Martin E. P. Seligman, Steven F. Maier and colleagues · 1967–2016

The learned helplessness experiments

If nothing you do changes what happens to you, does that experience make you stop trying later, even when trying would work?

  • University of Pennsylvania, Philadelphia (dog studies, supported by Richard L. Solomon's grants; human study of 1975); University of Portland, Oregon (Hiroto 1974)
  • Replicated, with caveats

The question

In the mid-1960s, in Richard Solomon’s laboratory at the University of Pennsylvania, researchers studying fear and avoidance noticed something unexpected. Dogs that had first received electric shocks they could not avoid often failed, a day later, to learn a simple escape: jumping a shoulder-high barrier to the other side of a box. Most dogs without that history learned it quickly.

Martin Seligman, then a doctoral student at Penn, together with Steven Maier and J. Bruce Overmier, set out to explain it. Was it the shock itself that did the damage? Or something about the experience of having no control over it?

How it was done

Overmier and Seligman (1967) first checked that the effect was reliable. Across three experiments with 88 dogs, a group would receive a series of brief shocks through the back paws while held in a fabric hammock, with no way of stopping them. The next day each dog was tested in a two-compartment shuttle box: a light dimmed, and if the dog crossed the barrier it avoided or ended the shock. They varied the shocks, tested at different delays, and in one experiment kept the dogs immobilised with a drug during the shocks, so that they could not learn any particular movement.

Seligman and Maier (1967) then used what became the classic triadic design, with three groups of eight dogs:

Because the escape and yoked groups received identical shocks, the only difference between them was control. Twenty-four hours later, all three groups went into the shuttle box.

The human studies. In 1974 Donald Hiroto adapted the design for 96 students at the University of Portland. In place of shock, he used a loud tone played through earphones. One group could switch it off with a button; a second could not; a third heard nothing. In the test, they could escape or avoid the tone by sliding a knob across a small box, a hand-sized shuttle box. He also told some students the test depended on skill and others that it depended on chance, and compared “internals” and “externals” on a questionnaire about whether outcomes depend on one’s own actions.

In 1975 Hiroto and Seligman ran four linked experiments with 96 students at Penn. The first task was either the tone (escapable or not) or reasoning problems (solvable or not); the second was either the hand shuttle box or anagrams.

What came out

In the dogs. In the triadic study, the yoked dogs were much slower to escape than the other two groups, and six of the eight never escaped at all. The escape group did about as well as dogs that had never been shocked. Because both shocked groups had received the same shocks, the authors concluded that it was the lack of control, not the shock itself, that produced the failure.

Overmier and Seligman found the same pattern. Only 1 of 8 unshocked dogs (12.5%) never escaped, against 37.5% to 62.5% in the groups that had had inescapable shock. Dogs held still by the drug during the shocks still showed the effect, so the authors argued it was not simply a learned habit of keeping still. The effect was also short-lived: dogs tested 24 hours after the shocks failed often, but those tested after 48 hours or more behaved like dogs that had never been shocked.

In their 1976 review, Maier and Seligman reported that about two-thirds of the roughly 150 dogs given inescapable shock between 1965 and 1969 failed to learn to escape. They also reported that dogs that had first learned to escape were protected against later uncontrollable shock (“immunisation”), and the failure could be undone by forcibly exposing the animal, again and again, to the fact that crossing the barrier ended the shock.

In people. In Hiroto’s study, students who had heard the inescapable tone failed to escape it on over half the test trials, against about 13% for the escapable group and 11% for the group that had heard nothing. Being told the task depended on chance, and scoring as an “external”, each produced a similar slowing. Hiroto and Seligman then found that the effect crossed over between tasks: unsolvable problems made students worse at escaping the tone, and inescapable tone made them worse at anagrams.

The first explanation. The authors concluded that animals and students had learned that outcomes did not depend on their actions, and carried that expectation into new situations: a learned “helplessness”.

What really happened (replications and critiques)

Rival explanations. In the 1970s others proposed simpler accounts: that the animals had learned a movement, such as freezing, that competed with jumping, or (Weiss and colleagues) that severe stress temporarily depleted a brain chemical, noradrenaline, needed for vigorous movement. Maier and Seligman argued against both in their 1976 review.

Limits of the effect. In dogs, the effect of a single session of inescapable shock faded within about two days, though after several sessions it was still present a week later. Rats were at first hard to make “helpless”; the effect appeared only when the escape task was made harder. Hiroto and Seligman themselves cautioned that students saw both tasks as one experiment, so the results might not hold outside the laboratory.

The 1978 reformulation. Abramson, Seligman and Teasdale argued that the original theory did not fit people well. It could not explain why helplessness is sometimes brief and sometimes lasting, sometimes narrow and sometimes broad, or why it sometimes lowers self-esteem. They proposed that what matters is how people explain the lack of control: caused by themselves or by circumstances, lasting or temporary, affecting everything or one area.

The 2016 reversal. Fifty years on, Maier and Seligman reviewed decades of brain research, mostly in rats, and concluded in Psychological Review that “the original theory got it backwards.” Their evidence:

So, they argue, passivity under prolonged stress is the default, not something learned. What is learned is control, and that learning can leave a lasting change that protects against later stress. Work on the same circuit in humans is still limited.

A darker chapter. After 2001 two psychologists, James Mitchell and Bruce Jessen, worked with the CIA on interrogating terror suspects, and a 2015 independent review for the American Psychological Association (the Hoffman report) found that the CIA programme drew on learned helplessness. Seligman met Mitchell and a CIA psychologist, Kirk Hubbard, at his home in December 2001 and again in 2002, and in May 2002 he gave a lecture at a conference of the US Joint Personnel Recovery Agency, he says on how captured Americans could resist torture. He has written that he strongly disapproves of torture and that interrogation was never discussed with him; Hubbard and Mitchell told the review the same. The philosopher Tamsin Shaw has argued that his account is implausible. The review said it would have been difficult for him not to suspect that the CIA’s interest was partly about interrogation, but that it did not have enough information to know what he knew.

Ethics

The animals were adult mongrel dogs, kept in individual cages with free access to food and water. They received repeated shocks they could not control, and in the test many of them did not try to escape further shocks. The 1967 paper by Overmier and Seligman mentions no ethics review.

From the start, the researchers suggested that giving animals control might prevent or reverse the effect, and their later findings on immunisation and recovery pointed that way. But the cost was borne by the animals. Today, research on animals in many countries must replace animals where possible, reduce their number and refine procedures to limit suffering. Under such rules the dog procedures would face very strict review and are unlikely to be approved in their original form; later work used mainly rats.

The human studies used a loud tone that students in the 1974 study had been told was unpleasant but not harmful, and that they rated, on average, near the middle of a seven-point scale. Some students were given problems that had no solution, and some were misleadingly told their success depended on skill. They were paid $2 and debriefed afterwards.

What it means today

The core comparison, the same stress with or without control, remains one of the clearest designs in stress research. The 2016 reappraisal changes the moral: not that failure teaches us to give up, but that under stress holding back comes first, and that real control is what we learn from.

This is a laboratory account, not a guide to anyone’s mood or mental health, and the steps below go beyond what was tested:

Would you have passed?

Coming soon. First comes a short series of puzzles where, depending on the version, your choices do or don’t affect the outcome. Then comes a new puzzle that can be solved. See how quickly you try, compared with the students in 1974 and 1975. One honest caveat: you know this is a game with no real noise and nothing at stake, so it shows the logic of the design, not what it feels like.

Questions people ask

What is learned helplessness?

It is the name Overmier and Seligman gave in 1967 to a laboratory finding: animals that had received shocks they could not control later failed to escape shocks they could have escaped. The original explanation was that they had learned their actions did not matter. In 2016 Maier and Seligman revised this: the brain's default response to prolonged stress is passivity, and what is learned is control.

What was the triadic design?

Three groups of eight dogs. One group could end each shock by pressing a panel. A second, 'yoked' group received exactly the same shocks at the same moments but could do nothing about them. A third group received no shocks. Comparing the first two isolates the effect of control, because the amount of shock was identical.

Was learned helplessness found in humans?

In laboratory tasks, yes. Students who had been exposed to an inescapable loud tone, or given problems with no solution, did worse afterwards at escaping noise and at solving anagrams. The authors themselves noted that students saw both tasks as one experiment, and that they did not know whether the effect carried outside the laboratory.

Did Seligman and Maier later reject their own theory?

In part. In a 2016 paper in Psychological Review, after decades of brain research mostly in rats, they wrote that the original theory got the mechanism backwards. Passivity under prolonged stress is not learned; it is the default. What animals learn, through a circuit in the prefrontal cortex, is that they have control, and that learning can protect them later.

Sources

  1. Seligman, M. E. P., & Maier, S. F. (1967). Failure to escape traumatic shock. Journal of Experimental Psychology, 74, 1–9. doi:10.1037/h0024514
  2. Overmier, J. B., & Seligman, M. E. P. (1967). Effects of inescapable shock upon subsequent escape and avoidance responding. Journal of Comparative and Physiological Psychology, 63(1), 28–33. about.illinoisstate.edu/vfdouga/files/2019/10/overmeier_seligman-_-LH.pdf
  3. Maier, S. F., & Seligman, M. E. P. (1976). Learned helplessness: Theory and evidence. Journal of Experimental Psychology: General, 105(1), 3–46. doi:10.1037/0096-3445.105.1.3
  4. Hiroto, D. S. (1974). Locus of control and learned helplessness. Journal of Experimental Psychology, 102(2), 187–193. www.appstate.edu/~steelekm/classes/psy5300/Documents/Hiroto1974-learned-helplessness.pdf
  5. Hiroto, D. S., & Seligman, M. E. P. (1975). Generality of learned helplessness in man. Journal of Personality and Social Psychology, 31(2), 311–327. www.appstate.edu/~steelekm/classes/psy5300/Documents/Hiroto&Seligman1975-learned-helplessness.pdf
  6. Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87(1), 49–74. core.ac.uk/download/357200055.pdf
  7. Maier, S. F., & Seligman, M. E. P. (2016). Learned helplessness at fifty: Insights from neuroscience. Psychological Review, 123(4), 349–367. doi:10.1037/rev0000033
  8. Seligman, M. E. P., & Shaw, T. (2016, April 21). 'Learned helplessness' & torture: An exchange. The New York Review of Books. www.nybooks.com/articles/2016/04/21/learned-helplessness-torture-an-exchange/
  9. Hoffman, D. H., et al. (2015, July 2). Report to the Special Committee of the Board of Directors of the American Psychological Association: Independent review relating to APA ethics guidelines, national security interrogations, and torture. Sidley Austin. content.news.harvard.edu/gazette/wp-content/uploads/2016/04/apa-final-report-7.2.15.pdf
  10. Seligman, M. E. P. (2018). The Hoffman report, the Central Intelligence Agency, and the defense of the nation: A personal view. Health Psychology Open, 5(2). doi:10.1177/2055102918796192
  11. Senate Armed Services Committee (2008). Inquiry into the treatment of detainees in U.S. custody. United States Senate. www.armed-services.senate.gov/imo/media/doc/Detainee-Report-Final_April-22-2009.pdf