Senoi Dream Theory: Myth, Scientific Method, and the Dreamwork Movement

G. William Domhoff

March, 2003

• Chapter 1: Introduction and Overview
• Chapter 2: What Do We Know About the Senoi?
• Chapter 3: The Life and Mind of Kilton Stewart
• Chapter 4: Why Did Stewart's Ideas Have a Strong Appeal?
• Chapter 5: Does Senoi Dream Theory Have Any Validity?
• Chapter 6: What Should We Make of All This, If Anything?

Chapter 5: Does Senoi Dream Theory Have Any Validity?

The fact that Stewart's theories of dream sharing and dream control are not practiced by the Senoi does not invalidate his ideas. Nor does his sloppy thinking and dubious credibility as an anthropologist and psychologist mean that the ideas are wrong, although his tendency to fabricate should lead people to be skeptical. The fact that his ideas were embraced uncritically by an eager new generation of ever hopeful Americans is interesting in and of itself, but that is neither here nor there as far as the validity of the ideas. It may be that sharing dreams is beneficial for individuals, groups, or societies. It may be that Stewart's principles of dream control are useful in ridding dream life of the aggression and negative feelings that predominate over friendliness and positive feelings for most dreamers, as has been found in dream reports from all over the world studied systematically using the Hall/Van de Castle coding system.[1]

The ideas put forth by Stewart are worthy of investigation whatever their origins or appeal. Even those who agree with me that the anthropological, biographical, and sociological findings raise serious questions ought to concede that we owe it to these ideas to give them a fair hearing on the evidence. Thus, in this chapter I will assess the independent evidence that is available on the efficacy of sharing and shaping dreams. This evidence comes from classrooms, dream groups, and experimental studies.

There is not much systematic evidence on the usefulness of dream sharing, but the little that exists is interesting. In particular, the idea that the sharing of dreams can lead to creativity and social harmony in small groups received support in a classroom exercise carried out over a period of several months in 1961 by a future grade school teacher working on her M.A. degree in psychology. Based in part on her knowledge of Stewart's work, but also drawing on her reading of experts on creativity, Elena Werlin spontaneously asked her students early in the school year if they would like to share their dreams:

The children were all sitting at their desks waiting for me to tell them what the next activity would be when suddenly it came to me, in a playful-serious way, "What about asking them if they would like to tell their dreams?" Originally I had been planning to have them draw, or listen to a story. But then my old conviction about the value of telling dreams began bubbling and before I knew it I was asking the children if they would like to tell a dream. I think at that moment my curiosity about the validity of my readings concerning the preconscious and creativity just got the better of me.[2]

Werlin then kept a classroom log over a six-month period, in which she reports on the effects of dream sharing among twenty-five first-grade children. Although there was "no wild enthusiasm" for the idea of "dream time" in the first few weeks, the children were often eager for these sessions after a few months, and many sessions seemed to result in reduced tensions in the classroom or heightened creativity for some individuals. However, it was not only dreams that were shared, but as often daydreams or made-up stories. Nor was any attempt made to interpret the dreams. In effect, she found that the surest way to discourage children from sharing their dreams is to interpret them. Children don't seem to find such a task very interesting.

Instead, the dreams were seen by Werlin and the children simply as a means of communicating personal thoughts or as a basis for stories or drawings. For these purposes the children often related to the dreams they told and heard in a very constructive way. At the same time, Werlin felt that she gained an awareness of aspects of the children's personalities that she would not have been aware of if she had not instituted dream sharing:

During all times of the day except dream time, Brenda and Robin and Patty pull in their feelers and become nearly completely withdrawn. But even if these quiet children are reticent for the rest of the day, the fact remains that they have participated in dream time and as a result a feeling of warmth and familiarity is able to grow between us.[3]

Werlin did not do any further work with dream sharing in the classroom. She told me that larger class sizes and the pressures to meet specific reading and arithmetic requirements rendered the open-ended sessions unfeasible, especially because the children often urged that dream discussions be continued beyond their allotted time. The same factors probably explain why nobody else apparently tried the idea.

There are no comparable reports on the effects of dream sharing among older students, adults, or therapeutic groups. Many such dream groups have existed, but they go unreported or are discussed in a brief fashion with a very few examples. One of the few published reports in the journal literature provides three dreams that were elaborated or completed by dream groups. This 1973 article begins with the following introductory comment:

For the past three years I have conducted classroom and therapeutic groups based on a combination of these principles. Here I will discuss the organization of work in such groups and some instances of successful use of dream materials in effecting personal change and group cohesiveness.[4]

Although those who direct dream groups of course say that dream sharing is useful, they do not show that there is anything unique about dream sharing. However, contrary to their emphasis, the general conclusion that can be drawn from a wide range of studies of personal sharing in small discussion groups is that any personal disclosure in an open and supportive atmosphere is found by some participants to be personally helpful and conducive to group cohesiveness.[5] Viewed from this perspective, dream sharing may be useful because it is one of many intimate disclosures that can have positive effects.

Even here, however, there is need for caution. It may be that people feel enthusiastic about sensitivity groups, therapy groups, and dream groups even when there is no objective evidence that anybody or anything changed. As a skeptical clinical psychologist suggested on the basis of a detailed survey of the literature on psychotherapy, supplemented by his own interviews with therapists and former patients, there is a strong tendency to emphasize the positive and ignore the negative even when the group or individual therapy has not reached its stated goals. First of all, people are pleased to find they have been able to say personal or shocking things without being censured. For example, to admit to a murderous or deceitful dream action and not be scorned can lead to a sense of relief. Second, they are glad to know they are not alone in their problems, which they learn when other people say they have had somewhat similar dreams. Third, people have a strong will to believe that they must be getting something out of an experience to which they have committed time, effort, and probably money. Finally, and most subtly, therapists and group leaders have ways of suggesting that the responsibility for success or failure lies with the patient or group member, so people are reluctant to admit to negative outcomes.[6]

Still, it seems that dream sharing could be a somewhat unique kind of sharing because a dream is at once so personal and yet not seen as something for which the person is responsible. In some languages people say "It dreamed to me," and English speakers in effect express the same distance when they say things like "I had this dream." A dream group leader of the 1970s expressed the point very well:

What puzzles is that dreams, unlike other intimate communications, are unlikely to meet with criticism, ridicule or shocked surprise when told to strangers. Members of a dream group would no sooner call another's dream foolish than they would say having brown eyes was foolish. There is that "givenness" about dream material.[7]

Based on the peculiar status that dreams enjoy as personal statements for which people do not feel responsible, further studies of their usefulness in groups could be conducted within the framework of the intimacy and self-disclosure literatures. These studies could directly compare dream sharing and other types of disclosures within experimental groups on the variety of dependent variables that are used in sensitivity studies, including individual ratings of others in the group and experimenter ratings of individual reactions and group interactions. Studies of this kind are not likely to determine whether or not dream sharing could lead to the larger, societywide harmony that Stewart hoped for, but they would be a starting point on the personal and social usefulness of this idea. However, in the years since I first made this point in 1985, there has been no effort on the part of anyone in the dreamwork movement to make any form of systematic study of what they do.

I turn now to the small literature on shaping and controlling dreams, only part of which derives from Stewart's claims. First, there is some evidence that people can dream about a topic if they are encouraged to do so, but it should be noted at the outset that dreaming about a suggested topic is one thing and controlling dream content is another. In the first of three studies at the University of Brunswick in the 1970s, for example, sixteen students were asked by eight of their friends to dream about one of four suggested topics on specified nights. They were instructed to set their alarm clock for 6:30 A.M. and to record any dreams they remembered. The experiment lasted for three consecutive nights. Ten of the nineteen reported dreams contained suggested topics. The two pleasant topics (riding a bicycle, going fishing) were incorporated more frequently than the two unpleasant topics (a car accident, a world war erupting).[8]

The second study included seventy students over five nights who were asked to dream about eating a favorite food. which led to similar positive results. In addition, it was found that those who were best able to dream about the suggested topic had a more positive attitude toward dreaming, a higher frequency of dream recall, and a greater tendency to be introspective.[9] The third study involved twenty-four students who slept two nights in a laboratory setting. Eight of the participants were asked to dream about eating a favorite food, eight were asked to dream about eating an unpleasant meal that made them sick, and eight were given no instructions. The subjects were awakened four times during the night. Seven of the eight subjects who received the positive suggestion reported themes of eating, but only three of the eight who received the negative suggestion did so. (But four of the eight people who received no suggestion dreamed about eating anyway, so the results are not conclusive by any means with such a small sample).[10]

Still another study showed the incorporation of pre-sleep suggestions, but no evidence that the content could be controlled. Inspired in part by Stewart's writings, the researcher asked seventeen college participants to list personal traits they would like to change and then urged them to have positive dreams about one of these traits in the sleep laboratory. They were told to wish over and over again as they fell asleep for the change they wanted to make; for example, "I wish I were not so hostile, I wish I were not so irritable, I wish I were more poised." Fifteen of the seventeen participants reported at least one instance of apparent incorporation of the target trait when they were awakened in the laboratory. However, only two of the dreams incorporated the trait in the positive way that was wished for. Instead of having positive incorporations, more often than not the participants incorporated the negative trait and seemed "to be getting some gratification out of maintaining it."[11] In the example that is provided, a participant who wished to be less sarcastic had two different dreams in which she was very sarcastic.

Another piece of seemingly positive evidence from the laboratory turned out to be inconclusive. This second study first came to public attention as one that seemed likely to lead to positive results. In "Happy Endings for Our Dreams," which appeared in Psychology Today in 1978, psychologist Rosalind D. Cartwright reported that she and her associates were attempting to alter the dream plots of sixty recently divorced women. Her comments on the ongoing study began by saying that some of the women "occasionally do succeed."[12] Beyond this preliminary report, however, no results were published. The study was discontinued at the pilot stage because it was not fund by the National Institute of Mental Health. This study ends up as another example of how popular magazine reports can create false impressions.

In the 1980s, claims by enthusiasts for lucid dreaming, such as Stephen LaBerge, raised the hope that lucid dreaming can lead to dream control, as it allegedly does for LaBerge himself, who was by far the most successful subject in his various studies.[13] However, the few other systematic studies that exist do not give much support to this hope. In one such study, for example, experienced lucid dreamers were instructed to turn on a light in the dream. Only two of sixteen subjects reported that they were successful.[14] There have been few attempts since the early 1990s to demonstrate dream control during lucid dreaming, where the emphasis is now on becoming and staying lucid to enjoy what unfolds.

As against these few hopeful glimmers in a few studies, there is stronger evidence that dreams cannot be controlled even to a small extent within dream groups or experimental situations using Stewart's principles. Two of the earliest leaders of Senoi dream groups, Joel Latner and Meredith Sabini, wrote as follows in an article that is very positive toward dream discussion groups as a way of heightening personal sensitivity and enhancing creativity:

We have had scant success with instructed dreaming, but we have harvested some fruits in them. One of us has awakened from a dream with artwork patterns which could be carried out in pastels, and the other has awakened with lines, words, or music, or both, for songs he was writing.[15]

Ironically, a strong piece of negative evidence comes from a 1974 study by Garfield herself. In this study a good dream recaller spent five months trying to increase the frequency with which his hands appeared in dreams. (Casteñada claimed that his mythical sorcerer, Don Juan, often focused on his hands while dreaming.) The dreamer also spent twelve months trying to increase the frequency of flying dreams. But the frequency of hand images stayed the same and the frequency of flying dreams rose only from 2 to 4 percent. Garfield attempts to rescue these findings by claiming that some of the hand dreams became more vivid and some of the flying dreams included intense sensations, but the frequencies speak for themselves against her after-the-fact interpretations.[16]

Two very careful and detailed studies of dream control by psychologists David Foulkes and M. L. Griffin were also unable to report any positive results. In the first study, twenty three participants were taught the Stewart control techniques as described in Garfield's book. They were asked to dream about a randomly assigned target selected from a list of six dream suggestions. They kept daily records of the dreams they remembered over ten consecutive nights. Two independent judges attempted to match the dream reports with the target suggestions. Their matchings did not exceed what would be expected by chance.[17]

The second study used twenty-nine highly motivated students who claimed some previous success in dream control or great interest in the topic. They too tried to dream about targets from a list of six dream suggestions. This time, however, the chosen targets were more carefully monitored by the experimenters "so as to be better equated for emotional tone, amount of detailed elaboration of content, and degree of specific personal relevance." This study also covered ten nights, but this time subjects were allowed to pick the nights on which they felt they were most likely to be able to control their dreams. The subjects reported an average of seven dreams. Four independent judges attempted to match these dreams to the target suggestions. Once again, the correct matches did not exceed what would be expected by chance. The authors reached the following conclusion:

These results cannot, of course, "disprove" the possibility of deliberate pre-sleep dream control. They do indicate, however, that if such control is possible, it must be much more difficult to achieve than enthusiasts such as Garfield generally intimate.[18]

A similar lack of results was reported in a laboratory study by a team of Canadian researchers, who assessed physiological reactions as well as dream content. In this study, seven females and three males aged nineteen to twenty-nine were instructed to either increase or decrease their emotional involvement in their dreams on the fourth and sixth nights of seven consecutive nights in the sleep laboratory. There was a slight increase in the variability of heart rate and respiration on experimental nights, but dream content measures did not show the anticipated changes in emotionality. The authors concluded that their instructions induced stress, but no changes in the dreams.[19]

Just when it seemed that any form of dream control was near hopeless, a new researcher reported success in the 1990s in reducing nightmares with a technique very different from the one advocated by Stewart. In this approach, nightmare sufferers are asked to write out a new ending of their own choosing, and then to imagine the dream, complete with the new ending, several times each day. This "imagery rehearsal" is a technique of cognitive-behavioral therapy. As noted in the first chapter, it was developed without any influence from the advocates of Senoi Dream Theory. Studies by its originator, Barry Krakow, conducted with the help of several colleagues, suggest that it works to at least some degree for many patients.[20]

This finding supports Stewart in the general sense that at least some people seem to have the capacity to "control" nightmarish dreaming if provided with the right techniques. However, the program by which nightmare frequency can be reduced does not fit with Stewart's principles of dream control. First, it is based on allowing people to create their own new endings, not on following rules about what to dream about (confronting danger, going toward pleasure, extracting creative gifts). In some cases, in fact, people pick very unlikely endings, such as being killed by their pursuer, but it works for them. Second, the technique is very individualistic, whereas Stewart claimed that control is learned through socialization within a group. Third, the technique relies on the cognitive strategy of imagery rehearsal, whereas Stewart claimed that dream control is acquired through social reinforcement by esteemed role models. As Stewart put it in his final published statement in 1962:

The analysis of the cross-cultural data proves that the individual cannot change, simplify, and reorganize the inwritten social patterns without the cooperation, permission, and assistance of inwritten social authorities. It not only requires the cooperation of the dream model of the dream interpreter in the dream to effect progressive reorganization; also, apparently this dream reorganization is largely confined to fulfilling the directives received from and agreements entered into with the dream interpreter while the dreamer was awake.

It is also necessary, in Stewart's view, to share and evaluate dreams in order to gain control of them:

Furthermore, the reorganizing effect of each individual dream appears to diminish and lose its validity as the foundation for a further step if the dream is not socially expressed, evaluated, and approved by a respected peer or authority.[21]

Neither of these claims by Stewart is supported by the work using imagery rehearsal. The clock has run out on Senoi Dream Theory as a technique of dream control.

Continue on to Chapter 6


  1. For the prevalence of negative kinds of dream content across the globe, see T. Gregor, "A Content Analysis of Mehinaku Dreams," Ethos 9 (1981):353-90; and G. William Domhoff, Finding Meaning in Dreams (New York: Plenum Publishing, 1996), Chapter 6. [<<]
  2. E. G. Werlin, "An Experiment in Elementary Education," in Contemporary Educational Psychology, ed. R. M. Jones (New York: Harper & Row, 1966), p. 233. [<<]
  3. Ibid., p. 238. [<<]
  4. E. Greenleaf, "'Senoi' Dream Groups," Psychotherapy: Theory, Research, and Practice 10 (1973):218. [<<]
  5. E.g., J. R. Gibb, "Effects of Human Relations Training," in Handbook of Psychotherapy and Behavior Change, ed. A. E. Bergin and S. L. Garfield (New York: John Wiley, 1971); J. Bebout and B. Grodon, "The Value of Encounter," in Perspectives on Encounter Groups, ed. L. Solomon and B. Berzon (San Francisco: Jossey-Bass, 1972); M. A. Lieberman, I. D. Yalom, and M. Miles, Encounter Groups: First Facts (New York: Basic Books, 1973); P. B. Smith, "Controlled Studies of the Outcomes of Sensitivity Training," Psychological Bulletin 82 (1974):597-622; A. Zander, "The Psychology of Group Processes," Annual Review of Psychology 30 (1979):417-51. Most of these studies also show the importance of the leader in determining how members view the experience. A study by Dane Archer, "Power in Groups: Self-Concept Changes of Powerful and Powerless Group Members," Journal of Applied Behavioral Sciences 10 (1975):208-20, shows that the mixed results within some groups can be explained by power differentials within the group itself Those who are perceived as powerful members of the group, as measured by a questionnaire about group interaction patterns, are the ones most likely to improve in self-concept from the group experience. If Archer's results can be generalized, then it may be that not everyone receives the big benefits that are claimed for dream groups. [<<]
  6. B. Zilbergeld, The Shrinking of America: Myths of Psychological Change (Boston: Little, Brown, 1983), pp. 114-21. [<<]
  7. Greenleaf, "'Senoi' Dream Groups," p. 221. [<<]
  8. C. C. Hiew, "The Influence of Pre-Sleep Suggestions on Dream Content," paper presented to the New Brunswick Psychological Association Convention, October 24-25, 1974. [<<]
  9. C. C. Hiew, "Individual Differences in the Control of Dreaming," paper presented to the Association for the Psychophysiological Study of Sleep, 1976. [<<]
  10. C. C. Hiew and P. Short," Emot ional Involvement and Auditory Retrieval Cues in Pre-Sleep Dream Suggestion," paper pre-sented to the Association for the Psychophysiological Study of Sleep, 1977. [<<]
  11. R. D. Cartwright, "The Influence of a Conscious Wish on Dreams: A Methodological Study of Dream Meaning and Function," Journal of Abnormal Psychology 82 (1974):387-93. [<<]
  12. R. D. Cartwright, "Happy Endings for Our Dreams," Psychology Today, December 1978, p. 66. [<<]
  13. S. LaBerge, L. Nagel, W. Dement, and V. Zarcone, "Lucid Dreaming Verified by Volitional Communication During REM Sleep," Perceptual and Motor Skills 52 (1981):727-32. LaBerge notes in his dissertation that it took him two and a half years to learn to be a good lucid dreamer. In his popular book, Lucid Dreaming (Los Angeles: Jeremy P. Tarcher, 1985) he says that most people can become lucid dreamers following his set of cognitive instructions, but by the late 1980s he was selling goggles that blinked during REM sleep as the best way to induce lucid dreaming. [<<]
  14. K. Hearne, "Effects of Performing Certain Set Tasks in the Lucid-Dream State," Perceptual and Motor Skills 54 (1982):259-62; K. Hearne, "Lucid Dream Induction," Journal of Mental Imagery 7 (1983):19-24; P. Tholey, "Techniques for Inducing and Manipulating Lucid Dreams," Perceptual and Motor Skills 57 (t983): 79-90. For a book that shows the difficulties of controlling dream content during lucid dreaming, but has an emphasis on control in its title, see J. Gackenbach and J. Bosveld, Control Your Dreams: How Lucid Dreaming Can Help You Uncover Your Hidden Desires, Confront Your Hidden Fears, and Explore the Frontiers of Human Consciousness. (New York: Harper & Row, 1989). [<<]
  15. J. Latner and M. Sabini, "Working the Dream Factory: Social Dreamwork," Voices 18 (1972):43. [<<]
  16. P. Garfield, "Self-Conditioning of Dream Content," paper presented to the Association for the Psychophysiological Study of Sleep, 1974, as reported in Tart, "From Spontaneous Event to Lucidity," p. 241. [<<]
  17. D. Foulkes and M. L. Griffin, "An Experimental Study of 'Creative Dreaming,'" Sleep Research 5 (1976):129. [<<]
  18. M. L. Griffin and D. Foulkes, "Deliberate Presleep Control of Dream Content: An Experimental Study," Perceptual and Motor Skills 45 (1977):660-62. [<<]
  19. R. Ogilvie, K. Belicki, and A. Nagy, "Voluntary Control of Dream Affect?" Waking and Sleeping 2 (1978):189-94. [<<]
  20. B. Krakow, R. Kellner, D. Pathk, and L. Lambert, "Imagery rehearsal treatment for chronic nightmares," Behaviour Research & Therapy 33 (1995):837-843; and B. Krakow et al., "Imagery rehearsal therapy for chronic nightmares in sexual assault survivors with posttraumatic stress disorder: A randomized controlled trial," JAMA: Journal of the American Medical Association 286 (2001):537-545. In an interview on Feb. 13, 2003, Krakow told me that he first learned about this type of approach when he read an article by the British psychiatrist Isaac Marks, "Rehearsal relief of a nightmare," British Journal of Psychiatry 133 (1978):461-465. [<<]
  21. K. R. Stewart, "The Dream Comes of Age," Mental Hygiene 46 (1962):235. [<<]

Go back to the Dream Library index.
dreamresearch.net home page dreamresearch.net contact info