Orne, M.T., The simulation of hypnosis: Why, how, and what it means. International Journal of Clinical and Experimental Hypnosis, 1971, 19, 183-210.

The International Journal of Clinical and Experimental Hypnosis 1971, Vol. XIX, No.4, 183-210



Institute of the Pennsylvania Hospital and University of Pennsylvania

Abstract: The use of the real-simulating model with a "blind" hypnotist as a quasi-control for hypnosis research is discussed. The advantages and limitations of the procedure are set forth and the kind of problems for which it is appropriate is outlined. The nature of inferences that may legitimately be drawn from findings using this model are discussed. It is pointed out that, though it has been shown that Ss are able to simulate successfully and can deceive highly trained hypnotists, this observation does not challenge the reality of S's experience nor does it question the genuineness of hypnosis. Furthermore, in most contexts, simulation does not occur spontaneously and ought not to preoccupy either the therapist or the investigator.


Hypnosis is said to exist when suggestions from one individual seemingly alter the perceptions and memories of another. In its extreme form hypnosis is easily identified: appropriate suggestions will cause a S to perceive an individual who is not actually there and behave as if he were, or if the S is told to forget certain events that have transpired, he will suddenly seem unable to recall them. In a similar man-

Manuscript submitted February 13, 1971.

1 An earlier version of this paper was presented at the meeting of the Society for Clinical and Experimental Hypnosis, Chicago, November, 1968, as a part of the symposium on The Simulation of Hypnosis--Who, How, What, and Why. The idea for the symposium was originally conceived by Arthur Shapiro, who chaired the session. Theodore X. Barber, Erika Fromm, Ernest R. Hilgard, and Milton V. Kline participated in the discussion. For a considerably more detailed and definitive exposition of these issues, the reader is referred to the chapter (Orne, in press) in Fromm and Shor (in press).

2 The preparation of this paper, as well as the substantive research conducted at the Unit for Experimental Psychiatry, was supported in part by grant MH 03369 and by grant MH 19156-01 from the National Institute of Mental Health, Public Health Service.

3 The author wishes to express his appreciation to Harvey D. Cohen, Mary R. Cook, Frederick J. Evans, Charles Graham, A. Gordon Hammer, Ulric Neisser, Donald N. O'Connell, Emily Carota Orne, David A. Paskewitz, Arthur Shapiro, and Mae C. Weglarski for their helpful comments during the preparation of this manuscript.




ner, appropriate suggestions will alter any desired aspect of memory and perception, making possible an incredibly wide range of distortions of subjective experience. That such occurrences should strike the observer as peculiar, or even magical, is not surprising, nor is it surprising that they have fascinated scientists and laymen alike. If such a course of events followed the ingestion of an exotic drug or was the consequence of some delicate brain surgery or some form of unique electrical stimulation of the brain, it would, nonetheless, be puzzling and striking. However, that such a course of events should follow from a brief, innocuous, and simple procedure is even more remarkable. It is hardly strange, therefore, that since the time of Mesmer efforts have been made to explain or, conversely, to explain away the phenomenon: to claim that the behavior of a hypnotized S is not the consequence of an altered subjective experience, that he is not really seeing the person he seems to be hallucinating, that he really does know the information which he claims to have forgotten, and so forth. It is argued that, for reasons not necessarily understood, some individuals choose to behave as if they had these experiences but, in fact, they join with the hypnotist in perpetrating a magnificent hoax upon not only the hypnotist but also the observers, the world in general, and perhaps even themselves.

To the student of hypnosis, the behavior of the hypnotized individuals seems compelling, and it is hard to imagine that even a trained actor could as convincingly portray in his behavior events that he did not actually experience. Yet, to convince skeptics, as well as to resolve nagging doubts within themselves, early investigators found it particularly important to demonstrate that the hypnotized individual was capable of skills and feats beyond his ken in the normal waking state. Once it was demonstrated that the hypnotized S could do things that were beyond him while awake, any questions about the reality of hypnosis were automatically eliminated. All of the early investigators reported that the hypnotized individual had greater strength, greater memory, and was able to perform various feats not otherwise possible. The many rather extravagant claims that were made for the effects of hypnosis were gradually replaced by more sober, carefully documented findings as empirical methods of research were applied to the study of hypnosis. In the first major systematic effort to quantify hypnotic phenomena, Hull (1933) devoted a great deal of attention to what he called the transcendence of normal volitional capacity and was able to demonstrate significant increments under carefully controlled conditions.

These careful studies compared the performance of the hypnotized individual with his own performance in the waking state. Considerable



sophistication was used in order to control for order effects; however, it is important to note that S served as his own control. If one wished to ask a question about strength during hypnosis, one studied the individual's performance in the wake state and in hypnosis. While the effects of practice could be controlled by varying orders, such a procedure could not prevent S from recognizing that his performance during hypnosis was to be compared with his performance during the wake state. In fact, Hull, in commenting on Sears's (1932) study of analgesia, recognized that under some conditions Ss might alter their baseline performance in the waking state in order to make their hypnotic performance more striking. This effect, studied definitively in recent years by Zamansky, Scharf, and Brightbill (1964), raises serious methodological questions.

Another problem that characterized much of the early systematic work was the difference in the treatment of Ss when they were hypnotized as compared with when they were awake. Anyone who has worked with hypnosis will recognize that it is extremely difficult to refrain from giving instructions to hypnotized individuals with greater intensity and conviction than is given to the same Ss in the waking state. In a relevant study, Troffer and Tart (1964) have shown that judges can readily recognize the voice quality used by hypnotists in addressing hypnotized individuals since it is quite different from that normally employed.

Because of problems such as these, the appropriateness of the designs that had been used to evaluate transcendence of normal volitional capacity seemed questionable. The fact that a S might give a better performance under similar conditions in hypnosis than in the wake state could be a function of his interpretation of the total situation, differential instructions, or motivational effects, and would not necessarily reflect a greater capacity. It was possible to show in an experiment involving the holding of a kilogram weight at arm's length that Ss were able, in the waking state, to exceed their own maximum performance obtained during hypnosis (Orne, 1954). In this rather simple study, no attempt was made to equate the motivation of S in hypnosis and in the waking state. On the contrary, during hypnosis S was asked to rest his hand on a hallucinated table and was given suggestions of super-human strength; in this way the unique attributes of hypnosis were used in order to obtain a maximum performance. On the other hand, during the wake state appeals to male chauvinism, sportsmanship, monetary reward, and exhortation were employed. It was argued that, from the point of view of determining transcendence of normal volitional capacities, it mattered little what alternative motivational tech-



niques were used to elicit the maximal performance; it was sufficient to demonstrate that under some circumstances individuals in the waking state could indeed exceed their hypnotic performances.

Further, performances such as the ability to become sufficiently rigid to be placed across two chairs and to support someone on one's midriff and similar feats of strength previously ascribed to hypnosis could, with appropriate motivation, be duplicated in the wake state (Orne, 1962c). It seemed reasonably certain that, with appropriate motivational instructions, it would be possible for waking Ss to equal and surpass the kind of performance that had previously been used to demonstrate transcendence of normal volitional capacity. This observation was subsequently documented in exquisite detail by Barber (1969) and his associates.

Difficulties in Identifying Which Behaviors Are Induced by Hypnosis

In another line of research it had also become clear that much of the hypnotized S’s behavior during hypnosis could, as White (1941) had suggested, be understood in terms of S’s conceptions about how hypnotized Ss behave. Indeed, it was shown (Orne, 1959) that Ss who had observed hypnotized individuals manifest a plausible item of behavior never previously observed during hypnosis (catalepsy of the dominant hand) actually tended to manifest this behavior when they were later hypnotized. It seemed reasonable, therefore, to be concerned about the extent to which Ss might pick up subtle cues, possibly outside of the awareness of the hypnotist, as the basis for their behavior during hypnosis. An investigator could observe consistent behavior among his Ss and come to believe that such behavior reflected an essential attribute of hypnosis, when, in fact, it was a function of how the investigator himself treated Ss. The stages of hypnosis described by Charcot, the demonstration of the effects of magnets on mesmerized Ss, Mesmer's seizures, as well as many similar observations by others, seem to have been such self-fulfilling prophecies: the hypnotized Ss would produce whatever behavior was suggested either explicitly or implicitly. Therefore, the investigator who assumes that Ss respond only to explicit suggestions will inevitably observe behavior on the part of his Ss which he ascribes to hypnosis but which is merely a function of subtle cues that he himself generates. As Merton (1948) has pointed out in his analysis of the self-fulfilling prophecy, experience--increasing the number of Ss--does not help clarify these problems; it merely leads to even greater conviction on the part of the investigator (in this instance, that he is observing an essential attribute of hypnosis). It is extremely diffi-



cult for the investigator to place himself sufficiently outside of the situation to recognize this problem.4

Pattie (1935) in his classic study of hypnotically induced uniocular blindness demonstrated some of the problems inherent in using Ss as their own control. In one of his Ss hypnosis induced what appeared to be true uniocular blindness. Because he found it difficult to accept the findings, Pattie devised a series of test procedures to evaluate the phenomenon, on each of which the S responded correctly. Finally, when he devised a test procedure based on a particularly obscure phenomenon, the S failed to respond appropriately and then admitted she had been able to "psych out" the previous procedures.

A large number of studies purported to show that suggestions given to deeply hypnotized Ss were able to modify their basic skills and enable them to do things they were unable to do while awake. Although, in Pattie's study as well as in our investigation of transcendence of volitional capacity, it was possible to demonstrate that the S’s behavior could be explained by simple psychological mechanisms, such demonstrations required the investigator to have correctly guessed the nature of these mechanisms. Because it is essential to use deeply hypnotized Ss and a great deal of effort is required to select and train such individuals, economy virtually dictates that the S be used as his own control. Consequently the individual's own waking behavior is used to evaluate his hypnotic performance. The question inevitably arises not only whether the waking performance of the S is a fair measure of his ability, but also whether the S might not have figured out the purpose behind the whole experiment and, like Pattie's S, wittingly or unwittingly behaved in a manner which would make the hypnotic performance seemingly transcend his waking capabilities. In order to avoid such problems, some experiments employ test procedures based upon obscure phenomena to evaluate the S's capacities in hypnosis and in the wake state, the assumption being made that the S could not be familiar with the procedures. It becomes a matter of judgment whether one chooses to accept this view or whether one believes that Ss really could have figured out what was expected of them from cues in the

4 Problems of this kind are by no means unique to hypnosis, although they account for much of the acrimonious controversy that has characterized its history. A delightful example is the case of Clever Hans, a horse who apparently could perform simple arithmetic and would tap out the answers with his right front hoof. After many outstanding scientists had been convinced of the horse's ability, Pfungst (1911) finally solved the mystery when he recognized that the horse would continue to tap his foot as long as his trainer looked expectantly at it and would stop when the trainer ceased to observe it. Even a casual look at the history of the use of magnets in mesmerism reveals striking parallels.



experimental situation. A comparison group of nonhypnotized Ss was needed which could be exposed to the identical cues provided for the hypnotized group. This would include repeated exposure to the identical test procedures as well as the almost inevitable differential treatment in the wake and hypnotic conditions. Finally, since the purpose of the comparison group is to determine whether hypnosis is essential to produce the phenomenon under investigation, these individuals should be maximally motivated to accomplish this goal. Questions such as the capabilities of unhypnotized individuals, whether Ss can or cannot figure out the desired behavior in a particular experimental context, and similar issues then become empirical questions. The real-simulator model was an attempt to achieve these ends.

The basic concept was an adaptation of the placebo control procedure in psychopharmacology where it had long been recognized that patients respond to pharmacologically inert substances if they are presented as medication. Placebos were known to be effective in relief of minor pain, as sleep medication, tranquilizers, and so forth. However, it was also soon recognized that the physician had to remain blind as to which pill was pharmacologically inert since, otherwise, especially in the evaluation of new medication, the attitude with which the pill was given and the manner in which reported side effects were treated differed radically for the drug and the placebo. Therefore, unless the physician was truly blind, his behavior tended to result in differential effects of drug and placebo.

The intent of the real-simulator model was to have a control group of Ss who were not hypnotized but who would otherwise be treated in the identical manner as hypnotized individuals. The only way to assure identical treatment was to keep the investigator blind.5

Prior to this research (Orne, 1959), there was general agreement in the literature about the impossibility of simulating hypnosis in a way that would deceive the experienced hypnotist. When, nonetheless, we asked Ss to simulate hypnosis, it became clear why there was such agreement on this point. The Ss found this task both difficult and distasteful. Typically they gave a very transparent performance, tending to interrupt it by asking for reassurance with questions such as, “Am I doing all right?” or “Is this what you want?” When urged to resume the role, they would do so, frequently looking to E for encouragement,

5 The analogy between the use of placebos and our use of simulators is by no means perfect. The patient should not suspect and the physician should not know that the given individual is on placebo whereas the S instructed to simulate is not blind about his own status and differs in a variety of ways from the hypnotized individual. The hypnotist, however, is blind and thus it is possible to use the behavior of the simulator to evaluate the kind of cues made available to S.



smiling with embarrassment, and in many clearly discernible ways communicating their discomfort in the situation. In discussing these experiences, Ss indicated that they felt foolish in carrying out a task which was manifestly impossible. They had been asked to simulate hypnosis, but because the request had emanated from a hypnotist who knew that they were simulating, there was no conceivable way in which they could successfully comply with the request without appearing ridiculous.

It was necessary to make a small but absolutely vital change in the procedure: two Es had to be involved so that the hypnotist-experimenter could remain absolutely blind as to which Ss were in deep hypnosis and which Ss were simulating. Under these conditions, Ss are instructed by one investigator to simulate in order to fool another hypnotist who they knew would not be aware of their true status.

This altered situation produced a remarkable change in the performance of simulating individuals. Initially, I had intended to develop a method to evaluate the effects of nonhypnotized Ss’ perceptions of what was expected of them on the experimental tasks under investigation. However, as a clinician who had worked extensively with hypnosis, I never doubted that it would be easy for me personally to recognize those Ss who were, in fact, simulating. It came as a complete surprise as well as a considerable blow to my fantasies of omniscience to find that Ss were actually able to deceive me. It turned out initially to be impossible to distinguish between simulating and hypnotized Ss with a high degree of certainty or even better than chance. Many of the procedures which I had casually accepted as useful tests to determine the Ss’ depth of response and had assumed to be proof positive of the reality of the hypnotic response failed to distinguish between the groups. Not only were simulators able to perform the many feats of strength without hypnosis, but they also showed themselves capable of tolerating painful stimuli without flinching (Shor, 1962) and were able apparently to recall material that ought to have been beyond their ken (O'Connell, Shor, & Orne, 1970). Simple behavioral tasks did not effectively discriminate between them nor did a variety of procedures which attempted to evaluate the individual's trust in the investigator (e.g., whether, when told there was a chair behind him, S would sit down without actually testing its presence).

A very striking personal experience with this technique occurred when I tried to establish some simple differentiating criteria. I instructed Ss to simulate for colleagues and, while observing the Ss, carefully noted the myriad of clear, definable differences which I could easily perceive between the behavior of simulating and deeply hypnotized individuals. Indeed, it was hard for me to fathom how my



colleagues were so inexperienced and unsophisticated as to fail to recognize the dozen or so blatant mistakes that simulators made in pretending to be hypnotized. It was all the more surprising, therefore, when I again took the role of blind E to realize that I too was quite incapable of successfully distinguishing the two groups, even when attempting to apply what I had learned through careful observation of the behavior of simulators. Perhaps this experience, more than any other, imbued me with an abiding awareness of man's capabilities for self-deception and taught me not to place undue significance on impressions which are not rigorously tested.6

Why Simulating Subjects Should Not Be Able to Enter Hypnosis

Hypnosis cannot be defined simply in terms of its antecedent conditions (see Orne, 1966). When an attempt is made to hypnotize a group of individuals, a few will be profoundly affected, others will have no response whatsoever, while the majority will show varying degrees of response. Furthermore, in some circumstances, given appropriate expectations and subject selection, profound hypnotic phenomena may occur without any form of recognized trance induction (for example, among members of an audience listening to a suitable talk about hypnosis by someone who is considered a leading authority). It would seem foolhardy, therefore, to attempt to define the presence of hypnosis in terms of procedures which have varying effects. The more parsimonious course is to diagnose the presence or absence of hypnosis on the basis of S’s behavior7 and subsequent description of his experience. In other words, the condition of hypnosis is diagnosed on the basis of S’s response to test suggestions, not the mere fact that he has been exposed to a hypnotic induction procedure. If a S responds and describes the experience of hypnosis, he must be considered as having been hypno-

6 Several distinguished colleagues, well-known for their clinical skills, insisted that simulating Ss could easily be distinguished from hypnotized individuals by experienced clinicians. In each instance, when they actually tried to do so with experimental Ss that were set up for them according to the procedures outlined, they found themselves unable to make accurate and reliable differentiation on the basis of their clinical judgment during a single session.

7 The issue is closely analogous to the definition of sleep. It would seem strange to define sleep as that condition which occurs when S is told to go to sleep or even as that condition which occurs after S has gone to bed and turned off the light, as any insomniac will gladly attest. Not only does sleep fail to occur under such conditions sometimes, but it also occurs in some individuals under unlikely and sometimes embarrassing circumstances. Whether an individual is in sleep is diagnosed by observation of his physical appearance as well as by rousing him, seeing how he alerts, and inquiring what has happened. While the availability of EEG criteria has (apparently) simplified this problem, the diagnosis of sleep was made with high reliability long before such measures were available.



tized regardless of the manner in which this condition was brought about. If he fails to respond behaviorally and reports no alterations in his experience, he must be considered as having been awake regardless of the hypnotist's activities. In the final analysis, it is the response of S, not the behavior of E, that defines hypnosis.

Trained independent observers can reliably agree whether an individual is hypnotized if he is selected from the relatively small proportion of individuals who are profoundly responsive to suggestions, particularly if he manifests the following responses: subjectively compelling positive and negative hallucinations, inability to overcome challenge suggestions despite subjectively real efforts to do so, unequivocal suggested posthypnotic amnesia, and posthypnotic behavior carried out in an unconscious, automatic fashion without awareness of the original suggestion. Of course, the majority of individuals are able to respond successfully to only some suggestions while failing others. In such instances, there is likely to be some question about the extent to which the individual is hypnotized. If the intention is to study the effects of hypnosis, it seems logical to utilize that group which most clearly manifests the phenomenon, thereby avoiding some of the ambiguity which results from dealing with individuals who may be affected in varying degrees.

Elsewhere (Orne, 1959; Orne & O'Connell, 1967) we have discussed in more detail the criteria of deep hypnosis. The appropriate procedures to establish whether hypnosis is present under normal conditions are fairly clear-cut. It must be obvious, however, that when Ss are instructed to simulate and are able to deceive trained hypnotists, their behavior cannot then be used as a criterion of whether hypnosis is present. Therefore, an alternative strategy is demanded. Fortunately, an individual's response to hypnosis shows considerable stability over time. The test-retest reliability, for example, between the first and second administration of both the A and B forms of the Stanford Hypnotic Susceptibility Scale is .83 (Hilgard, 1965), which compares favorably with the reliability of IQ tests. Further, after several sessions, the reliability of S’s response in similar situations becomes far higher. After the fifth experience with hypnosis, the S who is able to enter deep trance will, with virtual certainty, be able to do so yet the sixth time. Equally important from the point of view of the investigator, a S who has been unable to enter hypnosis despite earnest efforts on his part will hardly ever respond during the sixth session. For this reason, potential simulators are selected from the extreme lower end of the continuum of hypnotizability after concerted efforts have been made repeatedly to induce hypnosis, using not only standardized procedures but also clinical methods adapted to S’s needs. As a result, one may be



virtually certain that these individuals will at a subsequent time be unable to enter hypnosis, providing the situation is roughly analogous.8

It is equally important at the end of each experiment for the simulating S to be seen again by the investigator who originally instructed him to simulate. At that time a very careful inquiry is carried out about his actual experiences. We have never observed such an individual's entering deep hypnosis; however, very occasionally a simulating S will report isolated phenomena such as: "When I was told I couldn't bend my arm, I tried and I really couldn't. I became so scared I almost stopped simulating." Although such individuals typically do not report successful hallucinations and other deep trance phenomena, they are excluded from the comparison group because it is no longer possible to be certain about the extent to which they might have entered hypnosis.

Objections have been raised about the simulating comparison group because these Ss are drawn from a different population in terms of hypnotizability. For this reason, Austin, Perry, Sutcliffe, and Yeomans (1963) argued that highly hypnotizable Ss should be used as simulators. Unfortunately, results obtained from such a procedure are ambiguous. In our own work, it was apparent that highly hypnotizable Ss, when asked to simulate, had great difficulty in distinguishing the extent to which they were following simulation instructions or were actually entering hypnosis. Hypnotizable simulators seem to "borrow" from their hypnotic experiences while simulating--at times unwittingly. Further, they may interpret the simulating instructions as an implicit request to actually enter hypnosis but make believe that they have not been hypnotized, or they may simply be very confused by instructions which to a S capable of deep trance necessarily question the reality of his previous hypnotic experiences. Since hypnotic behavior is within their repertoire, a failure to find differences between their behavior and that of the real hypnotic group could be because both groups in fact entered hypnosis; differences, on the other hand, could be ascribed to differential understanding of the experimental in-

8 It is important in this context to avoid negativistic Ss or individuals who, for purposes of their own, seem to need to demonstrate an inability to enter hypnosis. Such persons view the situation as some form of power struggle and they can readily be identified during the induction process: for example, they may sway forward when instructed to sway backwards; if told they will be unable to bend their arm, they bend it even before they are asked to test the suggestion. Such individuals, while superficially cooperative, sabotage the induction in a variety of ways and betray their underlying attitude by being manifestly pleased by their inability to experience suggestions. This relatively rare group of individuals may at times be capable of entering deep trance if special induction procedures are able to circumvent their defensive maneuvers (see Erickson, 1952). For Ss of this kind, simulating instructions sometimes provide the setting in which hypnosis becomes possible. They are, therefore, excluded during earlier screenings.



structions. Because of the impossibility of determining precisely what is happening with hypnotizable simulators, they cannot be used to clarify the cues inherent in the experimental situation which is the purpose of simulating Ss in the real-simulator model.

As long as the simulating Ss are drawn from a population of individuals unable to enter hypnosis despite several honest efforts to do so, the investigator may be confident that this group will not have been in hypnosis. This assumption should, of course, be corroborated by carefully conducted inquiries into the individual's experiences during the experiment, carried out in the absence of the hypnotist who is being deceived. The fact that the behavior of these individuals closely resembles that of hypnotized individuals ought not to lead to the seductively simple conclusion that there is no difference between these groups. Ironically, such a view was shared both by individuals whose views would be categorized by Sutcliffe (1960, 1961) as "skeptical" as well as individuals whose views he would categorize as "credulous." The former have argued that, because no differences were observed, hypnosis was no different from simulation, whereas the latter group have used similar logic to uphold the position that, because no differences were observed, the simulating Ss were hypnotized! In fact, however, these two groups of Ss had radically different experiences; and, indeed, under suitable circumstances behavioral differences were eventually identified.

These issues, as well as the kind of inference which may appropriately be drawn from experiments in which simulating Ss are used as comparison groups will be discussed later. First, however, it seems essential to clarify precisely how the simulating procedure is employed in actual practice.


The procedure for setting up the context for Ss to simulate seemed to follow clearly from an understanding of how simulating Ss ought to perceive the experimental situation and the purposes for which the group was designed. It has subsequently become clear that many important aspects of the technique were not spelled out in enough detail for replication, and investigators attempting to use the procedure have varied a number of apparently minor but actually crucial aspects. For this reason, it seems best to describe the procedure in some detail and outline some essential caveats.

Selection of Simulating Subjects

The same subject pool from which the highly hypnotizable Ss are drawn also provides the relatively unresponsive Ss who will eventually



become simulators. These Ss must score less than 4 on initial standardized tests. The Harvard Group Scale of Hypnotic Susceptibility (Shor & E. Orne, 1962) is efficient and useful for such an initial screening but must be followed by an individual standardized test of susceptibility such as the Stanford scales (Weitzenhoffer & Hilgard, 1959, 1962) on which S must again score less than 4. The Ss must then have two additional sessions in which a concerted effort is made to maximize response, using clinical techniques adapted to their individual needs. These sessions are preferably carried out by different hypnotists. The Ss should score no more than 2 on the diagnostic scale (Orne & O'Connell, 1967).

It is absolutely essential that these Ss be unaware of the importance of locating individuals who are unable to respond to hypnosis. Certainly they cannot be told that they might be asked to simulate "if they qualify," nor is it appropriate for the investigator to use any similar technique that would interfere with a careful evaluation of their inability to enter hypnosis when motivated to do so.

An Example of Simulating Instructions

Assuming that S has shown himself to be unable to enter hypnosis in several sessions with different Es and has agreed to participate in another experiment, the simulating instructions would be given along the following lines:

We much appreciate your participation in several of our sessions in the past. Today I would like you to take part in a very interesting experiment which is quite different from any in which you have participated to date. ...You have attempted to go into hypnosis several times and found it quite difficult to respond. Though I understand you have been able to experience a certain lightness in your arm and felt quite drowsy at times, it was not possible to experience much else. ...In this particular study there is a special group of subjects to which you will belong, all of whom were not able to enter hypnosis despite their honest efforts to do so. As you know, people vary in their ability to respond; some individuals find it very easy while some individuals find it quite difficult. This doesn't seem to be related to any other personality characteristics. In this instance your task will be to simulate being a very good hypnotic subject. You will be working with Dr. who is a very experienced and competent hypnotist and will be carrying out an important piece of research. Your task will be to behave as though you were one of those Ss who is able to enter deep hypnosis with ease. There will be only two kinds of Ss in this experiment: those who are highly hypnotizable and those, like yourself, who are not but who will be trying to simulate or fake being hypnotized.

Dr. does know that some subjects will be trying to simulate but has no idea who these subjects will be. Your task is to convince him that you are in fact an excellent hypnotic subject. Now, this is a difficult task and, at times, you may well do something which makes you think you



have given yourself away. Don't worry about this possibility because if Dr. recognizes the fact that you are simulating he will stop the experiment immediately. Therefore, as long as he continues with you, you know you have been successful in faking hypnosis. I point this out to you because in the past we have found that some subjects would suddenly stop, thinking they had goofed and given themselves away, when, in fact, their behavior had been quite appropriate and the investigator had no idea that they were simulating. Keep in mind, then, that as long as the experimenter continues with you, you are doing all right; if he catches on he will stop the study immediately. We realize that you have no experience in how to do this. You were asked simply because you were not able to enter hypnosis. However, we also know from previous studies--we have run a great many studies using this procedure--that intelligent subjects are able to do this. It is difficult but it is possible. ...I can't tell you how to behave or what to do; you have to use whatever you know about hypnosis, whatever cues you get from Dr. , and whatever you learn from the situation to figure out how a deeply hypnotized subject would behave, and your task is then to use this information to simulate hypnosis. Keep in mind that you will be simulating the behavior of an excellent, highly hypnotizable individual and that your task is to maintain that you are going into hypnosis, to perform during hypnosis, and, when you are awakened, to respond as if you had been in hypnosis. In other words, this means simulating not only while you are being hypnotized but afterwards as well. When Dr. asks you about your experiences, you should answer the way a deeply hypnotized subject would answer if he had actually been in trance. If Dr. asks you how you did the last time, keep in mind that you are a good hypnotic subject and you would have gone into deep hypnosis on your previous efforts. You would have had several previous experiences with hypnosis just as you actually have except that you would have entered deep hypnosis. All subjects will have had at least five such sessions with the laboratory.

At no time, once you leave this room, should you reveal to anyone that you are simulating. No one, but you and I, will know that you are simulating. Though it is known that some subjects will be simulating, no one knows who they are except for me. When you are completely finished with the experiment, the investigator will ask you to return to the waiting room, which you will do, and I will meet you there. I will eventually discuss your experiences with you back in this office. Until you are back here with me at the very end of the experiment, you are to reveal to no one that you are not actually hypnotized: this means the experimenter or anyone else who asks you about your experiences other than me when you are back in this room.

The S might ask something along this line:

S: Could I ask you something about how to do it?

E: Certainly.

S: Let's say I am told to make a physical movement. How fast does a good subject react?

E: I can't tell you that. I can tell you only to do what you think you should do.



S: If I am told that I feel no pain, how will I be able to prevent myself from responding?

E: I really can't tell you how to simulate or what to do. It is a difficult task but we have found that other intelligent subjects have been able to do it successfully. Good luck! I will see you back here at the end of the experiment.

Some Important Aspects of the Real-Simulating Model

1. The hypnotist who is working with S must truly be blind about the actual status. When the prospective simulator is told that the hypnotist-experimenter with whom he will work does not know which Ss are simulating, this must, in fact, be an accurate statement and, equally important, before the hypnotist-experimenter is unblinded, the experiment must be entirely over, all Ss having completed all relevant sessions. If this is not the case, the hypnotist often tries to second-guess the experiment, thinking, for example, “I had two reals, one simulator; the next one must be a simulator.” In our experience, it is simply not possible for the hypnotist to treat a S whom he knows to be a simulator the way he would a S whom he believes to be hypnotized--even though he may subjectively feel convinced he has been consistent in his approach. Subtle, but nevertheless real, differences exist which will, in turn, tend to affect S’s performance (see Troffer & Tart, 1964).

2. The simulator, as we have described him, is radically different from a S who is asked to role play hypnosis, simulate with the knowledge of E, or any variation of these. To emphasize this difference in studies with many different kinds of control groups, O'Connell, Shor, and Orne (1970) proposed the term “cryptosimulator” to further distinguish this control group--when the hypnotist is blind to which Ss are simulating--from role playing Ss on the one hand and from Ss who are “simulating” with an investigator who is aware of their true status, i.e., not blind, on the other. Neither of these groups can serve the purposes for which the original simulator procedure was designed.

3. While the concept of the simulating design is a single-blind adaptation of the double-blind placebo technique in psychopharmacology, there are important and very significant differences between the two techniques. Whereas both E and S are blind to the presence or absence of the active pharmacological agent, only the hypnotist, not S, is blind in the simulating design. Further, while the simulator unavoidably knows that he is in a special simulating group, the real S does not know of the simulator's existence.9 Neither real nor simulating Ss are deceived

9 While it is motivating and important in bringing meaning to the task for simulators to know of the existence of a real hypnotic group, when real Ss are confronted with the idea of a simulating group, of necessity the group is understood by them to cast doubt on their own performances. This inevitable conclusion does not allow them the necessary trust in the laboratory that is essential to their



at all. There must be no deception about the blindness of the hypnotist-experimenter involved in dealing with the simulating S. The S must correctly perceive that he has a chance of fooling the hypnotist-experimenter or there is no "percentage" in his trying and he is apt simply to give up.

4. While it is crucial that the simulating Ss have several previous opportunities to enter hypnosis, preferably with different hypnotists using varying techniques, no special training may be given to this group. The very point of having simulating Ss to help evaluate the nature of the cues in the experimental situation is to determine how these would be interpreted by Ss who do not share the hypnotic experience but otherwise have access to the same amount of background information. It is, of course, particularly important that the instructions to simulators in no way communicate how a deeply hypnotized S would actually perform, either on the experimental procedures being tested or on any other aspect of hypnotic behavior. While simulators as well as real Ss may have seen hypnosis at some time in the past, neither group should include individuals who have special knowledge or experience.

5. The Ss should be given simulating instructions at a time immediately preceding the experimental session. This contiguity will assure their understanding what is required of them and will tend to facilitate a high level of motivation. While in certain studies it may seem more efficient to give instructions to Ss well in advance of the "hypnosis" session, Ss then tend not to have a clear enough picture of their task in their minds at the time of the experimental run. Perhaps even more important, the spontaneity and enthusiasm of participating in an exciting study as a member of a very important control group are sometimes lost. With time, self-doubt about the ability to simulate successfully looms large, and S may become unduly concerned with possible failure. Instead of focusing on fooling E, the S may become troubled about looking foolish himself. The model is designed to overcome this problem by assuring S that E is blind, thereby making it possible for him to succeed and making it worth his while to try his best to do so.

It is possible, however, to have S simulate for several sessions extending over a period of weeks, provided that the "set-up" takes place immediately prior to the first session. In this instance S successfully completes the first session before he has an opportunity to become concerned. As a result, he will be less likely to fear failure--believing the experiment would have been terminated had he been unsuccessful; and having had actual experience in the role of simulation, he is unlikely to

Footnote 9 continued. concentrating on attaining full hypnotic depth, and it is therefore not good practice that the simulating group be discussed with them as it can distinctly hinder their cooperation.



forget the task. Further, in the pre-experimental "set-up" interview, it is very important for S to be told precisely when his task as simulator will be completed and also for the end point to be clearly marked by a final interview with the investigator who originally instructed him to simulate.

6. The S should clearly understand that simulation is a crucial part of the experimental situation, that it is his task as S to fool E. It is helpful in this regard for the E who is to be fooled to be seen as competent and senior; this is best accomplished implicitly by identifying the blind E as the responsible investigator. Yet S ought not to view his role only as making a fool of the investigator, but rather as a vital control procedure, contributing to the scientific process--success being difficult, but possible.

Under these circumstances we have not encountered any difficulty with Ss’ being concerned about deceiving the investigator. The only situation in which we have observed such problems was when the E to be fooled was a relatively young and somewhat insecure graduate student with whom some Ss tended to identify. Under these specific circumstances some Ss felt uncomfortable since they perceived the situation as a test of E rather than of themselves. It is important in setting up Ss to simulate not to reassure them about the legitimacy of faking hypnosis. To tell Ss beforehand not to feel guilty about simulating clearly conveys an implicit assumption that simulators have felt or should in fact feel guilty about faking. This problem is most effectively dealt with by clearly communicating to these Ss that they are an essential control group, that the existence of the group is known to the investigator, but that the individual identities of its members are not known by E.

7. We have found it to be essential for the investigator who sets up S as simulator to arrange, at that time, for a post-experimental interview with S at the completion of the experiment. This relationship provides a necessary audience for the S who is simulating--someone with whom he can eventually share his success, someone who will provide him with feedback about the experiment and with whom he can discuss how he actually felt. Since the information S provides at the very end has always been viewed as crucially important by our laboratory, arrangements for a final interview were routinely made whenever the simulator model was used. In retrospect, it seems very likely that an advance arrangement of this kind played a significant role in facilitating the simulator's performance.

8. Only in rare cases does a simulator behave so transparently that the blind E can recognize with certainty that he is faking. If this is the case, of course, the data from such a S must be excluded from the



analysis.10 However, we do not feel it appropriate to terminate the experiment at that point and embarrass S; instead, in these rare cases, E’s certainty is marked by his immediately shortening the normal experimental procedure, but not in such a way as to alert S. We believe that every effort should be made to have any S participating in an experimental study perceive his performance as a success experience; his

data are real, and he has thereby contributed to science even if his performance was not congenial to E’s aims. To bluntly inform S that he has been unsuccessful in carrying out an experimental requirement is, at best, unkind and frequently quite traumatic for S. A little more time and effort can insure that S leaves the study with a good feeling. For ethical reasons alone, therefore, it is inappropriate for the blind E ever to stop an experiment and inform S that he is faking. In addition, such behavior could make the task of subsequent simulating Ss more difficult to the extent that it could become campus scuttlebutt that Ss who fail at simulating hypnosis are rudely dismissed.

Even with great care it is possible--and we have seen two or three such instances--for a S instructed to simulate not to understand the instructions. In one instance, S erroneously assumed that the simulating instructions were intended to facilitate hypnotizing him. In another, S simply did not believe that the blind E actually would be blind. Under such circumstances, an experimental session is kept as short as possible, the blind E carrying out whatever experimental task might seem reasonable to S without making him aware of his inappropriate behavior. The post-experimental interview is then used to explore S’s actual perceptions, again without making him feel that he has failed in his task.

9. We have found it useful to ask the blind E to specify whether a given S is hypnotized or simulating at the end of each run, although no feedback to E is provided. At the conclusion of the experiment it is then possible to determine whether or not E's perceptions of S’s status affected S’s behavior, especially when, as is frequently the case, his judgments prove to be random.

In an optimal situation, the blind E knows as little as possible about the design, and he should certainly not know how many real and how many simulating Ss he will see. When technically possible, the blind E

10 In our experience such an occurrence is sufficiently unusual that there has never been more than one S in any given study who was identified with certainty by the blind E. This includes the kind of S who fails to understand instructions referred to below. Although such Ss would not tend to affect the group data to any significant degree, it is against the logic of the design to include them--except as a footnote. Unless the unhypnotized S instructed to simulate is able to perform the task required of him, he will not be treated equivalently by the blind E, thereby violating the assumption of the model.



should not even know about the existence of the simulating Ss; under such circumstances, of course, no judgments would be obtained.


The Simulator as a Quasi-Control

The basic logic of modern experimental research is derived from the physical sciences. The S is conceived of as a passive responder to stimuli, and, therefore, the precise effects of specific stimuli can presumably be established by comparing individuals who are exposed to a particular set of stimuli (experimental Ss) with another group of Ss (the control Ss) who are treated in a like manner but are not exposed to the experimental stimuli. While this approach is exceedingly powerful in the physical sciences, it has long been recognized implicitly that special problems plague psychological research with human Ss. Individuals are not merely passive responders; on the contrary, they are active, sentient beings who respond to stimuli differently, depending on their perception of the total context.

In a variety of studies (Orne, 1959, 1962b, 1969, 1970), we have been able to show that Ss' perceptions of the purpose and expected outcome of an experiment, as well as their views about what constitutes an appropriate response to the situation, are significant determinants of their behavior. In a vain attempt to avoid these issues, some investigators have simply refrained from informing S explicitly about the nature of the experiment and its purpose. Unfortunately, this procedure does not prevent S from drawing his own conclusions about the situation. Other investigators have tried to control the problem by purposively misinforming Ss. Indeed, this ploy has been sufficiently ubiquitous that a whole generation of Ss has become convinced that "psychologists always lie." The importance of S’s perception in deception studies seems self-evident since, if it were not important, there would be no reason ever to attempt to deceive S. However, ethics aside, lying to S in no way assures that he will believe what he is told. The crucial issue is not the information explicitly provided but what S perceives about the experiment.

What are the cues which determine how S views an experimental situation., what he construes to be the purposes of the study, and what he determines is the appropriate behavior for Ss in the situation? Certainly the explicit instructions given to S are not the only, nor necessarily the most important, basis for his conclusions. "Scuttlebutt" about an experiment, the behavior of the investigator, subtle cues E may make available during the experiment, and, above all, the ex-



perimental procedure itself--all interpreted by S in the light of his past knowledge and experience--contribute to S’s understanding of what an experiment means and what is expected of him. The sum total of these cues we have called the demand characteristics of the experimental situation (Orne, 1959).

Psychological experiments are carried out in order to permit inference about the behavior of individuals outside of experimental situations. It is crucial, therefore, to determine the extent to which a given finding is uniquely related to the demand characteristics intrinsic to the particular experimental situation and the extent to which it may legitimately be generalized to other contexts. We need to determine when we might be dealing with the kind of self-fulfilling prophecy discussed earlier which holds only in a particular experimental context. For example, if it were generally believed by investigators as well as Ss that hypnotized individuals are physically stronger in hypnosis than they are in the waking state, then a number of investigators would in all likelihood document this belief with experimental Ss regardless of the effects hypnosis might actually have on physical capacity. In other words, an experimental finding of increased physical capacity would then be a function of the demand characteristics of the experimental situation in which it is tested rather than of the independent variable under investigation, i.e., hypnosis.

A great deal of emphasis has been placed on statistical means of determining whether a given set of observations is likely to be related to some experimental treatment or could as easily be the consequence of chance fluctuations. Considerably less attention has been paid to the conceptually far more difficult problem of the validity of the experimental procedures, the extent to which a given experimental procedure adequately reflects the process outside of the laboratory about which we hope to draw inference.

Demand characteristics cannot be eliminated, nor can E establish what the demand characteristics are simply by looking at the procedure. While experience can help form educated guesses, each experiment must be evaluated as it is seen from the point of view of S. The authority on how Ss perceive experiments cannot be E; rather, it must be Ss who have actually been exposed to the totality of the cues available in the situation. For this reason, a variety of techniques, aimed at clarifying how Ss actually perceive a given experimental situation, have been developed. These all have the peculiar quality that they utilize the active cognitive mechanisms of one group of Ss (or the same group at a later time) to shed light on how the experimental Ss themselves could actually have perceived the particular experimental situation.



These procedures, then, use the identical mental processes which tend to confound psychological research to more clearly reflect the possible effects of Ss' perceptions on their behavior. They are designed to test the adequacy of an experimental procedure, not to permit definitive conclusions about the nature of the real world. In order to distinguish them from the more usual kinds of controls, they have been termed "quasi-controls."

The simulation procedure, despite its superficial resemblance to other more typical control procedures, is in fact a classic example of a quasi-control. Findings from simulating Ss, therefore, never permit a direct inference about the nature of hypnosis (the independent variable); rather, they are designed to evaluate the possible effects of S’s perception of what is expected of him on the observed experimental behavior. The specific purpose of a simulating subject group then is to evaluate the kinds of cues that are being communicated about what constitutes an appropriate response for a hypnotized individual.

A characteristic of quasi-controls in general, and especially the simulation design, is that they provide a very stringent test of the experimental procedure. Simulating Ss are purposively selected for their inability to enter hypnosis and cannot, therefore, base their behavior upon the kind of subjective experiences which are available to the hypnotized individuals. These Ss, however, are required to mimic the behavior of a deeply hypnotized S in order to deceive the hypnotist. Consequently, they become exquisitely responsive to all cues, be they subtle hints by the hypnotist, information inherent in the instructions, the nature of the experimental procedure, or any aspect of the situation other than the hypnotic experience. To the extent that these individuals are successful, it is clear that (a) the behavioral response in question lies within the unhypnotized individual's repertoire and (b) that sufficient information is available in the total procedure to inform the unhypnotized individual of the expected behavior. The appropriate conclusion to be drawn from such a finding is that the experimental procedure that was utilized was inadequate to definitively demonstrate anything about the nature of hypnosis since the behavior of hypnotized Ss could also be explained as a function of the demand characteristics of the situation. It must be emphasized, however, that a finding such as this in no way demonstrates that the behavior of the hypnotized group is also due to the demand characteristics. It could as easily be due to hypnosis. Under these circumstances, the only inference that can legitimately be drawn is that the experimental procedure does not permit the investigator to decide which of the two alternative sets of explanations best accounts for the hypnotized S’s behavior.



The Circumstances under Which the Real-Simulator Model is Useful

Because the simulating Ss are quasi-controls, the interpretation of research using such groups requires considerable care. Since the real-simulator model is also exceedingly tedious and difficult to use in practice, it may be tempting to conclude that there is no need for such a technique in hypnotic research. Indeed, for many experimental questions it is neither necessary nor useful. However, as a procedure for evaluating three specific types of questions, it is uniquely helpful, although the kinds of conclusions which can appropriately be drawn from findings of the real-simulator model differ somewhat in each of the three cases:

1. When a claim is made that the hypnotized individual is able to transcend his normal volitional capacity, the real-simulator model provides a needed control; for example, after anesthesia is induced and a long needle is pushed through the S’s hand, when another S in deep hypnosis maintains his outstretched cataleptic arm for a long period of time, or when another S who had not spoken his native tongue since age 5 suddenly appears fluent in it during age regression. In each instance the hypnotist is likely to assert, and the observer is likely to agree, that the S could not have behaved as he did were he not hypnotized. If simulating Ss, run by a hypnotist blind as to their nature, produce data indistinguishable from that of the hypnotized individuals, it is possible to assert with confidence that the particular behaviors under investigation are within the repertoire of the unhypnotized waking individual. Under these circumstances, questions of subject selection are irrelevant--the simulators demonstrate that the behavioral tasks being investigated do not transcend normal volitional capacity. The model, then, evaluates the adequacy of the particular procedure to test the specific claim under investigation. We know of no other procedure which permits the hypnotist to employ his full range of clinical skills to elicit a hypnotic phenomenon while, at the same time, putting the transcendence question to a rigorous test. Used in this fashion, the real-simulator model does not demand a rigorous standardized procedure which might interfere with the phenomenon being studied; instead, it illuminates the possible extra-hypnotic consequences that the procedure might have.

2. Whenever hypnosis appears to result in an unusual willingness of Ss to carry out behaviors requested of them, it is instructive to evaluate how an unhypnotized individual would behave in the identical situation. Since Ss' behavior is greatly influenced by the relatively subtle communication of expectations by E, it is useful to have a group for whom a wide range of behaviors has been legitimized, on the one hand,



and a situation in which E expects the behaviors to be carried out, on the other. These are attributes which the simulating situation shares with hypnosis but in the absence of hypnotic experience. The question of whether a nonhypnotized individual would be willing to pick up a poisonous snake or remove a penny from a beaker of nitric acid with his bare hands (Orne & Evans, 1965) cannot be resolved adequately by asking him whether he is willing to carry out these actions. Most Ss characteristically and appropriately reply with a resounding "No." Quite a different situation exists, however, if E peremptorily demands that S undertake these actions, communicating to S that he expects his requests to be carried out. The use of the simulating comparison group permits the evaluation of S ’s behavior in precisely such a situation. Similarly, in an age regression experiment, to instruct an adult to play in the sandbox and make mudpies is not a meaningful control for hypnotized age regressed Ss. Even instructing a S to role play a 6-year-old and make mudpies is quite different from legitimizing his behavior by inducing hypnosis and providing appropriate role support. This is possible to achieve in the real-simulator model: the blind hypnotist expects the S to behave as a child and treats him accordingly. The nonhypnotized simulating S is therefore automatically provided with these important aspects of the hypnotist's behavior (O'Connell, Shor, & Orne,1970).

In situations such as these, the real-simulating model helps to clarify the kind of extra-hypnotic components in the experimental situation that contribute to the hypnotized S’s behavior. Again, to the extent that simulating Ss can and do behave identically to deeply hypnotized individuals, the experimental tasks are insufficient or inadequate to demonstrate that hypnosis leads to an increased willingness of S to undertake actions demanded of him.

We have not observed differences between deeply hypnotized and simulating Ss in experiments directed at this issue. Perhaps this observation is a function of the kind of tasks we have selected; on the other hand, it also suggests that some investigators have been unduly concerned about the effects of subject selection, particularly because of an assumed but not, in fact, observed difference in the hypnotized S 's willingness to please the hypnotist (Orne, 1966, 1970).

3. Whenever an experimental deception is involved, the real-simulator model can provide an appropriate test of its effectiveness. In this context, the simulator group is used in a somewhat different fashion. Since simulators, by virtue of the instructions given to them, are hyper-alert to experimental subtleties, their behavior can help to evaluate in a very rigorous way the extent to which S might have figured out what is going on.



A study of the effect of a post-experimental suggestion on behavior occurring outside the experimental context (Orne, Sheehan, & Evans, 1968) is a case in point. The Ss, given a suggestion to run their hands through their hair every time they heard the word EXPERIMENT for the next 48 hours, were tested outside the experimental context by the secretary-receptionist, once as they left the laboratory and again as they returned for another experiment on the succeeding day. The Ss served as their own control, and it turned out that the most hypnotizable Ss responded as readily in the absence of the hypnotist as in his presence. The crucial question in the interpretation of these findings, however, is whether Ss realized that their behavior was being recorded by the secretary and that their interactions with her were in fact part of the experiment. Exclusively to test this experimental deception, a group of simulators was run concurrently. While all simulators responded in the presence of E, no simulator responded on both of the days in his absence, thus lending behavioral support to the assertion of both the hypnotized Ss and the simulators that they did not view the interaction with the secretary as part of the experiment. Under these circumstances, the sole purpose of the simulator was to test whether or not the experiment was carried out with sufficient care to successfully deceive even highly suspicious Ss. Obviously, if the simulators failed to recognize that the secretary was part of the experiment, they would not respond to the posthypnotic suggestion. Once the simulator had served the purpose of demonstrating the ecological validity of the experiment, it became possible to use the experiment to draw conclusions about how deeply hypnotized Ss respond to posthypnotic suggestions outside the experimental context. The crux of the experimental demonstration, then, did not lie in the statistical difference between the responses of the two groups but only in the behavior of the hypnotized group itself, once the behavior of the simulating group had established the adequacy of the experimental situation for the purpose of studying the question under investigation.

While, conceptually, it is possible to separate the three kinds of situations in which simulating Ss can be useful, it should be recognized that in studying some phenomena more than one aspect of the simulator model is involved.

Conclusions from Work with Simulating Subjects

The difficulty even experienced clinicians encounter in distinguishing between simulating and deeply hypnotized Ss should not be taken to suggest that there are no differences between these two groups or, for that matter, that simulation can be used to explain hypnosis. It indicates only that observations made in the usual clinical context do not



hold when Ss are motivated to deceive. This is similarly true of the remarkably stable and highly useful objective measures of hypnotic susceptibility such as the Stanford scales (Weitzenhoffer & Hilgard, 1959, 1962). If Ss choose to do so, it is an easy matter for them to falsify their responses. Fortunately, under most experimental conditions, such deception does not occur and, therefore, S’s behavior on these tests remains a highly useful measure. Similarly, clinicians have learned to recognize how a patient behaves as he goes into hypnosis. The fact that a S can purposively mimic these behaviors--without having the experiences of hypnosis--does not affect the utility of these behaviors for evaluating hypnotic responses in the clinical situation. Spontaneous simulation is a rare event in clinical practice, much in the way that purposive lying to psychotherapists is highly unusual except in circumstances under which the patient might derive substantial benefit from doing so, as in cases involving compensation neuroses.

While the fact that hypnosis can be successfully simulated ought not to trouble a clinician using hypnosis, this knowledge should help him recognize the limitations of his experience. For example, it would be quite inappropriate to use hypnosis in an attempt to prove the honesty of a witness's statement. Not only are some individuals capable of willfully lying despite being deeply hypnotized, but also none of the clinical tests usually employed can guarantee that the patient is, in fact, hypnotized. This limitation extends to simple tests of pain tolerance so often erroneously employed as proof that Ss are in deep hypnosis.

Studies with simulating Ss, as well as other recent research (e.g., Barber, 1969; Sutcliffe, 1960, 1961) have demonstrated that hypnosis does not magically increase an individual's capabilities beyond those available to him in a motivated waking condition. Consequently, they obviate the need for theorists to explain how the hypnotized individual gains superhuman capabilities. Further, these studies have helped to emphasize the tremendous importance of subtle cues in producing many of the behaviors observed in hypnotized individuals. In the past, investigators had been largely preoccupied with S’s response, but the concern with subtle cues has helped focus attention on what the hypnotist actually does. In a clinical context, when the therapist employs hypnosis, many significant aspects of his behavior are altered. Some of the therapeutic effects attributed to hypnosis may be direct consequences of changes in the therapist's behavior and depend less upon the patient's actually experiencing hypnosis (Orne, 1962a).

Although we have emphasized the failures to observe differences between real and simulating Ss, differences do exist, and the meaning of



these experimental findings must be considered. Simulating Ss differ from deeply hypnotized individuals not only in that the latter are in deep hypnosis but also in that the former have been instructed to simulate. This set to simulate may be as responsible for differential behavior as the presence of hypnosis; for example, simulators are more likely to answer with "I don't know" when asked questions which require committing themselves on some ambiguous issue (Orne, 1959). Recently, Sheehan (1971) has demonstrated differences between real and simulating Ss on psychological tests which resulted from the instruction to simulate rather than the presence or absence of hypnosis. Finally, in interpreting observed differences in behavior, the fact that these two groups differ in hypnotizability and the possible consequences of this selection factor should be borne in mind.

Despite the necessary caveat about the interpretation of findings from the real-simulator model, certain kinds of differences are of great theoretical importance. The behavior of simulating Ss concretely reflects how Ss interpret the sum total of cues available in an experiment in order to best act the part of a hypnotized individual. When the deeply hypnotized S behaves differently, his actions are counter- expectational and can no longer simply be attributed to his effort to play the role of a hypnotized individual; his behavior must involve some other mechanism and reflect some of the processes which are the essence of the hypnotic phenomenon.11 To date such differences have been observed when one appropriately tests the deeply hypnotized individual's remarkable tolerance for logical incongruity and the mixture of hallucination with percept, which I have referred to as "trance logic" (Orne, 1959). A somewhat related finding is source amnesia: Ss who are unable to recall what has transpired during the course of hypnosis unhesitatingly correctly answer questions about information which they acquired while hypnotized (Evans, 1971; Evans & Thorn, 1966). Finally, it has been possible to demonstrate with the real-simulating model that in hypnotized individuals the effect of hypnosis and post-hypnotic suggestions is not dependent upon Ss' belief that they are being observed or that their actions are part of an experiment (Evans & Orne, 1971; Orne, Sheehan, & Evans, 1968). In other words, in contrast to simple instructions, the effects of hypnosis or suggestions given in hypnosis persist beyond the interpersonal relationship in which they were initiated.

11 It is important to establish that the observed differences are not due either to the set to simulate which we discussed earlier or to differences in S selection. The relevance of these alternative explanations may need to be determined by the use of additional control groups.



With some effort it has been possible to create experimental conditions which clearly demonstrate that the hypnotized individual's behavior cannot be explained simply as a response to his beliefs about what is expected of him. Due to the large amount of knowledge available about hypnosis and the amount of information inadvertently transmitted during most experimental sessions, experimental demonstrations of this kind are difficult. On the other hand, it would be strange indeed if the dramatic alterations in the subjective experience of the hypnotized individual did not result in demonstrable behavioral differences. Future research should identify other counterexpectational behavioral consequences of hypnosis. At the present time, however, sufficient evidence is already available to dismiss the view that hypnosis can be explained as a magnificent hoax or some kind of a simulation. It is perhaps ironic that simulating Ss have proven to be a powerful tool in showing that hypnosis is not simulation!


AUSTIN, M., PERRY, C., SUTCLIFFE, J. P., & YEOMANS, N. Can somnambulists successfully simulate hypnotic behavior without becoming entranced? Int. J. clin. exp. Hypnosis, 1963, 11, 175-186.

BARBER, T. X. Hypnosis: A scientific approach. New York: Van Nostrand Reinhold, 1969.

ERICKSON, M. H.. Deep hypnosis and its induction. In L. M. LeCron (Ed.), Experimental hypnosis. New York: Macmillan, 1952. Pp. 70-112.

EVANS, F. J. Contextual forgetting: A study of source amnesia. Paper presented at the meeting of the Eastern Psychological Association, New York, April 1971.

EVANS, F. J., & ORNE, M. T. The disappearing hypnotist: The use of simulating subjects to evaluate how subjects perceive experimental procedures. Int. J. clin. exp. Hypnosis, 1971, 19, 277-296.

EVANS, F. J., & THORN, W. A. F. Two types of posthypnotic amnesia: Recall amnesia and source amnesia. Int. J. clin. exp. Hypnosis, 1966, 14, 162-179.

FROMM, E., & SHOR, R. E. (Eds.), Current trends in hypnosis research. Chicago: Aldine Press, in press.

HILGARD, E. R. Hypnotic susceptibility. New York: Harcourt, Brace & World, 1965.

HULL, C. L. Hypnosis and suggestibility: An experimental approach. New York: Appleton-Century, 1933.

MERTON, R. K. The self-fulfilling prophecy. Antioch Rev., 1948, 8, 193-210.

O'CONNELL, D. N., SHOR, R. E., & ORNE, M. T. Hypnotic age regression: An empirical and methodological analysis. J. abnorm. Psychol. Monogr., 1970, 76 (3, Pt.2).

ORNE, M. T. Die Leistungsfahnigkeit in Hypnose und im Wachzustand. Psychologische Rundschau, 1954, 5, 291-297.

ORNE, M. T. The nature of hypnosis: Artifact and essence. J. abnorm. soc. Psychol., 1959, 58, 277-299.

ORNE, M. T. Implications for psychotherapy derived from current research on the nature of hypnosis. Amer. J. Psychiat., 1962, 118, 1097-1100. (a)

ORNE, M. T. On the social psychology of the psychological experiment: With par-



ticular reference to demand characteristics and their implications. Amer. Psychol., 1962, 17, 776-783. (b)

ORNE, M. T. Problems and research areas. In Group for the Advancement of Psychiatry, Medical uses of hypnosis. New York: GAP, 1962. Pp. 676--689. (c)

ORNE, M.T. Hypnosis, motivation and compliance. Amer. J. Psychiat., 1966, 122, 721-726.

ORNE, M. T. Demand characteristics and the concept of quasi-controls. In R. Rosenthal & R. L. Rosnow (Eds.), Artifact in behavioral research. New York: Academic Press, 1969. Pp. 143-179.

ORNE, M. T. Hypnosis, motivation and the ecological validity of the psychological experiment. In W. J. Arnold & M. M. Page (Eds.), Nebraska Symposium on motivation. Lincoln: University of Nebraska Press, 1970. Pp. 187-265.

ORNE, M. T. On the simulating subject as a quasi-control group in hypnotic research: What, why and how. In E. Fromm & R. E. Shor (Eds.), Current trends in hypnosis research. Chicago: Aldine Press, in press.

ORNE, M. T., & EVANS, F. J. Social control in the psychological experiment: Anti-social behavior and hypnosis. J. Pers. soc. Psychol., 1965, 1, 189-200.

ORNE, M. T., & O'CONNELL, D. N. Diagnostic ratings of hypnotizability. Int. J. clin. exp. Hypnosis, 1967, 15, 125-133.

ORNE, M. T., SHEEHAN, P. W., & EVANS, F. J. Occurrence of posthypnotic behavior outside the experimental setting. J. Pers. soc. Psychol., 1968, 9, 189-196.

PATTIE, F. A. A report of attempts to produce uniocular blindness by hypnotic suggestion. Brit. J. med. Psychol., 1935, 15, 319-334.

PFUNGST, O. Clever Hans (The horse of Mr. van Osten). [Re-issue of 1911 ed.] (R. Rosenthal, Ed.) New York: Holt, Rinehart &; Winston, 1965.

SEARS, R. R. An experimental study of hypnotic anesthesia. J. exp. Psychol., 1932, 15, 1-22.

SHEEHAN, P. W. Task structure as a limiting condition of the occurrence of the treatment effects of simulation instruction in application of the real-simulating model of hypnosis. Int. J. clin. exp. Hypnosis, 1971, 19, 260-276.

SHOR, R. E. Physiological effects of painful stimulation during hypnotic analgesia under conditions designed to minimize anxiety. Int. J. clin. exp. Hypnosis, 1962, 10, 183-202.

SHOR, R. E., & ORNE, E. C. The Harvard Group Scale of Hypnotic Susceptibility, Form A. Palo Alto, Calif.: Consulting Psychologists Press, 1962.

SUTCLIFFE, J. P. "Credulous" and "skeptical" views of hypnotic phenomena: A review of certain evidence and methodology. Int. J. clin. exp. Hypnosis, 1960, 8, 73-101.

SUTCLIFFE, J. P. "Credulous" and "skeptical" views of hypnotic phenomena: Experiments on esthesia, hallucination, and delusion. J. abnorm. soc. Psychol., 1961, 62, 189-200.

TROFFER, S. A., & TART, C. T. Experimenter bias in hypnotist performance. Science, 1964, 146, 1330-1331.

WEITZENHOFFER, A. M., & HILGARD, E. R. Stanford Hypnotic Susceptibility Scale, Forms A and B. Palo Alto, Calif.: Consulting Psychologists Press, 1959.

WEITZENHOFFER, A. M., & HILGARD, E. R. Stanford Hypnotic Susceptibility Scale, Form C. Palo Alto, Calif.: Consulting Psychologists Press, 1962.

WHITE, R. W. A preface to the theory of hypnotism. J. abnorm. soc. Psychol., 1941, 36', 477-.505.

ZAMANSKY, H. S., SCHARF, B., & BRIGHTBILL, R. The effect of expectancy for hypnosis on prehypnotic performance. J. Pers., 1964, 32, 236-248.



Que Significa Simular la Hipnosis: Porque y como se la Ejecuta

Martin T. Orne

Resumen: Se discute en este trabajo el modelo simulador-real ante un hipnotista "ciego" (vale decir que no sabe quien es quien de los sujetos en estudio) como una forma de control experimental en los estudios sobre hipnosis. Se da cuenta de las limitaciones y ventajas de este proceder, asi como aquellos campos en que debe usarse. Asimismo se discuten las conclusiones que legitimamente puedan sacarse de este modelo. Se enfatiza especialmente que, en modo alguno, el simular bien la hipnosis signifique que quien la experiencia realmente, no la este sintiendo como tal. Por ultimo, se destaca el hecho que en la mayorIa de los contextos en que ocurre la hipnosis, la simulacion no sucede en forma espontanea y, por ello, tanto el investigador como el clinico no deben preocuparse mayormente de ella.


Die Hypnosesimulation: Warum, Entstehung und Bedeutung

Martin T. Orne

Abstrakt: Die Anwendung cines echt-simulierenden Modells mit einem "blinden" Hypnotiseur als "Quasi"=kontrolle in der Hypnoseforschung ist diskutiert. Die Vorzuge und Begrenzungen des Verfahrens werden dargelegt, und die Art der Probleme, fur die es zweckmassig ist, wird im Umriss wiedergegeben. Die Natur der Schlussfolgerungen, die berechtigterweise von Befunden aus der Anwendung dieses Modells gezogen werden konnen, wird behandelt. Es wird darauf hingewiesen, dass, obgleich man zeigen konnte, dass Vpn. durchaus in der Lage sind, erfolgreich zu simulieren und einen hoch trainierten Hypnotiseur tauschen zu konnen, diese Beobachtung jedoch nicht das wahre Erleben der Vp. bestreiten kann, noch die Echtheit der Hypnose in Frage stellt. Weiterhin ensteht die Simulation in vielen Umstanden nicht spontan und sollte daher weder den Therapeuten noch den Forscher ausschliesslich beschaftigen.

The preceding paper is a reproduction of the following article (Orne, M.T. The simulation of hypnosis: Why, how, and what it means. International Journal of Clinical and Experimental Hypnosis, 1971, 19, 183-210.). It is reproduced here with the kind permission of the Editor-in-Chief of The International Journal of Clinical and Experimental Hypnosis.