Clark L. Hull, Hypnotist

A review of

[book cover]

Hypnosis and Suggestibility: An Experimental Approach
by Clark L. Hull

New York: Crown House 2002. 416 pp. ISBN 1-899836-93-4. $44.95

Reviewed by
John F. Kihlstrom

Kihlstrom, John F. S-1 #1 Clark L. Hull is known to most psychologists for his enormously influential theory of learning. But in the 1920s and early 1930s, long before he promulgated the formula SER = SHR × D, Hull embarked on a systematic program of experimental research on hypnosis. Many of the individual studies, chiefly conducted by undergraduates at the University of Wisconsin for their senior theses, were published in scholarly journals—more than 30 of them—in most cases with the student as the sole author (an act of generosity few of us would emulate today). In 1933, Hull published a comprehensive account of the entire project in the prestigious Century Psychology Series. Notwithstanding its tables, graphs, and high-level scientific vocabulary, the monograph was “intended to make this body of experimental material available to the general public” (p. ix). In fact, the book was very well received, and remained in print long after Hull's death, and has since inspired several generations of hypnosis researchers. Now it has been reissued in a handsome new edition with an introduction by Michael D. Yapko, a distinguished hypnotherapist.

Kihlstrom, John F. S-1 #2Exactly how Hull became interested in hypnosis is not clear (Hull,Kimble, 1991). His graduate research on concept learning had proved that a rigorously controlled, quantitative science could be made out of the study of thought processes, just as Fechner and Ebbinghaus had done in the 19th century for sensation and memory. But the field was not ready for it, and the work fell flat, not to be revived until the cognitive revolution after World War II. Hull also did research on the effects of tobacco smoke on human performance—in the process, inventing the active placebo control. Hull then turned his attention to aptitude testing, but the work quickly bored him, though not before he invented a machine for calculating correlation coefficients from data coded on punched paper tape that is on display in the Smithsonian Institution. Nothing in any of this work betrays any interest in hypnosis.

Kihlstrom, John F. S-1 #3Hull might have been introduced to hypnosis by Joseph Jastrow, first his graduate mentor and then department chair, who had included the topic in his book on unconscious mental life. But in fact, Jastrow never demonstrated hypnosis to him, and Hull had never observed hypnosis at all until a student asked him to try it to rid the student of a phobia (Hull, 1952). Swinging a “hypnotic crystal,” which he had received as a gift from a student, and reading instructions from a book, Hull succeeded. Hull included hypnosis in his lectures to Wisconsin undergraduates and medical students as early as 1921 (Triplet, 1982), and he ran a graduate seminar on the topic as well: One of his early students was Milton Erickson, an undergraduate at University of Wisconsin who later took medical training and became famous as a hypnotherapist. But there is no reference to hypnosis in the published excerpts from the “idea books” that Hull had kept since his college years (Hull, 1962) until early (probably January) 1927, when he listed “a volume on hypnosis” (p. 822) as his top-priority project; and not until January 1928 did he propose an “intensive series of investigations on hypnotic phenomena” (p. 824) to be carried out over the next two years, pointing toward completion of the hypnosis book in 1929.

Kihlstrom, John F. S-1 #4The answer to the question of origins appears to be, quite simply, that Hull was intrigued by hypnosis, just as James and Pavlov had been, and he thought it would serve as a kind of proving ground for the new experimental psychology that he and his fellow behaviorists were trying to promote. If scientific psychology were worth its salt, he seems to have thought it ought to be able to tell us something about interesting “higher mental processes” (Hull,Schneck, 1972; IWeitzenhoffer, 1972), and he even developed a mechanical device for objectively recording the degree of response. Today we recognize that proper assessment requires sampling from a wider domain of suggestions, including cognitive as well as motor items and, in each domain, items involving both positive and negative suggestions. Multi-item hypnotizability scales were introduced only after Hull had completed the major portion of his research program, culminating in the “Stanford” and “Harvard” scales that are the gold standard for hypnosis research today. It is interesting that the first of these scales was developed by Davis and Husband at Wisconsin, but apparently independently of Hull's group. Hull commented favorably on the Davis-Husband scale in his book, but it appeared too late for him to use it in his own research. Still, it is not clear that any of his experimental findings would be changed if he had.

Kihlstrom, John F. S-1 #5Finally, there is little inquiry into the subjective experience of hypnotic subjects, as they respond to the hypnotist's suggestions. Today we understand that subjective experience is central to hypnosis. The point of hypnotic age regression is not to behave like a child; it is to feel like a child. Hull's preference for behavior over self-report was mandated by his goal of making psychology an objective science, but it led him to misinterpret some of his findings. He thought he had disproved Janet's dissociation theory of hypnosis by showing that posthypnotic amnesia did not affect retroactive inhibition (some investigators hold similar views today). When he found that posthypnotic amnesia did not reduce practice effects or savings in relearning, he concluded merely that the forgetting was “by no means complete” (Hull, 1933, p. 138). And when he observed that hypnotic analgesia and anesthesia did not eliminate reflexive responses to stimuli, he concluded simply that the suggestions affected voluntary but not involuntary behavior. Now we understand that dissociation is not about interference between divided streams of consciousness, but rather about awareness of the contents of the dissociated stream. And we understand that retroactive inhibition, practice, and savings effects reflect a dissociation between explicit (conscious) and implicit (or unconscious) memory of the prior learning experience, just as psychophysiological variables such as heart-rate variability and galvanic skin response reflect the dissociation between explicit and implicit perception of the pain stimulus.

Kihlstrom, John F. S-1 #6Hull was a methodological behaviorist, but he was not a radical like Watson or Skinner, who sought to abolish mentalistic thinking from psychology. For Hull, the task of psychology was not merely to describe behavior, but rather to describe the internal mental processes that caused voluntary behavior to occur. Volition, or will, was central to his view of behavior, and suggestion—hypnotic or not—was central to his analysis of volition. Hull embraced William James's theory of ideomotor action, and like James he thought that hypnosis was a particularly good place to study how ideas were translated into action. In hypnosis, Hull could trace the source of these ideas to the hypnotist's suggestions, which were then taken up as autosuggestions, but he is clear that the critical processes are internal, rather than external. As Triplet (1982) has noted, the theme of ideomotor action was carried into Hull's later theory of learning, which was really intended to be a theory of complex behavior of all sorts and which Hull was already developing before he finished the hypnosis book. Moreover, Williams (1953) cites a comment by the British sexologist Havelock Ellis, to the effect that someone should do for psychoanalysis what Hull had done for hypnosis, as the genesis of the interdisciplinary investigation of psychoanalysis pursued at the Institute for Human Relations that led to the development of social learning theory. Rather than being a topic Hull stumbled on for want of anything else to do, and later abandoned for the greener and safer pastures of learning theory, in this view the work on hypnosis was just one step on the way to a general theory of complex human behavior—a journey that began with the work on concept formation and ended with the theory of habit and drive.

Kihlstrom, John F. S-1 #7Hull was not the first to do experiments on hypnosis. William James, who generally loathed experimentation, loved to tinker with hypnosis, and James's colleagues at Harvard, Morton Prince and Boris Sidis, reported some interesting findings around the turn of the century. More to the point, P. C. Young completed an extensive series of experimental investigations for his 1923 doctoral dissertation at Harvard. Still, Hull's research was the first to use control groups, and it set the standard for future experimental work. Hull abandoned hypnosis research, but not before outlining 40 experiments on nonhypnotic suggestibility (Hull,Hull, 1930a, IDb), for others to pursue. Hull had actually debated whether the latter paper should include 100 research projects or merely 99 (Hull, 1962, p. 823). In the time since, some of the studies Hull proposed have actually been carried out, not always with attribution; many of the others remain as interesting today as they were more than 70 years ago. Hull was correct to write that “the possibilities of research in hypnotism not only are not exhausted but that, comparatively speaking, they have as yet scarcely been touched” (Hull, 1930a, p. 200).

Kihlstrom, John F. S-1 #8With so much left to do, why did Hull give up hypnosis? For one thing, some authorities at Yale's medical school were concerned about the dangers of hypnosis research (concerns we now know were invalid) and put severe restrictions on Hull's ability to recruit new subjects in New Haven to replace the ones he left behind in Madison. The situation was worsened when one of Hull's subjects, described as mentally disturbed by Williams Hull, 1962, p. 852). The writing also went badly. Hull had originally intended to publish a 200-page book in 1929, but in 1931 he referred to “the terrible chapter on waking suggestion” (Hull, 1962, p. 848) and in 1933 to “the terrible book on hypnosis and suggestibility” (Hull, 1962, p. 849), which eventually contained more than 400 pages. Throughout this long period of gestation, Hull fretted that a rival might claim credit for “the new birth of scientific hypnosis” (Hull, 1962, p. 831; the competitor's name was unfortunately redacted from the published version of the idea books, but was possibly W. R. Wells, P. C. Young, George Estabrooks, or even Hull's former student, Milton Erickson; in any event, none of them produced a book to rival his). Hull's final assessment of the book is a mix of pride and regret:

I believe that it is an important contribution, that it may mark the beginning of a new epoch in that form of experimentation, and that it will be read and quoted for a long time, possibly a hundred years. At all events it probably will be read after the work of those here at Yale who have thrown obstacles in the way of the experimental work upon which it is based, has long been forgotten. But even if all this should take place, I have paid a high price and would hardly do it again. (Hull,Hull, 1962, p. 823), and to a considerable extent he succeeded, spawning generations of “neoHullian” and social learning theorists, including John Dollard, Neal Miller, Hobart Mowrer, Kenneth Spence, Abraham Amsel, and Frank Logan. Nevertheless, his formal theory of learning—weighed down as it was with all those postulates and formulas—fared poorly in the competition with Skinner's radical behaviorism, and it did not long outlive its author. By contrast, Hull's hypnosis book is still going strong, more than 70 years after its publication—something that a new generation of hypnosis researchers, like the several that went before, can read with awe and respect.


Hull, C. L.. (1929). Quantitative methods of investigating waking suggestion. Journal of Abnormal & Social Psychology, 24, 153-169.

Hull, C. L.. (1930a). Quantitative methods of investigating hypnotic suggestion. Part 1. Journal of Abnormal & Social Psychology, 25, 200-223.

Hull, C. L.. (1930b). Quantitative methods of investigating hypnotic suggestion. Part 2. Journal of Abnormal & Social Psychology, 25, 390-417.

Hull, C. L.. (1933). Hypnosis and suggestibility: An experimental approach. New York: Appleton-Century-Crofts.

Hull, C. L.. (1952). Clark L. Hull. In E. G. Boring, H. S. Langfeld, H. Werner, & R. M. Yerkes (Eds.), A history of psychology in autobiography (Vol. 4, pp. 143-162). Worcester, MA: Clark University Press.

Hull, C. L.. (1962). Psychology of the scientist: IV. Passages from the “idea books” of Clark L. Hull. Perceptual & Motor Skills, 15, 807-882.

Kimble, G. A.. (1991). Psychology from the standpoint of a mechanist: An appreciation of Clark L. Hull. In G. A. Kimble, M. Wertheimer, & C. L. White (Eds.), Portraits of pioneers in psychology (pp. 209-225). Washington, DC: American Psychological Association.

Schneck, J. M.. (1972). A historical comment on Clark Hull, his postural sway test for suggestibility, and earlier descriptions of the maneuver. International Journal of Clinical & Experimental Hypnosis, 20, 25-30.

Triplet, R. G.. (1982). The relationship of Clark L. Hull's hypnosis research to his later learning theory: The continuity of his life's work. Journal for the History of the Behavioral Sciences, 18, 22-31.

Weitzenhoffer, A. M.. (1972). The postural sway test: A historical note. International Journal of Clinical & Experimental Hypnosis, 20, 17-34.

Williams, G. W.. (1953). Clark L. Hull and his work on hypnosis. Journal of Clinical & Experimental Hypnosis, 1, 1-3.

CLARK L. HULL, deceased.

JOHN F. KIHLSTROM, Department of Psychology, University of California, 3210 Tolman Hall, Berkeley, CA 94720-1650. E-mail:

0010-7549© 2004 by the American Psychological Association
Previously published in Contemporary Psychology: APA Review of Books, April 2004, Vol. 49, No. 2, 141-144

For personal use only--not for distribution.