by Morton Hunt
But the strict behaviorist view proved to be an inadequate explanation of much human behavior—and even of some animal behavior. Tolman, though a behaviorist, saw his rats acting at right-left choice points in a maze as if they were remembering, weighing information, and making decisions; and even before midcentury he and other behaviorists were trying to include internal mental processes in the stimulus-response paradigm.
An important effort of this kind was made by two Yale scientists, the sociologist John Dollard and the psychologist Neal Miller, who in the 1940s jointly worked out a theory of “social learning” as an expansion of behaviorism. Under certain conditions, they noted, rats—contrary to Thorndike’s experience with cats—will imitate each other, evidently learning not by means of S-R conditioning but by means of cognitive processes. In human beings, said Dollard and Miller, much learning is social and takes place through high-level cognitive processes as well as drives and needs that underlie motivation.47
From the 1950s on, a number of other behaviorists further developed social learning theory, particularly its cognitive aspects. Central to all versions of the theory is the concept that human personality and behavior are shaped not only by rewarded actions but by the individuals’ predictions or expectations, based on what they have observed, that specific ways of behaving will yield certain rewards. Although this view is much more cognitive than strictly behaviorist, it differs from both trait theory and psychoanalytic theory in that it still sees experiences and situations—external influences—as the major determinants of personality and behavior.
But in the 1950s, a trait-like modification of the social learning view of personality was made by Julian Rotter (1916–1987), then a professor in his mid-thirties at Ohio State University. Rotter was both a psychotherapist and an experimentalist, and although a behaviorist in the laboratory, his experience as a therapist gave him a respect for cognitive processes and emotions that researchers who are best acquainted with mice and rats often lack. Like most other clinicians, Rotter had found that often his patients’ basic attitudes toward life had been formed by critical experiences, some good and others bad. Recasting this in behavioral terms, he theorized that when particular acts are either rewarded or not rewarded, people develop “generalized expectancies” about which kinds of circumstances and behaviors will or will not be rewarding.48 A student who studies diligently and gets good grades, wins praise, and feels good about himself may come to expect that hard work in other situations will be similarly rewarding; a student who studies hard but fails to get good grades and their associated benefits may come to expect that in general hard work does not pay off.
Rotter and his graduate students conducted a series of experiments that demonstrated the pervasive influence of such generalized expectations. In a typical study, he or his collaborator would tell the volunteers—undergraduate men and women at OSU—that they were being tested for ESP. (This was a cover story to camouflage what they were really doing.) The experimenter would hold up, with its back toward the volunteer, a card on which there was either a square or a circle, the volunteer would guess which it was, and the experimenter would say he or she was either right or wrong. After a set of ten, he’d ask the volunteer to estimate how many he or she would get right of the next ten. Some students regularly estimated that they would do worse the next time because, as they later revealed in questionnaires and interviews, they ascribed their right guesses to luck. Others estimated that they would do better the next time because they attributed their right guesses to their skill at ESP, which they expected would increase with practice.
At about the same time, Rotter was supervising a psychotherapist in training, E. Jerry Phares, one of whose patients was a single man in his twenties who complained of having no social life. Phares urged him to attend a free dance held on campus. He did so, and several girls danced with him, but he told Phares, “It was just lucky—it would never happen again.” When Phares reported this to Rotter, it crystallized an idea he had been forming. Reminiscing some thirty years later, he recalled that moment:
I realized there were always some subjects in our experiments whose expectancies, like this patient’s, never went up even after successes. My graduate students and I had run various experiments in which we rigged the volunteers’ success or failure—we did so in the ESP series, and also in an angle-matching test in which we could control the number of supposedly “right” or “wrong” responses because the angles were so close that they looked alike and the volunteers would believe whatever we told them. Some volunteers, whether we told them they were right or wrong most of the time, didn’t change their expectation that they’d get most of them wrong on the next set. Others, whatever we told them, thought they’d do better the next time.
At that point I put together the two sides of my work—as practitioner and as scientist—and hypothesized that some people feel that what happens to them is governed by external forces of one kind or another, while others feel that what happens to them is governed largely by their own efforts and skills. Phares and I then worked out a test to measure the degree to which any individual perceives reward or the lack of reward as the result of his own behavior or as having nothing to do with it.49
Rotter called this crucial attitude—the major discovery of his career— “locus of control.” The test he and Phares developed to measure it, the Internal-External (I-E) Locus of Control Scale, is made up of twenty-nine items, each of which comprises two statements; the person taking the test says which of each pair of statements seems more true to him. Some typical items:50
2. a. Many of the unhappy things in people’s lives are partly due to bad luck.
b. People’s misfortunes result from the mistakes they make.
4. a. In the long run people get the respect they deserve in this world.
b. Unfortunately, an individual’s worth often passes unrecognized no matter how hard he tries.
11. a. Becoming a success is a matter of hard work; luck has little or nothing to do with it.
b. Getting a good job depends mainly on being in the right place at the right time.
25. a. Many times I feel that I have little influence over the things that happen to me.
b. It is impossible for me to believe that chance or luck plays an important part in my life.
Choices 2a, 4b, 11b, and 25a indicate that the respondent feels he or she has little control over events, the others that the respondent feels in charge of his or her own life. People who score high on external locus of control tend to attribute their successes and failures to fate, luck, or the power of other people; people who score high on internal locus of control attribute their successes and failures to their own intelligence, hard work, or other personal traits. Locus of control, a generalized attitude affecting many aspects of personality and behavior, is thus like a “central trait” in Allport’s scheme and a “source trait” in Cattell’s.
The concept of locus of control and the I-E scale struck a responsive chord among personality psychologists. Since the scale appeared in 1966, some two thousand studies using it have been published, and for at least two decades it remained one of the more popular personality tests. Then it was largely supplanted by other and more sophisticated tests, but the locus of control concept has remained a staple in personality assessment.51 Many of the research studies using the I-E Scale showed how locus-of-control expectations affect behavior. For instance, elementary school students who scored as Internals got higher grades, on the average, than Externals; “helpless” children (Externals) did worse after failing a test containing difficult problems, “mastery-oriented” children (Internals) tried harder and did better. In experiments where volunteers were confronted with a dilemma, Internals were more likely to seek useful information, Externals to rely on others to help them. Among people hospitalized with TB, Internals knew more about their illness and asked more questions of their doctors than Externals. Internals brushed and flossed their teeth more than Externals. Internal
s were more likely than Externals to use seat belts in automobiles, get preventive shots, engage in physical fitness activities, and practice effective birth control.52
On the negative side, some studies have found that Internals are less likely than Externals to sympathize with people in need of help, since Internals believe that needy people brought their troubles on themselves. And although Internals feel proud when they succeed, they feel ashamed or guilty when they fail; Externals, in contrast, feel less strongly about either success or failure. (Normal and healthy personalities, some researchers believe, are balanced between Internal and External and explain their lives to themselves in a self-protective, if inaccurate, way: they tell themselves, the social psychologist Fritz Heider has said, “I caused the good things; the bad things were forced upon me.”)53
Social learning theory and locus-of-control research led to some notable developments in personality theory and clinical psychology. One was a growing recognition that conscious attitudes and ideas, not just unconscious ones, account in considerable part for the individual’s traits and actions. What the psychologist George Kelly called “personal constructs”—sets of conscious ideas about one’s own abilities and character, the behavior other people expect of us in various situations, how others are likely to behave in response to us, what they mean by the things they say, and so on—are important determinants of personality and behavior.54
Research based on this view has produced such interesting findings as the laboratory demonstration in 1978, by Edward E. Jones and Steven Berglas, of the ego-protecting tactic they labeled “self-handicapping.”55 Self-handicappers, faced with a situation in which they fear they may fail, protect their self-esteem by arranging things so that other people will attribute their failure to forces they could not control. A mediocre tennis player may choose only partners distinctly better than he so that losing will do him no discredit; a student about to take a final exam may burden himself with campus duties when he should be studying so as to have an ego-defending excuse if he performs poorly. The self-handicapper defeats himself in the effort to protect himself.
A particularly noteworthy byproduct of locus-of-control theory was the elucidation of a disabling phenomenon known as “learned helplessness.” We have all known people who, hopeless and passive, are unable to make the effort to cope with their problems even though they possess sufficient abilities and resources. Many clinicians had offered conjectures about the cause of such passivity, but in 1967 Martin Seligman, a twenty-one-year-old graduate student at the University of Pennsylvania, had a flash of insight that led, after many years of work, to valuable understanding of such passivity.
Seligman had just gone to his professor’s laboratory for the first time and found him and his graduate assistants deeply troubled. Their experimental dogs wouldn’t perform. The dogs had been conditioned by tone and shock given together until they associated the tone with the shock. Now they were in a “shuttlebox,” a large cage with a low fence dividing it into two compartments, where they were being subjected only to the tone. When dogs are put in such a box and given a shock in one compartment but not the other, they rapidly learn to leap the fence to escape the shock; the experiment was intended to find out whether they would do the same when they heard the tone without the shock. But at the tone, these dogs lay still and whimpered. No one could understand it, but young Seligman had a sudden thought. While the dogs were being conditioned to tone and shock, they couldn’t escape the shock; they had learned that nothing they did mattered. Now, in a situation where they could have escaped the shock, they still acted as if nothing they did would help.56
With a fellow student, Steven Maier, and later with another colleague, H. Bruce Overmier, Seligman conducted a series of experiments in the creation of learned helplessness. A central experiment had the dogs placed, one at a time, in a cage where, harnessed and unable to escape, they received a series of electric shocks to their feet through the metal floor. Then each dog, and a number of others that had not had the shock treatment, were put in a shuttlebox where, from time to time, a light would go on in the compartment the dog was in, followed in ten seconds by a shock. All the animals quickly associated the light with an imminent shock; when it went on, the untreated dogs scrambled about wildly and soon found that they could escape the shock by jumping over the fence into the other chamber, but the dogs that had been subjected to unavoidable shock stayed put and let themselves be shocked without making any effort to escape. They had developed the expectation that nothing they could do would avoid the shock; they had learned to be helpless.57
That seemed to explain learned helplessness in human beings as well as dogs. But Overmier and Seligman went further. Depression in human beings, they boldly suggested, might often be due less to actual inability to cope with problems or sorrows than to learned helpless-ness—the feeling or belief that there is nothing to be done. This theory was immediately rebutted by psychologists and psychiatrists, who pointed out that some people never become helpless when bad things happen to them; some do but bounce back rapidly; some become helpless not only in the given situation but in new and different ones; some blame themselves and some blame others for their misfortunes.
Seligman, in collaboration with one of his critics, a British psychologist named John Teasdale, and another colleague, set out to find a better explanation of human depression. They worked out a new hypothesis combining learned helpless and locus of control. When human beings have painful experiences they can do nothing about, they either interpret them as the result of external forces or blame themselves, and the erroneous latter interpretation induces depression.58 The team tested the hypothesis by means of a complicated locus-of-control questionnaire; the information supported the hypothesis, and after their study was published in 1978, a rash of similar and confirmatory studies—more than three hundred in the next twenty years—with dogs, rats, and people confirmed and extended it.59
One such study, for instance, rated a group of pregnant women, on the basis of personality testing, as either Internalists or Externalists, and found a markedly higher rate of postpartum depression among the Internalists. These women blamed the difficulties of the period on their personal attributes; the Externalists blamed them on the situation and, though feeling helpless, were not particularly depressed.60
Later, Seligman broadened his theory into what he called “explanatory style.” It accounts for the basic aspect of personality that appears as overall optimism or general pessimism. In Seligman’s words:
Take a bad event such as a defeat in business or love, for instance. Pessimists attribute it to causes that are long-lasting or permanent, that affect everything they do, and that are their own fault. Optimists regard the causes of a defeat as temporary, limited to the present case, and the result of circumstances, bad luck, or the actions of other people.
Optimism leads to higher achievement than pessimism. Optimistic life insurance agents, we found, sell more insurance and stay at it longer than pessimistic agents. Optimistic Olympic-level swimmers, when they’re defeated, get faster; pessimistic swimmers, when they’re defeated, get slower. Optimistic professional baseball and basketball teams do better, particularly after they’re defeated, than pessimistic teams.61
On this basis, Seligman developed his own distinctive approach to personality: helplessness or pessimism could be learned (as the dogs had learned it)—but so could the opposite, optimism; one could learn mental skills that would change one’s view of life in a positive and self-directing fashion. “Learned optimism” became the basis of what he views as a new discipline, Positive Psychology, which studies positive emotion, positive character traits, and therapeutic methods of achieving them. Since 2000 this has been Seligman’s main interest, and at the University of Pennsylvania he now heads a training and research facility known as the Positive Psychology Center.62
One other issue on which social learning theory cast new light is that of the differences between male and female persona
lities. Ostensibly wise persons have had a lot to say about this subject throughout the ages, most of them men, who have spoken well of their own sex and ill of the other. Their views have ranged from Plato’s mild denigration of woman (“The gifts of nature are alike diffused in both [sexes], but in all of them a woman is inferior to a man”) to Clement of Alexandria’s tirade against woman’s sinful nature (“You are the gate of hell, the unsealer of that forbidden tree, the first deserter of the divine law”) to Lord Chesterfield’s genteel sneer at woman’s mind and personality:
Women are only children of a larger growth; they have an entertaining tattle, and sometimes wit; but for solid, reasoning good sense, I never in my life knew one that had it…A man of sense only trifles with them, plays with them, humors them, and flatters them, as he does with a sprightly, forward child.
A number of traditionally female traits—emotionality, timidity, vanity, nurturance, perceptiveness, deviousness, and so on—have always been assumed to be innate. In the early decades of psychology, most psychologists, Freud among them, believed that these traits were the inescapable outcome of women’s hormonal and biological endowments and of the special experiences these brought about. As late as 1936, Lewis Terman and a colleague, C. C. Miles, published a well-received and influential study of male and female personalities, Sex and Personality, based on the results of a test they had constructed. Many of the answers were scored on the basis of traditional beliefs about gender differences. In the word-association section of the test, for instance, if the respondent’s association to “tender” was “meat,” the answer was scored as masculine, if “kind” or “loving,” feminine. Reading detective stories and liking chemistry were scored as masculine; reading poetry and liking dramatics, feminine.63