Psychology (1914). One might expect that instinct would emerge as a key concept in a textbook on comparative psychology. Watson defined an instinct as a behavior on the part of an animal that did not require learning, and he devoted two chapters in Behavior to a discussion of instincts. (See Dewsbury, 1994, for a portrait of Watson as a comparative psychologist, including his interests in instincts.) However, Watson reserved the most praise in Behavior not for instincts, but for the topic of learning, whose importance he described as follows: 

On account of its bearing upon human training, learning in animals is probably the most important topic in the whole study of behavior. Entirely apart from this connection, this division contains the behaviorist’s most important group of problems, since by means of habit formation he finds the most direct way of controlling animal activity. (1914, p. 45) 

Consider three points in the preceding quotation. First, notice that Watson was not interested in animal learning per se as a topic for scientific exploration, but primarily as the topic related to human learning. Second, because the goal of Watson’s ( 1913a) behaviorism was the prediction and control of behavior Watson found the topic of learning relevant because it provided him with a tool for controlling behavior. Third, Watson described the topic how called learning as habit formalion. Habit formation is no longer used as a synonym for learning. As Malone (1990b) pointed out, Watson borrowed the concept of habit formation from William James (1890/1950). 

Watson ( 1917a) had another opportunity to spotlight learning when he reviewed Holt’s (1915) book on psychoanalysis. Holt was an American psychologist from Harvard who attended Freud’s lectures at Clark University in 1909. After he returned to Harvard he wrote a book on ethics that was influenced by Freud’s idea of wish fulfillment. Holt called his book The Freudian Wish and Its Place in Ethics. The book contained a discussion of the Meynert problem, a hypothetical and philosophical explanation of how a child might learn not to put his or her finger into a candle flame despite the impulse to reach out and touch an attractive object. Holt highlighted the topic of emotional learning in children in his book and pointed out that “fear is a normal and necessary ingredient of the learning process” ( 1915, p. 71 ) . However, Holt did not of fer any data, and his discussion of learned fear in children was philosophical. The significance of Holt’s book was that after listening to Freud, Holt talked about learned fear as a psychological topic worthy of the attention of psychologists. Holt (1915, p. 74) went on to conclude his theoretical discussion of the Meynert problem with the statement that “the mechanism of learning is by no means understood as yet.” 

In a review of Holt’s book, Watson’s response to the challenge of Freudian psychoanalysis was a call for psychologists to turn to research on learning. Watson (1917a, p. 86) believed that psychoanalytic concepts, which he considered mystical entities, could be explained in terms of “well-known principles of habit formation.” Watson followed Holt and discussed the Meynert problem. In the context of understanding the mechanism by which children’s fears were acquired, Watson made another call for elevating the status of the topic of the mechanisms of learning within experimental psychology: 

In these few experiences a genuine learning process is involved and the explanation of this learning process-regardless of whether the act is acquired in few of many trials-is what I consider one of the chief problems in psychology. (1917a, p. 89) 

Watson has transformed his research interest in children’s fears from an old philosophical conundrum, how children learn to keep their hands out of a burning candle, into “one of the chief problems of psychology.” 

Watson was explicit in indicating that behaviorism represented a shift in the importance of the topics psychologists selected for research. For Watson’s behaviorism represented a shift away from sensation and perception toward a focus on learning and habit formation. Watson described the change as a shift from the structural analysis of the unit of sensation toward the units of learning and habit formation. He rejected 

the introspectionist unit concept [of] sensation. The reason for the omission is clear. When a unit changes in a science the problems and points of interest shift . . .The mechanism of habit formation . . must await the working out of just those factors which the behaviorists insist upon. ( 1917a, p. 92). 

Here Watson offered to set the theoretical agenda for successive generations of behavioristic learning theorists. Thus a characteristic of Watson’s rhetoric was to call attention to the importance of learning as a topic for research within experimental psychology. Another illustration of this practice occurs in a paper on maze learning, where Watson wrote, “The control of habit is one of the most vital problems in every system of psychology” (1917b, p. 59). Although Watson was a great salesman for research on the topic of learning, he was less successful as a learning theorist. His first theoretical failure was with maze learning. 

Watson’s failure to explain maze learning behavioristically 

Why did Watson abandon consciousness as an explanatory concept? One good place to start considering this question is by comparing the discussions of Small and Watson on maze learning in the rat. Small (1901 ) constructed a maze for white rats from a diagram of the Hampton Court Maze in the Encyclopedia Britannica. He placed food at the center of the maze and found that his rats rapidly learned to run from the start location, through a complex series of passages, until they reached the food. Looking down on the uncovered maze, Small recorded the errors when a rat entered a cul-de-sac and the amount of time the rat required to find the food. The maze was complicated and Small was impressed by the proficiency with which the rats mastered the task. Watson (1907) duplicated Small’s maze, replicated Small’s experiment, and began an experimental analysis of the factors that influenced maze learning in rats. Watson and Small obtained essentially the same data, but their interpretations of the mechanism of learning and even their choices for presenting the data were profoundly different. Small did not trust the learning curve of Thorndike (1898) as an adequate representation of the process of learning, but Watson proudly filled his paper with beautiful learning curves. Watson found that the average length of time to find the food was 16 min on the first trial and that the time dropped to about 1 min after only 10 trials. For both Small and Watson, the question was, What is the mechanism underlying the process of learning? 

Small’s mechanisms for maze learning: Consciousness and sensations 

In Small’s and Watson’s day, comparative psychologists were sharply divided over anthropomorphism and the law of parsimony. Small was an eclectic theorist who belonged in the camp of those who were willing to attribute components of human consciousness to animals. Unlike Thorndike (1898), he thought animals could reason and he chided the advocates of parsimony as inflexible extremists in very clear language: 

Some modern comparative psychologists’ abhorrence of anthropomorphism leads to the opposite extreme. . . . .Is not a certain amount of chastened anthropomorphism a wholesome specific, a kind of saving grace against scientific pedantry?. . . . The law of parsimony is important no doubt, but it may be employed too rigorously. (Small, 1901, p. 228) 

Small used what would now be called cognitive language to describe the process of maze learning: “The selection of paths begins to be purposive” (Small, 1901, p. 212). In other words, Small anticipated the language of Tolman’s (1932) purposive behaviorism to the extent of using the term purposive. Small never mentioned the construct of Tolman’s (1948) cognitive map, but he did use a similar cognitive concept, imagery, to describe maze learning when he wrote, “There seemed to be some kind of an image in his mind that he was trying to follow . .He apparently knew when he was on the right and when on the wrong road” (Small, 1901, pp. 212-213). Small mentioned consciousness frequently in his discussion. 

Small’s explanation of maze learning also included a second, much more molecular level of analysis. He discussed the various modalities of sensation, including smell, sight, and tactual motor sensations. The tone of this discussion was quite different from the discussion of consciousness. Small believed that the rats acquired a motor memory in learning the maze. This memory included a representation of distance intervals and proper turns. Small concluded his paper with a call for other researchers to consider the possibility of thinking about maze learning in tactual motor terms. Watson answered Small’s call for a sensory analysis of maze learning. 

Watson’s mechanism for maze learning: Kinesthetic sensations only 

Simply put, Watson applied the law of parsimony to Small’s mechanisms by rejecting Small’s discussion at the level of consciousness. In other words, Watson was willing to run what Small called the risk of “scientific pedantry.” Watson (1907, p. 94) rhetorically asked, “Why, then, in the case of the rat, need we assume the presence of motor images?” Watson followed Small by concluding that the rat learned the maze on the basis of kinesthetic sensations. Watson recognized that he arrived at this conclusion by a process of eliminating other sensory candidates. Watson rejected all of Small’s cognitive processes such as imagery as follows: “No one would dream of affirming that such a complexity in the cortical processes as this would call for could exist in the case of the rats” (Watson, 1907, p. 94). Of course, we now know from the work of O’Keefe and Nadel (1978) and other neuroscientists that the hippocampus is the structure in the brain of rats that is involved in maze learning, but this information was not available to Watson. 

Watson discovered that extramaze visual cues were important in maze learning for normal rats. When he rotated the maze by 180 for normal rats, he found that these rats were “absolutely lost” (1907, p. 87). However, this discovery was made rather late in the collection of the data and Watson decided to simply present the data rather than to speculate about a possible interpretation. 

In conclusion, Watson’s research on maze learning was basically empirical. Although Watson rejected interpretations of the data that included consciousness, he left the problem of maze learning without a satisfactory explanation. 

Although the Chicago school of maze learning in rats produced several additional publications (Carr & Watson, 1908; Peterson, 1917) the theoretical interpretation of the results in terms of Watsonian concept remained vague. At the end of an extensive monograph on maze learning, Peterson, who had worked as a graduate student with Watson at Chicago, reluctantly concluded that “the great problem of how learning takes place is yet largely unsolved” (1917, p. 46). Watson sought to impress the readers of Behaviorism (1925) with the learning abilities of rats in the maze by presenting an impressive learning curve for maze learning, but he remained fixated on his theoretical position that habit formation could be reduced to a kinesthetic sensation in which “the muscular stimuli coming from the movements of the muscles themselves are all we need to keep our manual responses occurring in proper sequence” (1925, p. 176). Watson’s school of maze learning came to a dead end before 1920. In Behaviorism, Watson (1925) reprinted the learning curve from his paper of 1907 and then concluded that the mechanism underlying maze learning and other manual habits “have never been worked out in a wholly satisfactory manner” (1907, p. 171). Watson’s legacy was to pass along to the next generation of researchers in animal learning the problems he had been unable to solve. 

Most historians of learning theory would conclude that the problem of maze learning remained unsolved until Tolman (1948) introduced the construct of the cognitive map. After discussing and rejecting the point of view advocated earlier by Watson, Tolman advocated a point of view called field theory that assumed that “in the course of learning [a maze,] something like a field map of the environment gets established in the rat’s brain” (1948, p. 191). Tolman’s (1932) purposive behaviorism represented a return to the kind of theorizing Small had introduced into learning theory at the beginning of the 20th century. However, Small does not appear in Tolman’s writing as the source of purposive behaviorism. Tolman (1932, p. 12) credited McDougall (1926) for the origin of the term purposive behaviorism. After Watson’s failure to explain maze learning, Tolman successfully proposed a cognitive explanation of maze learning that remains viable today. 

Watson’s failure to improve