Learned Helplessness: A Critique of Research and
Kathryn Hahner, Ph.D.
Martin Seligman and colleagues at the University of Pennsylvania were leading investigators of an animal model of human depression known as "learned helplessness." After the definitive studies (e.g., Seligman and Maier, 1967; Overmier and Seligman, 1967), there was a proliferation of research in learned helplessness in animals, as well as research that claimed to demonstrate learned helplessness in humans. In 1978, Seligman and colleagues denounced the animal model of learned helplessness and proposed a "reformulated model" of human depression (Abramson et al., 1978). The term "reformulation" was evidently used to imply that the old learned helplessness research was valuable and relevant to the new model of depression. However, it will be shown that the new human model was not a reformulation of the old learned helplessness model. Seligman and colleagues linked learned helplessness in animals to clinical depression in people with a series of questionable assumptions. Because the "reformulated" model of human depression relied on factors that cannot be studied accurately in animals, the "reformulation" could not have been derived from the animal model. While learned helplessness is still being used as a means to stress animals in fields such as physiological psychology, behavioral pharmacology, and immunology, its original use, as a model of human depression, has been abandoned.
In one of the studies that established the learned helplessness paradigm, Seligman and Maier (1967) divided dogs into three groups. The first two groups consisted of "yoked pairs." That is, one dog of each pair received an electric shock that it could terminate, and the other dog in each pair received the same shock. To this second dog, the shock seemed to stop at random, because it was the first dog that was ending the shock. The dogs with no control over the shocks were said to receive "inescapable shock." The third group of dogs were control subjects who received no shock in this phase of the experiment.
Next, all three groups were tested in a shuttle-box apparatus, in which the dogs could escape electric shock by jumping over a partition. The only dogs that tended to perform poorly in the shuttle-box were those that had received inescapable shock in the pre-treatment phase of the experiment. They did not try to escape, but rather passively accepted the painful shocks. Dogs in the control group, as well as dogs that had been pre-treated with controllable shock, tended to jump over the partition and escape the shocks. Since the dogs which had experienced escapable shock behaved in the same manner as the control dogs, Seligman and Maier claimed to have demonstrated that it was the perceived inescapability of the shocks, and not the shocks alone, which explained the passive behavior.
Overmier and Seligman (1967) addressed the question of whether the inescapably shocked dogs' poor performance in the shuttle-box was due to an interference effect. In other words, had the dogs, while struggling in the harness against inescapable shock, learned to make responses that later interfered with normal escape behavior? In order to rule out the possibility of this interference effect, the authors made sure that the dogs could not move at all as they received inescapable shock. The dogs were immobilized with curare, a paralyzing drug. Later, they showed the predicted "helpless" response in the shuttle-box.
Seligman and colleagues inferred that inescapably shocked dogs learned to be helpless. In other words, according to the investigators, the dogs perceived that outcomes were not related to efforts. Although the mental state of the dogs was difficult to ascertain at best, the researchers readily applied the learned helplessness label to other, very different phenomena. Abramson et al. (1978) provided, as an example of human learned helplessness, a man who is fired from his job and then neglects his household duties.
Buchwald et al. (1978) discussed the confusion over the term "learned helplessness." They identified three different meanings, all of which were used interchangeably by Seligman and colleagues. The first meaning of "helplessness" (from the original dog experiments) was that exposure to inescapable shock resulted in subsequent failure to learn to avoid (escapable) shock. The second meaning of "helplessness" referred to the cognitive and motivational deficits exhibited by the animals. The third meaning was that the animals expected that outcomes were independent of effort (were uncontrollable). According to Buchwald et al., "...it is sometimes not clear which of these three meanings is intended."
The Seligman group believed that learned helplessness in animals was analogous to human depression, but there are other ways to interpret the animals' mental states. Pratt (1980) suggested that Seligman et al.'s dogs acted more like trauma victims than depressed people. A possible analogy is post-traumatic stress disorder, as described in the American Psychiatric Association Diagnostic and Statistical Manual III-R (1987). For example, political prisoners are often immobilized and tortured with electric shocks, much like Seligman et al.'s procedure with dogs. As in post-traumatic stress disorder, the dogs may have been debilitated by anxiety or paralyzed by terror. Another possibility is that the dogs did learn that shocks in the lab were inescapable and that efforts to try to escape subsequent escapable shock were futile. Even if this third explanation were correct, it would have little in common with depression, which is characterized by sadness and pessimism.
Maier and Seligman (1976) listed alternative explanations of the helplessness effect in animals. For example, there was the theory that inescapable shock is an extreme stressor that depletes a neurochemical needed by the animals for movement. The authors attempted to refute the competing theories, but they stated, "It must seem to the reader that a great deal of theoretical confusion surrounds the learned helplessness phenomenon." (1976, p.40). Two years later, Seligman was to abandon the animal model altogether (Abramson et al., 1978). This was a wise decision. Because of the difficulties assessing the animals' motivational and emotional states, there was no way of knowing which explanation was correct.
After 1967, investigators demonstrated the so-called "helplessness effect" in species from fish to cats (see Abramson et al, 1978, for a summary of the literature). Subsequent studies modified and refined the original methodology. The investigators generally assumed that "helpless" animals learned that, since outcomes were independent of effort, it was useless to struggle. This assumption could be seen as the first weak link in a chain from Seligman et al.'s animal model to their human models.
The second weak link was the claim for learned helplessness in human subjects. Seligman and colleagues contended that they had produced a "human analogue to the animal model of learned helplessness" (Abramson et al., 1978). For example, Hiroto (1974) exposed human subjects to loud noises under inescapable conditions. He found that these people, when later exposed to escapable noise, tended to do worse at escaping from the noise than control subjects did. However, it is unlikely that these helpless humans were analogous to the helpless animals. Barber and Winefield (1976) were among those who thought that the helpless humans exhibited "selective inattention." Apparently, the experimental subjects reasoned, "I can't turn off noise in this experiment, so I'll ignore it." Later, accustomed to the loud noise and practiced at ignoring it, they did not attempt to escape from the noise as often as did control subjects. In contrast, it is likely that the helpless animals still found the painful electric shocks highly noxious during both the inescapable and the escapable phases of the experiment. While we do not know the psychological basis for learned helplessness behavior in animals, it is probably not analogous to this "model" in humans.
Seligman and others extended the animal model of helplessness to humans, suggesting that perceived "noncontingency" (i.e. the subject's perception of the uncontrollability of events) is the basis of human depression. In a series of studies in the mid-1970s, investigators (for example, Klein and Seligman, 1976) showed that depressed human subjects perceived noncontingency more often than non-depressed individuals. However, some researchers failed to replicate or get comparable results (Smolen, 1978; Willis and Blaney, 1978).
In 1978, the Journal of Abnormal Psychology devoted an entire issue to learned helplessness research. In this issue, the Seligman group introduced what they called their "reformulated learned helplessness model of depression" (Abramson et al., 1978). They also rejected the animal model of learned helplessness. They stated:
...investigators of human helplessness...have become increasingly disenchanted with the adequacy of theoretical constructs originating iii animal helplessness for understanding helplessness in humans. And so have we. We now present an attributional framework that resolves several theoretical controversies about the effects of uncontrollability in humans. We do not know whether these considerations apply to infra-humans. In brief, we argue that when a person finds that he is helpless, he asks why he is helpless. The causal attribution he makes then determines the generality and chronicity of his helplessness deficits as well as his later self-esteem.
Abramson et al. acknowledged that the animal model of learned helplessness did not account for the observation that depressed people blame themselves for bad events. While their phrase "reformulated learned helplessness model of depression" implies that the decade of animal research had some value, they admitted that the animal model failed to address this fundamental aspect of human depression. Indeed, blaming oneself for bad events seems to be the opposite of learned helplessness, in which "perceived noncontingency" is supposed to account for the passive behavior.
Seligman and colleagues drew their reformulated model not from the learned helplessness paradigm, but rather from that body of literature known as "attribution theory" (e.g., Jones et al., 1972). Attribution theory is concerned with the way people attribute causality to events. For example, is the attribution "internal" (the person is responsible) or "external" (person not responsible)? Also, is the attribution "global" (event seen as typical of life in genera!) or "specific?"
The reformulated model stated that the basis of depression and helplessness deficits is a person's causal attribution to the self for bad events - an internal attribution ("It's my fault"). The model also predicts that depressed people will make more global attributions ("It's like this in every situation"). In addition, depressed people will make more "stable" attributions; i.e., things are seen as always staying the same. Clearly, the attribution theory framework could not be farther from the old animal model. There is no reason to believe that the dogs blamed themselves for the electric shocks. While they were certainly upset by their situation, it is highly doubtful that they felt responsible for it.
The reformulated model essentially proposes that depressives make internal attributions for bad events and are pessimistic about the future. Interestingly, this reformulated model is the same as cognitive models such as that of psychiatrist Aaron Beck (1967, 1974). While Abramson et al. (1978) acknowledged that Beck's perspective was compatible with their own, they did not seem to notice the fundamental identity of the two. Before 1978, Seligman and Beck were often contrasted (Blaney, 1977), because Beck hypothesized that part of the basis of human depression is internal attribution for bad events, whereas helplessness theory saw perceived lack of control as the basis. The only possible conceptual link between the reformulated model and the old model of human helplessness is that the depressed person's pessimism is due to the person's perception that events are uncontrollable. But pessimism is a complex human response and trait and not the automatic result of a person's experience with uncontrollable outcomes. Once again, the link is weak.
Subsequent studies with human subjects by Seligman and colleagues (e.g. Seligman et al., 1988) have produced considerable support for the prediction that depressives tend to make causal attributions that are internal, global, and stable. Despite some conflicting evidence (Patrick and Moore, 1986), it appears that Seligman et al. are finally on firm ground. They got there via a cognitive model developed by Beck and others, a model that was based on research with humans and could not have emanated from animal research.
Seligman and colleagues now appear to be doing valid research (which refines and extends the attribution approach to depression). Unfortunately, they imply that there is continuity between their early helplessness models and the current one. The continuity is, at best, a chain of weak links: First, they assumed that helpless animals perceived noncontingency; next, they supposed that human helplessness was analogous to animal helplessness; next, they proposed that this human model of helplessness was analogous to human depression; finally, they implied that the animal based research has provided a basis for their reformulated model of human depression.
The final leap from learned helplessness to the "reformulated model" is more than a weak link; the chain is broken and a new (that is, Beck's) model adopted. Learned helplessness, as Seligman (1978) correctly pointed out, is a "hypothetical construct." Given the way Seligman and colleagues have invoked and stretched the learned helplessness construct, it probably has low construct validity; in other words, the measurements may be precise, but we do not know what is being measured. This may be the fate of any human construct derived from an animal model.
Abramson, L.Y., Seligman, M.E.P., and Teasdale, J.D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74.
American Psychiatric Association. (1987). Diagnostic and Statistical Manual of Mental Disorders (Third Edition Revised). Washington, D.C.: American Psychiatric Association.
Barber, J.G. and Winefield, A.H. (1986). Learned helplessness as conditioned inattention to the target stimulus. Journal of Experimental Psychology: General. 115, 236-246.
Beck, A.T. (1967). Depression: Clinical, experimental and theoretical aspects. New York:
Beck, A.T. (1974). The Development of Depression: A cognitive model. In R.J. Friedman and M.M. Katz (Eds.) The Psychology of Depression: Contemporary theo,y and research. Washington, D.C.: V.H. Winston.
Blaney, P.H. (1977) Contemporary theories of depression: Critique and comparison. Journal of Abnormal Psychology, 86, 203-223.
Buchwald, A.M., Coyne, J.C., and Cole, C.S. (1978). A critical evaluation of the learned helplessness model of depression. Journal of Abnormal Psychology, 87, 180-193.
Hiroto, D.S. (1974). Locus of control and learned helplessness. Journal of Experimental Psychology, 102, 187-193.
Jones, E.E., Kanouse D., Kelley, H.H., Nisbett, R.E., Valins, S., and Weiner B. (Eds.) (1972). Attribution: Perceiving the causes of behavior. Morristown, N.J.: General Learning Press.
Klein, D.C., and Seligman, M.E.P. (1976). Reversal of performance deficits in learned helplessness and depression. Journal of Abnormal Psychology, 85, 11-26.
Maier, S.F., and Seligman, M.E.P. (1976). Learned helplessness: Theory and evidence. Journal of Experimental Psychology: General, 105, 3-46.
Overmier, J. B. and Seligman, M.E.P. (1967). Effects of inescapable shock upon subsequent escape and avoidance responding. Journal of Comparative and Physiological Psychology, 63, 28-33.
Patrick, L.F., and Moore, J.S. (1986). Life-event types and attributional styles as predictors of depression in elderly woman. Journal of Geriatric Psychiatiy, 19, 241-262.
Pratt, D. (1980). Alternatives to Pain in Experiments on Animals. New York: Argus Archives.
Seligman, M.E.P. (1978). Comment and integration. Journal of Abnormal Psychology, 87, 165-179.
Seligman, M.E.P. and Maier, S.F. (1967). Failure to escape traumatic shock. Journal of Experimental Psychology, 74, 1-9.
Seligman, M.E.P., Castellon, C., Cacciola, J., Schulman, P., Luborsky, L., Ollove, M., and Downing, R. (1988). Explanatory style change during cognitive therapy for unipolar depression. Journal of Abnormal Psychology, 97, 13-18.
Smolen, R.C. (1978). Expectancies, mood, and performance of depressed and nondepressed psychiatric inpatients on chance and skill tasks. Journal of Abnormal Psychology, 87, 91-101.
Willis, M.H. & Blaney, P.H. (1978). Three tests of the learned helplessness model of depression. Journal of Abnormal Psychology, 87, 131-136.