Watson's applications of
behaviourism to advertising and to children



Although in his behaviourist call to arms in 1913 Watson was extremely scathing about ‘pure’ psychologists who were ‘not interested in a psychology which concerns itself with human life’, his own career switch from academic life to the hurly-burly of Madison Avenue was not altogether intentional. In one of his later books he uses as an example of ‘when and how we think’ the following dramatic episode:


    “ R’s employer called him in one day and said, ‘I think you would become a much more stable member of this organization if you would get married. Will you do it? I want you to answer me one way or the other before you leave this room because you either have to get married or I am going to fire you.’ ”(Watson, 1931, p. 242)


Something not entirely unlike this had happened to Watson in October 1920. The president of Johns Hopkins University had called him in one day, and made certain remarks about Watson’s personal life, with the result that Watson wrote out a brief note


tendering his resignation before leaving the president’s office. The problem was that Watson was having an affair with one of his research students, his wife had found love letters, and somehow or other this led to his resignation, remarriage and withdrawal from academia (Cohen, 1979). No one knows quite what happened. Tichener, the doyen of the traditional introspectionists (but, to his great credit, Watson’s most loyal personal friend and supporter during the crisis), thought that things were especially sensitive at Johns Hopkins because only a few years earlier the previous head of Watson’s department, Baldwin, had been forced to leave not only the university but the country after being caught with a child prostitute in a negro brothel (pausing only to hand over the editorship of the Psychological Review to Watson). The university was heavily dependent on the goodwill of the local and conservative financial community, and no doubt wanted to avoid further scandal. It seems unlikely that Watson was a martyr in defence of behaviourism; but quite probable that his fierce disagreement with prohibition, and his half-completed project designed to demonstrate the harmlessness of alcohol, together with his research into sexual attitudes and advocacy of greater sexual activity among students, inclined his employers, and many of his former friends and colleagues, against him.

Watson had already advocated field research, and the application of objective methods to human psychology, with the goal of finding ‘general and particular methods by which behaviour may be controlled’ (1914/1967, p. 11). But he cannot have anticipated doing a research project on the brands of rubber boots most favoured by the inhabitants of the banks of the Mississippi between Cairo, Illinois, and the sea. This was a temporary job in preparation for an address to the Boot Sellers League of America by the head of an advertising agency. Watson was an expert at preparing papers for conventions, and the agency head was sufficiently impressed with the work to offer him $10,000 a year for life.

Thus, by a quirk of fate, Watson the university professor became Watson the Madison Avenue executive, applying the ‘laws of human behaviour’ to advertising. In financial terms, this worked very well, since by 1930 Watson’s salary had been increased to $70,000, 4.5 per cent of the total wage bill of J. Walter Thompson, then as now a not inconsiderable agency. In addition Watson


blazed the trail in the 1920s for the popularization of the psychology of such matters as child-rearing techniques and sexual relations, commanding fees in the region of $1000 apiece for articles in magazines like Harpers and Cosmopolitan. Whether Watson should be credited with any permanent influence on advertising practice is difficult to say, but there seems little doubt that he himself was able to put his behaviourist principles to effective use in the commercial world. He spent some time lecturing Madison Avenue colleagues about the basics: ‘You must never lose sight of your experimental animal — the consumer.... What we are struggling with is the finding of the stimulus which will produce the reaction.’ And what is the reaction? ‘We want the man to reach in his pocket and go down and purchase. This is the reaction.’

Watson stressed a number of other points that now seem fairly obvious. The greater the quantity of a product that the consumer uses, the more he or she will purchase, and therefore advertising should encourage intensive usage. Apparently babies less than a year old were rarely dusted with talcum powder before 1924 —Watson recommended that authorities be found to suggest that infants should be powdered from the moment of birth onwards, and as often as possible. Johnson and Johnson could encourage the use of their brand by selling the idea of purity, and suggesting that a mother who did not use their powder was less of a mother. Toothpaste should be used after every meal, not just once a day, and in this case Watson himself spoke on the radio as an eminent scientist advising listeners on the care of their teeth. Usage can be encouraged for all sorts of toiletries, and Watson early on directed the advertising for Odorono, one of the very first deodorants.

Watson made use of both the observational and the experimental techniques of his former life in arriving at recommendations. He spent two months as a counter clerk at Macy’s department store in order to observe the consumer in its natural surroundings, and deduced that the position and arrangement of goods had a remarkably powerful effect on purchasing behaviour — clients were thereafter advised that placing their product by the entrance or by the cash register could in itself increase sales. But it took systematic experimentation to discover that smokers are surprisingly insensitive to the taste of their favourite brand — even after special training, Watson’s subjects were unable to identify the brand of cigarettes which they smoked without being able to


see from which packet they came. But this did not mean that smokers were indifferent to the identity of the cigarette; on the contrary, their ‘brand loyalty’ existed over and above characteristics of the product itself.

Using yet another method of investigation. Watson interviewed over a hundred potential purchasers of life insurance, and discovered that they almost all detested life- insurance salesmen. Therefore, he inferred, it was the customers’ reaction to the person doing the selling, rather than to the product being sold, that would determine purchasing responses. Arthur Miller’s Willie Lomax, who, in the play Death of a Salesman (1949), found that being ‘well-liked’ had turned to dust and ashes in middle age, might have attended one of Watson’s lectures as a young man in the 1920s, and been told ‘You are primarily selling yourself to these fellows and anything else you may be selling, any product you may be selling, is simply secondary. If you can sell yourself, you will have no trouble in selling them 12 dozen’ (Cohen, 1979). [see Watson's entry at

Childhood fears and neuroses

Although Watson was able to make a fortune by selling himself and his ideas to the American advertising industry, it is hard to believe that in doing so he contributed significantly to the general alleviation of human misery, even if he demonstrated that methods nurtured in the animal laboratory could be applied in the rough-and-tumble of the commercial world. His influence on behavioural methods of influencing emotions is also equivocal in some respects, since his pioneering efforts languished for several decades. But the limited experimental work he was able to perform on emotional reactions in young children supplied material which has been quoted in countless textbooks. It has recently been emphasized that Watson’s experimental work with young children was not above reproach, both in terms of its statistical methodology and in terms of its kindness to the infants involved. But there is certainly no doubt of its importance in the development of psychological theories, both in behaviour therapy and in the popular imagination (Harris, 1979; Samuelson, 1980; Kazdin, 1978).

Although many more than two children were observed in the course of Watson’s studies, his investigations were limited by his


move into advertising, and two infants have pride of place, Little Albert and Little Peter. Watson’s theoretical approach was of course to break down all emotional development into basic reactions. He thought that at birth there were only three different forms of emotional response, which would later develop into ‘fear’, ‘rage’ and ‘love’, these categories remaining the strongest bases for emotional appeals even in the adult consumer. Babies are most easily frightened either by very loud noises or by being dropped — both these produce startle, changes in breathing and blood circulation, and crying. Rage starts as a reaction to hampered bodily movement. Watson found that if he held a baby’s legs tightly together, or pressed its arms to its sides, a variety of movements were elicited, especially in the second week after birth. The infant usually stiffens its body, may lash about with its free limbs and often holds its breath, even with its mouth open, until its face turns blue (Watson, 1931, p.154, reassures us that ‘The experiments are discontinued the moment the slightest blueness appears in the skin’). The initial reactions of ‘love’ in the young baby include responses that might otherwise be labelled ‘affectionate’ or ‘good natured’, such as smiling, gurgling and cooing. However, the initial stimulus for these responses, the stroking, tickling and patting of the skin, is most effective, in Watson’s view, when applied directly to ‘the erogenous zones, such as the nipples, the lips and the sex organs’, even in the first few months of life (Watson, 1931, p. 155).

Thus the three basics of fear, rage and love were in Watson’s eyes the most instinctive human emotions, and the first to show themselves in young children; equally fundamental were the human necessities of food, shelter and sex. Even for the literate adult consumer, ‘every piece of good copy must be some kind of combination of these factors’, and the three emotions plus the three needs means that there are 720 combinations (Cohen, 1979, pp. 188—99). But adult emotional life is clearly more complex than combinations of infantile reactions to insecurity, constraints or tickling (although infant carry-overs are ‘the general cause of unhealthy personalities’), and the main mechanism of ‘how our emotional life becomes complicated’ is of course the mechanism of conditioning. The child’s unconditioned reactions to simple stimuli are only the starting points ‘in building up those complicated habit patterns we later call our emotions’ (Watson, 1931, p.


165). There is direct conditioning of emotional reactions to new stimuli, transfer of emotional effects indirectly to related sensations, and then a further level of complexity is introduced when the same stimulus (that is, often, the same person) causes fear in one situation and a little later comes to be associated with love — or rage. ‘The increasing complexity brought about by these factors soon gives us an emotional organization sufficiently complicated to satisfy even the novelist and poet’ (Watson, 1931, p. 165). Emotional life, normal or abnormal, was built up, Watson believed, by ‘the wear and tear of the environment’ taking its toll by the conditioning mechanism. As an extreme example, Watson supposed that the emotion of jealousy is not innate, or inevitable, but rather ‘Jealousy is a bit of behaviour whose stimulus is a (conditioned) love stimulus the response to which is rage’. When someone sees or hears of a loved object being tampered or interfered with, we expect to observe stiffening of the whole body, reddening of the face, pronounced breathing, verbal recrimination and possibly shouting (Watson, 1931, p. 194). This is hardly an exhaustive treatment of the psychology of jealousy, but illustrates how far Watson was prepared to go in breaking down what others would consider to be natural whole units of human emotion into supposedly more primitive parts.

In the case of jealousy, Watson supposed that this only arose through conditioning experiences, instead of being innate as the Freudians insisted. To test this out he and his wife tried making love in front of their one-year-old son, without ever eliciting jealousy (‘This was tested again and again’: Watson, 1931, p. 191). When they instead tried fighting in front of him they got only ‘behaviour of the fear type partly visually conditioned’. When this child was 2¼, Watson’s second son was born, and Watson did not detect any jealousy when the older brother was taken along to watch his infant brother breast-feeding. However, by the time the older boy was 2½ he had started to attack his father, crying out ‘my mama’ whenever he observed Watson kissing Mrs Watson, and Watson took this to indicate that jealousy developed gradually. Many of these ideas seem weird today, and Watson’s treatment of his own children is certainly not to be recommended to modern parents, but his book The Psychological Care of the Infant and Child was a widely used bestseller in the 1930s. Watson’s general approach was much admired by Bertrand Russell, and his more


extreme intentions for child rearing and the treatment of criminals made a considerable impression on Aldous Huxley, who seems to have used them for the basis of his novel Brave New World (1932). It was Watson’s ambition, though not, as far as I know, one that was ever fulfilled, to condition the behaviour of normal infants by electric shock. (‘I hope some time to try out the experiment of having a table top electrically wired in such a way that if a child reaches out for a glass or a delicate vase it will be punished, whereas if it reaches for its toys. . . it can get them without being electrically shocked’: Watson, 1931, p. 185; Huxley, 1983, pp. 27—30.)

This sort of treatment would hold equally well, Watson believed, ‘for adults in the field of crime’. Either people committed crimes because they were sick or they committed crimes because they were wrongly trained. In the first case the individuals concerned should be made well or, in hopeless cases, ‘etherized’. On the other hand, individuals with inadequate training should be given ten to fifteen years to learn a trade and ‘put on culture’. If this failed, ‘they should be restrained always, and made to earn their daily bread, in vast manufacturing and agricultural institutions, escape from which is impossible. Individuals put aside thus for additional training should of course be kept in the hands of the behaviourists.... Naturally such a view does away completely with the criminal law. . . and with courts for the trial of criminals’ (Watson, 1931, pp. 185—6).

It is hard to resist the conclusion that Watson was a Hitlerian monster but, fortunately, he was never in a position to put his more ambitious ideas into practice, and had to content himself with running his advertising accounts, and voting Republican.

The Little Albert experiment

There were two particular cases where Watson was able to gather experimental data on how conditioning experiences may alter emotional life in young children. In the first, the Little Albert experiment, fear was conditioned where none had been before. The second, the treatment of Little Peter, was one of several in which an attempt was made to remove fear and anxiety by special kinds of therapeutic experience. ‘Little Albert’ is the better known, partly because there are very few other re-


ported cases of the deliberate inculcation of fear in human infants.

Albert B. weighed 21 pounds and was 11 months old when the conditioning experiment began (Watson and Rayner, 1920). He was a son of a wet nurse at a hospital for invalid children, had lived most of his life at the hospital and was selected as being exceptionally stable emotionally — ‘In all the months we worked with him we never saw him cry until after our experiments were made!’ (Watson, 1931, p. 159). Albert’s first recorded crying occurred when he was almost 9 months old and Watson hit a 4-foot-long steel bar with a hammer just behind his head, while he was distracted by Rayner’s hand waving in front of him. Even then Albert remained comparatively unmoved until this procedure had been performed for the third time. He showed no sign of emotional after-effects between then and the conditioning test five weeks later. In this, a white rat was taken out of a basket and put in front of him. As the infant reached for the rat with one hand the iron bar was struck behind his head, causing him to jump violently. Albert then reached for the rat with his other hand, the bar was banged again, and he flinched and began to whimper. He was next shown the rat a week later, when he was judged to be more tentative in reaching for it, but seemed happy to play with wooden blocks. Now the blocks were taken away, and the rat was put in front of him and the bar banged five more times, at which point it was discovered that Albert would burst into tears and crawl away from the rat at great speed without any banging. Five days later Albert was brought back and allowed to play with his blocks, which he did quite happily, but every so often the blocks were removed and replaced in turn by the rat, a rabbit, a dog, a sealskin coat, cotton wool, Watson’s hair and a Santa Claus mask. He reacted negatively to all these, although not actually crying in response to Watson’s hair or the cotton wool. In between these now unpleasant stimuli, Albert continued to play with his blocks, and the hair of the assistants who were tried as alternatives to Watson.

What ought to have been concluded from these results? What ought to have been concluded and what Watson did conclude are not necessarily the same thing. It is now being said that Watson should not have concluded anything at all from an experiment on just one child, that Little Albert might have developed an aversion to rats and rabbits even if there had been no enormous iron bar<


being clanged, and so on and so forth (Harris, 1979; Samuelson, 1980). For present purposes, our interest is in the theory that the data were used to support, and not in the reliability of the data. But the reader should note two points about the Little Albert experiment. First it supports a general belief that infants, even of less than I year of age, are not indifferent to their experiences — the child is father to the man — and that nasty experiences, especially, may have peculiar after- effects. But second, the Little Albert experiment certain does not prove, as Watson sometimes implied, that everything which happens to an infant has long-lasting effects, or that all the phenomena observable in infants correspond exactly to the descriptions of the dog’s salivatory reflexes given by Pavlov. In fact, conditioning procedures used with infants less than a year old are not always reliable, although certain standard results, such as anticipatory mouthing at the sight of a feeding bottle, leave little doubt that normal life is producing conditioned effects.

A more benign technique of demonstrating simple conditioning in an infant at about Albert’s age is to clap one’s hands several times over its eyes. This should at first produce blinking in response only to the clap, but after four or five responses the infant should close its eyes if the hands are moved, but stopped just short of a clap (see the chapter on ‘Conditioned reflexes’ in Pieper, 1963, which begins with the proverb ‘A burnt child dreads the fire’). It seems fairly safe to conclude that Little Albert became distressed by the sight of both the rat and the Santa Claus mask because of the experience of being made to jump out of his skin when the rat was previously presented. In Watson’s hands this became a ‘prolific goose for laying golden eggs’ since it provided him, he thought, with ‘an explanatory principle that will account for the enormous complexity in the emotional behaviour of adults’ (Watson, 1931, p. 161). In my view it would be foolish to claim that experiences as extreme as those suffered by Albert are not likely to have some carry-over effects in infants, but even more ridiculous to assume that conditioning is a sufficient explanation for all adult emotions.

The Little Peter treatment

The original plan had been to remove the conditioned emotional responses from Albert before they became chronic, but this was


never carried out (either because of the removal of Albert from the hospital, when he was adopted, or because of the removal of Watson from Johns Hopkins University). Watson speculated that Albert might still have a phobia for sealskin coats when he grew up, and that Freudian psychotherapists might try to cure him by providing him with the insight that his fear of coats was caused by his having been scolded for attempting to play with his mother’s pubic hair in the Oedipal phase. It was Watson’s own view that unpleasant early experiences associated with maternal pubic hair might indeed produce a conditioned fear response which could be ‘transferred’ to fur coats in adult life, and that one of the distinctive things about such internal emotions is that ‘We have never learned to talk about them’ (Watson, 1931, p. 166, original italics).

But partly, I think, because most of his experimental work was done with young children, Watson did not set much store by talking as a method of therapy, and recommended instead more direct methods of treatment. Had Albert continued under Watson’s care, he was due to be exposed to several methods designed to make his fear of rats, rabbits and sealskin disappear. These included constant confrontation with the feared stimuli, the encouragement of manipulation of the feared object by the experimenter setting an example or by the experimenter taking the child’s hand and forcing it to touch the rat, and the ‘reconditioning’ method. The reconditioning was to take the form of presenting the original stimulus (the rat) but calling forth different emotional responses, either by tactile stimulation of the child’s lips, nipples and genitals, or by giving the child food.

After Watson was settled in New York, and working full time on Madison Avenue, he obtained a research grant to continue work on the emotional life of children, and supervised the efforts of Mary Cover Jones (1924a, 1924b, 1974) to alleviate specific fears in children living in a charitable institution. Most of the children studied by Mary Cover Jones were under 5. The ones she was interested in were those who cried and trembled (the fear response) when an animal, such as a frog, rat or rabbit, was shown to them. Individual children often had individual fears, but a jumping frog was the animal most likely to frighten these urban under-5s. Various methods of curing these fears were tried. Simply giving repeated presentations of the same animal, in the


hope that familiarity would breed contempt (the method of disuse) led usually to repeated tears. A 5-year-old girl afraid of rabbits was given a daily dose of pictures of friendly rabbits, plasticine toy rabbits and pleasant rabbit stories. This had the desired effect in so far as the child happily talked about rabbits, and even claimed ‘I touched your rabbit and stroked it and never cried’ (this was not true, but might count today as positive cognitive restructuring). However, when the real rabbit was brought in, she sobbed for it to be taken away, as she had done before. Thus the ‘method of verbal organization’ was discounted. Jones herself was inclined to recommend social methods of treatment (1924b). Small boys afraid of rabbits on their own sometimes developed an interest if shown rabbits in the company of small girls who liked rabbits. Vicarious social learning of this sort has been much studied more recently by Bandura (1973). However, Watson himself was distrustful of the social methods, partly because they could go wrong. Occasionally the social companions, who were supposed to be merely setting a good example, would turn on the infant patient. Little Arthur, afraid of frogs, was not helped when his friends picked up a frog and chased him with it; he screamed and fled. This, Watson said, was liable to breed more negative reactions, not just to the animal, but to human society as a whole. Also, at this young age there was the possibility of social transmission of the phobia — sometimes well-balanced infants, whose presence was supposed to reassure their friend who was afraid of rabbits, themselves joined in the crying.

Watson was much more in favour of ‘direct conditioning’ as used to treat Little Peter. This child was active and eager, at 3 years of age, and judged to be well adjusted, apart from the fact that he was extremely afraid of rats, rabbits, fur coats, feathers, cotton wool, frogs and fish. From this list, Watson said ‘one might well think that Peter was merely Albert B grown up’, but of course the source of Peter’s fear was unknown. Initially Peter fell flat on his back in a paroxysm of fear when a white rat was dropped into his playpen. Then, after a two-month stay in hospital with scarlet fever, he was attacked by a large dog when getting into a taxi to go home. After this all his fears became even worse and the method of direct unconditioning was then begun. It started with the display of a rabbit in a wire cage while Peter was eating his mid-afternoon lunch and ended, forty sessions later, with Peter


eating his lunch with one hand while fondling the rabbit in his lap with the other.

The idea behind the method was that there was a gradual scale of closeness to the fear-provoking stimulus, and that the positive emotions aroused by eating lunch should allow a progression up this scale until even intimate contact with the rabbit took on some of the pleasant associations. The scale had seventeen steps — at first the rabbit was in its cage, which was brought closer and closer each day; then it was free in the room; and eventually it sat on the lunch tray. Six different observers independently rated Peter’s adjustment to the various stages, and it was quite clear that his emotions underwent genuine change. We cannot be so sure about the reasons for these changes. Watson felt they were the direct result of the enjoyment of eating, and that the method would have been even more successful if Peter had been hungrier, or if he had been ‘stroked, petted and rocked’ in order to counteract his fears with sexual feelings. Mary Cover Jones, who performed the therapy, laid more emphasis on social factors of encouragement and imitation. Peter made his most striking advances after particular social incidents. In one case watching another child hold the rabbit led to imitation, and the most dramatic improvement of all occurred after a Dr S, who was a particular favourite of Peter’s, had come along to watch. Peter insisted on calling Dr S ‘my papa’, which suggests a somewhat more complex form of emotional support than digestive contentment. A more straightforward factor was at work when Peter subsequently underwent a relapse, since the relapse occurred just after Peter had suffered a scratch when he was at the stage of helping the experimenter put the rabbit back in its cage. But, this setback was only temporary, and after another dozen lunch sessions, Peter was showing what Jones took to be genuine fondness for the rabbit, was heard to say ‘I like the rabbit’, and allowed it to nibble at his fingers.

There are thus good grounds for claiming that these studies of the Heckscher Foundation children, published by Jones but initiated and supervised by Watson, are the first case histories of behaviour therapy, and the first successes (as well as the first failures) of the application of learning theory to clinical problems. Watson was a wayward and peculiar man, and the swashbuckling success of his behaviourist methods in advertising cigarettes and baby powder may have owed more to his physical resemblance to


Errol Flynn than to the soundness of his psychological principles. But the Heckscher studies anticipated both the method of in vivo desensitization and many later analyses of social learning in children, which are still in use as reliable and safe procedures for the relief of anxiety (Kazdin, 1978).


End of Chaper 7 | Contents | Start of Chaper 8