Damaser, E., Whitehouse, W.G., Orne, M.T., Orne, E.C., & Dinges, D.F. Behavioral persistence in carrying out a posthypnotic suggestion beyond the hypnotic context: A consideration of the role of perceived demand characteristics. International Journal of Clinical and Experimental Hypnosis, 2010, 58, 1-20.

The International Journal of Clinical and Experimental Hypnosis, 2010, LVIII, No. 1, 1-20



Private Practice, Yellow Springs, Ohio, USA


University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA

Abstract: Compliance with a posthypnotic suggestion (PHS) to carry out a specific behavior in a subsequent nonhypnotic setting was investigated in high and medium hypnotizable participants. The target behavior -- solicited by either a PHS given during hypnosis, a waking social request, or both -- was to be performed daily for an unspecified period of time. Findings indicated that the waking request alone yielded a high level of compliance, particularly among medium hypnotizable participants. In contrast, highly hypnotizable participants who received the PHS coupled with instructions for posthypnotic amnesia exhibited considerable variation in responding, whereas high hypnotizables, who received either a waking request, or a combination of PHS and waking request, performed similarly to medium hypnotizables. Postexperimental interview data suggest that perceived demand characteristics may contribute to variation in the persistence of posthypnotic behavior outside the hypnotic context.

Manuscript submitted March 29, 2009; final revision accepted April 8, 2009.

1 This research was supported in part by Public Health Research Grant M-3369 from the National Institute of Mental Health, in part by Contract Nonr3952 (00) from the Office of Naval Research and in part by a grant from the Institute for Experimental Psychiatry Research Foundation. We are grateful to Robert W. White, Ronald E. Shor, Donald N. O'Connell, George Smiltens, Eleanor DeRubeis, Elizabeth L. Scherer, Teresa Lucatorto, and Nancy Leppert for their various contributions to the design and execution of this study.

2 Deceased.

3 Address correspondence to Emily Carota Orne, Unit for Experimental Psychiatry, University of Pennsylvania School of Medicine, 423 Guardian Drive, 1013 Blockley Hall, Philadelphia, PA 19104-6021, USA. E-mail: eorne@mail.med.upenn.edu




The use of hypnotic procedures has demonstrable benefits when employed by trained professionals as an adjunct in the application of diverse psychotherapeutic approaches, ranging from behavior therapy to psychodynamically oriented treatments (e.g., Ascher, 1977; Gibson & Heap, 1991). Gains achieved in the clinician's office, however, often need to be supported in real-life situations, and these, too, can be bolstered by hypnotic techniques, such as training the client in self­hypnosis or the use of a therapeutic posthypnotic suggestion (PHS), that maintain adaptive behavior in everyday circumstances.

Although PHS has long been a well-known phenomenon, very little has been established regarding its limits and little is understood about its mechanism. PHS is here operationally defined as instruction given to a participant in the hypnotic condition requiring the performance of specific behavior at a time subsequent to the termination of hypnosis.

From the participant's point of view, there are two major characteristics of the phenomenon: (a) the experience of a compulsion to perform the suggested behavior and (b) a lack of awareness of the motivational basis of the behavior. However, the latter characteristic is not a necessary condition for either the performance of the behavior or for the feeling of compulsion to carry it out (see LeCron & Bordeaux, 1947, pp. 124-125, for a description).

Apart from various methodological weaknesses, much of the earlier research on PHS assessed the subsequent level of behavioral compliance in the same context in which hypnosis was conducted (e.g., Kellogg, 1929; Patten, 1930). This raises the possibility that responding may have been partially controlled by the participant's perception of situational demand characteristics (Orne, 1962). These refer to subtle cues in the experimental context that convey, essentially, how to behave in a way so as to conform with the experimenter's hypothesis (Orne & Whitehouse, 2000), thereby calling into question whether the behavior was truly and solely a product of hypnotic suggestion. An additional concern is that comparisons of the posthypnotic responding of hypnotic and nonhypnotic groups might be compromised due to systematic differences in the participants' perceptions of the "purpose of the proceedings." To be useful clinically, it is important to examine the durability of posthypnotically influenced behavior beyond the setting in which hypnosis was introduced, where one might expect demand characteristics to be comparatively less salient.

Several studies have attempted to investigate posthypnotic responding outside the context in which the suggestion was given. Orne, Sheehan, and Evans (1968) used a real-simulator design in which high hypnotizables experienced hypnosis but low hypnotizables were instructed by a different experimenter to try to fool the hypnotist into believing they were actually hypnotized. Under these conditions, 5 out of 17 highly hypnotizable participants responded appropriately to the



prearranged PHS cue when administered postexperimentally by a scheduling secretary. In stark contrast, none of the simulators did so. The authors interpreted these results to indicate the compulsive quality that some hypnotizable individuals experience to carry out the suggested response even if outside the experimental context.

Spanos, Menary, Brett, Cross, and Ahmed (1987) postulated that Orne et al. (1968) did not sufficiently differentiate the PHS testing context from the hypnosis setting, as both took place in the same laboratory, and the secretary, although disinterested, was not necessarily unaware of each participant's group assignment and might have inadvertently cued their responses. Thus, Spanos et al. designed a replication real-simulator study with a second simulator condition -- one intended not only to deceive the hypnotist but anyone perceived to be associated with the experiment. Their results document virtually no response to a PHS when student confederates administered the cue in situations divorced from the study, whereas considerable PHS responding occurred in the presence of the hypnotist as much as 1 week later. Spanos and colleagues concluded that enactment of post­hypnotic behavior is strategic and goal directed. Nevertheless, this interpretation does not adequately explain their findings that nonvolition ratings regarding PHS responding between high hypnotizables and low-hypnotizable simulating groups differed significantly when honesty instructions were administered postexperimentally. Thus, the contention of Orne et al. that posthypnotic behavior is experienced by some highly hypnotizable persons as compulsory and automatic would seem not to be discredited by the outcome of this study.

More recently, Barnier and McConkey (1998) adapted the methodology of the current article (derived from Damaser, 1964) to examine some of the parameters of posthypnotic behavior in other contexts subsequent to the administration of a PHS. In one study, high-hypnotizable participants were assigned the task of returning one postcard each day following the experimental session. One group of participants received a PHS with a specified duration: to continue the task until the experimenter contacted them. A second group received a PHS that implied the task should be carried out for an unlimited duration. A final group did not receive a PHS but, upon arousal from hypnosis, received a waking social request to perform the task for an unlimited duration.

The results showed that the group receiving the social request with unlimited duration returned the highest number of postcards over the 16-week testing period, significantly more than either of the PHS groups. Among the latter, those given the unlimited-duration suggestion exhibited consistently the lowest performance, whereas participants in the limited-duration condition initially showed a high level of compliant behavior, followed by a marked deterioration of responding



as the weeks progressed. Those who agreed to a follow-up interview explained their experiences with the task. Generally, participants in the two PHS groups reported a sense of compulsion to complete the task, which was automatic and required little effort. Those in the waking social-request group said they responded because they had agreed to do so, and the act was a conscious one. A second experiment reported by Barnier and McConkey (1998) employed a real-simulator design as well as a waking control group unselected for hypnotizability. Here they obtained essentially the same results, with high hypnotizables who received a PHS returning a higher percentage of postcards than low-hypnotizable simulators who also received a PHS. But the waking control group who received a social request performed comparably to the high hypnotizables. In the follow-up interview, the reasons for responding followed the same themes as in Experiment 1: general feelings of compulsion and automaticity among high hypnotizables, and a commitment to fulfilling a contract with the experimenter among the social-request group; response rates for simulators were quite low as many claimed they either forgot or were not motivated to carry out the task.

Despite the consistency of findings, two aspects of Barnier and McConkey's (1998) study bear further examination. First, although all participants received instructions to mail one postcard every day, the authors do not mention whether some individuals might have mailed several on the same day, thereby failing to comply appropriately with the instruction. Also, on the basis of posthypnotic interviews, the researchers learned that many participants had divulged the nature of their task to family and friends who, in turn, provided assistance in helping them to complete the task or, in some cases, discouraged them from carrying it out. Either way, such participation by others creates a problem for evaluating the endurance of self-initiated posthypnotic behavior outside an experimental or clinical setting.

The study described in the current report is based on the first author's doctoral dissertation (Damaser, 1964). Much pilot testing and careful screening of participants preceded the main study reported below in order to minimize procedures that might obscure interpretation of our primary findings. The principal aim of the research was to evaluate the relative effectiveness of a PHS versus a nonhypnotic request in establishing and maintaining a specified behavioral performance outside the laboratory. In attempting to accomplish this, the design of the study took into account hypnotizability, comparability of prior hypnotic experiences, the convenience and naturalistic features of the target task, stressing the importance of not discussing the experiment with others who might intervene in any way, and extralaboratory demand characteristics. Ancillary objectives were (a) to ascertain whether behavioral performance controlled by a PHS was more



automatic than the same behavior pursuant to a nonhypnotic request and (b) to determine if posthypnotic behavior could be prematurely terminated by instructions given by another credible authority figure or whether its cancellation required the involvement of the hypnotist.



Undergraduate and graduate-student volunteers between the ages of 17 and 28 were recruited from Boston-area colleges and universities to participate in a paid psychological experiment involving hypnosis. Of the original enrolled sample of 182 volunteers, 31 were selected for the final sample on the basis of achieving total scores of 7 to 12 on the Harvard Group Scale of Hypnotic Susceptibility, Form A (HGSHS:A; Shor & Orne, 1962), as well as meeting additional criteria, whereby objective performance during later hypnosis sessions was validated by subjective reports in order to justify assignment to either medium- or high-hypnotizable experimental conditions (see Damaser, 1964, pp. 83-89). Data analyses were based on only 29 participants (15 males, 14 females), because 2 individuals were subsequently disqualified.


In addition to the participant characteristics previously described, general selection criteria included (a) being unmarried or not sharing residence with another person who had been accepted in the study; (b) not being in psychotherapy currently, in the past year, or contemplating therapy during the coming year; and (c) expecting to remain in the local area for at least 3 months after the experimental session. Criterion a was established to reduce the risk that someone might remind the participant to carry out the task targeted by the PHS. Criterion b was included to avoid any possible conflict a participant might encounter with his or her therapist over the issue of participating in a hypnotic experiment. Finally, Criterion c was incorporated to maximize the likelihood that the participant would be available for a postexperimental interview.

The first phase of the study involved hypnotizability assessment with the HGSHS:A in 12 groups, which ranged in size from 2 to 31 participants. Each session involved completing a general information form to provide contact and demographic information, as well as several questionnaires unrelated to the purpose of the present study.

4 Two participants were enrolled in the final sample without the initial screening by the HGSHS:A because they had prior experience in hypnosis in earlier studies conducted by our laboratory.



After this, the experimenter made some general introductory remarks and introduced Dr. Martin Orne, the director of the laboratory, who described the project, answered questions about hypnosis and invited the participants to make appointments to see him if they wished to discuss their experiences with him. The purpose of this invitation was to make it possible for the participants to contact someone (other than the experimenter) who had expertise in hypnosis should they be puzzled or concerned by anything they experienced in hypnosis. This was deemed important as the design of the study attempted to insure that the experimenter would have no contact with the participants from the time the PHS or nonhypnotic request was administered until the final postexperimental interview.

The HGSHS:A was then administered and, at its completion, participants rated their objective responses to each scale item in a response booklet, which included space for a description of any subjective experiences that they felt might be of interest. When this was completed, any additional questions were answered, participants were thanked, and a token monetary reimbursement was given as the session was terminated.

The general information forms were then examined and any participants who did not satisfy the general selection criteria were eliminated -- e.g., married individuals, in psychotherapy, not college students, etc. Objective scores on the HGSHS:A were then considered. Participants who scored between 8 and 12 were contacted and appointments for a follow-up session were made for those who agreed to participate further. For those participants with an objective score of 7 on the HGSHS:A, their subjective comments were taken into account and, if their description implied relatively vivid hypnotic experiences, they were also contacted for further participation.5 Of the original recruitment sample of 182 volunteers, 110 were rejected on the basis of scoring too low (i.e., 0-6) on the HGSHS:A or of not meeting general selection criteria.

Because the HGSHS:A is primarily a screening measure for hypnotic ability, those volunteers who were invited for further participation received additional individual evaluation sessions aimed at more thoroughly assessing and developing their hypnotic potential. Therefore, participants underwent multiple hypnosis sessions -- as many as deemed necessary by the experimenter -- until they either satisfied the hypnotizability requirements or were considered lacking

5 During the early screening phase with the HGSHS:A, there was considerable overlap in the objective and subjective scores of participants ultimately differentiated into high- and medium-hypnotizable categories. The critical distinguishing performance characteristic was the ability of high hypnotizables consistently to maintain suggested amnesia for the events of the hypnosis sessions for a period of 1 week, whereas medium hypnotizables were unable to do so.



the potential to satisfy them, in which case their participation in the study was terminated. An additional 41 participants were discontinued following the individual hypnotic evaluation phase. The remaining 31 participants were eligible and participated in the final experimental phase, but 2 were later disqualified due to technical problems.

The general procedure for all individual evaluation sessions included an induction using the hand-levitation technique followed by a variety of hypnotic suggestions that proceeded from those that are more easily experienced to gradually more difficult items. No specific list of items was adhered to, but each session included a PHS and a suggestion for total posthypnotic amnesia (PHA). The participant was permitted to "sleep" for 2 minutes before being dehypnotized. The hypnotist would then ask the participant how he or she felt, and what the hypnotic experiences were like but made no attempt to probe for memories of the hypnotic events, which might break down whatever PHA existed. During this discussion the experimenter gave the signal for the PHS and observed whether the participant responded. At the end of each individual session, participants were thanked, paid a nominal remuneration and asked not to discuss the experiment with anyone. The experimenter then rated the participants on the objective criteria of hypnotic ability.

Prior to each individual session following the first one, the participant was engaged in an interview conducted by someone other than the experimenter, during which a subjective report of the participant's prior hypnotic experiences was elicited. This allowed for the assessment of the subjective dimensions of the participant's experience during hypnosis (e.g., how vivid or real the suggested experiences were) as well as whether the suggestion for PHA had remained intact for a full week, which would be used in conjunction with the hypnotist's objective criteria for final classification of individuals into medium- or high-hypnotizable conditions. As soon as the participant satisfied all the requirements for either hypnotizability category, he or she was admitted to the study, and the experimental session was held. A total of two individual sessions (including the experimental one) was needed by 17 volunteers; 7 needed a total of three, and 5 needed a total of four.

The experimental session, during which the postcard instructions were given, was conducted in a manner similar to the previous sessions. The specific suggestions utilized were purposively not identical to the set of suggestions used in the previous session. According to the participant's predetermined group assignment, the PHS and/or nonhypnotic request was given. The PHS was given as the last item for the medium-hypnotizable participants and was the next-to-last hypnotic item for the high-hypnotizable participants-amnesia always



being the last for the latter group. Amnesia was not induced in the medium hypnotizables during the experimental session so that they would not have a sense of failure in attempting to comply with the experimenter's suggestions. Each participant was allowed to "sleep" for a period of 2 minutes prior to being aroused from hypnosis. If the nonhypnotic request was to be given, this was done some time after the participants were alerted from the hypnotic condition, when their behavior and interaction with the experimenter was clearly "awake." No discussion took place about the hypnotic session.

The experimenter placed a set of 150 postcards, bound with a rubber band, on the desk next to the participant before arousal from hypnosis. Thus, the participants saw the postcards for the first time when they opened their eyes upon awakening. The postcards remained there until the participant removed them. For those individuals who received only the PHS, no discussion whatsoever took place about the postcards after hypnosis was terminated. For those receiving the non­hypnotic request, the only mention of the postcards was the instructions regarding them. No participant failed to take the postcards before leaving the laboratory and everyone was asked not to talk about the experiment until they received a summary of the purpose and findings of the study in the mail.

An abbreviated version of the PHS follows:

I am going to give you some instructions now. I am going to give you some postcards; they are on the desk. After you wake up and before you leave this room you will take the postcards with you, bring them home, and put them in a safe place. They are already addressed. You are to mail one postcard every day, but you must mail it before 8:00 p.m. You will mail one a day, no more, no less. You will do this every day, including Sunday, and you will start tomorrow. If you travel you will take enough postcards with you to mail one a day.

If you absolutely cannot get to a mailbox; for example, if you are very ill, you will save the postcard for that day and mail it as soon as you can along with the next day's postcard. You will not feel bad that you have not mailed a postcard that day, but you will mail it as soon as you can along with the next day's postcard.

Another very important thing is that you will not tell anyone what you are doing, or talk to anyone about this experiment.

The abbreviated nonhypnotic request follows:

This is a very important part of the experiment. I'm going to ask you to do something every day for a long time. This won't take much of your time each day, and you can do it at home, that is, you needn't come down here. It is really very important that you do this. This is what it is. Here are some postcards. You are to take the postcards with you, bring them home, and put them in a safe place. They are already addressed. Your part is to mail one postcard every day, but you must mail it before



8:00 p.m. You should mail one a day, no more, no less. You should do this every day, including Sunday, and you should start tomorrow. If you travel take enough postcards with you to continue to mail one a day.

However, if you absolutely cannot get to a mailbox; for example, if you are very ill, you should save the postcard for that day and mail it as soon as you can along with the next day's postcard.

It is also very important that you do not tell anyone about this experiment.

Assignment of participants to experimental conditions was dictated by their demonstrated level of hypnotizability but otherwise followed a pseudorandom sequence, in which the first slot to be filled was the PHS condition, followed by the nonhypnotic request condition, and then the PHS + nonhypnotic request condition as qualified candidates were identified. A few exceptions to this scheme occurred in an effort to balance gender representation among the groups.

As noted earlier, both the PHS and nonhypnotic request instructed participants to begin mailing postcards "tomorrow" (i.e., the day after the experimental session), which, for data collection purposes, was designated "Day 1." On Day 69, a letter was mailed from the director of the research project thanking the participants for their help in the study and informing them that the study was now completed and that they could therefore stop mailing postcards. This served as the test of whether the PHS could be cancelled by someone other than the hypnotist. The letter further conveyed that the experimenter would contact them soon for a final follow-up interview to inquire about their experiences.

Participants were contacted on Day 85 to schedule appointments for the final interview. Twenty-four participants took part in the final interview in person. Five individuals from the original sample had moved out of the area, but all completed and returned a questionnaire version of the interview. The final interview began with a written free recall of all the hypnotic experiences participants could remember with regard to the study. Then they were asked about their perceptions of the purpose of the study, their memories of the postcard instructions, their experiences with the postcard task, their interpretations of the letter from the project director, and any general impressions and feelings about the study. All participants were offered a final opportunity to be hypnotized, during which they would be instructed to recall any previous hypnotic experiences that they wished to recall and would have the PHS removed, if one had been given. Those participants who declined to be hypnotized again were instructed that they no longer needed to mail any postcards connected with this study and were thanked for their help with the study.

To determine compliance with the postcard task, each participant was assigned a unique six-digit identification number that was visibly



stamped on the message side of each of the 150 postcards in his or her possession. In addition, each set of 150 postcards was consecutively numbered from 1 to 150 in ink visible only under an ultraviolet light source. This "invisible" numbering system was intended to provide a means to examine an aspect of "automaticity" of responding -- the extent to which the postcards were returned in consecutive order.


The originally planned dependent variables were: total and corrected number of postcards mailed prior to the project director's letter (for corrected number, postmarks were checked to assure that multiple postcards were not mailed on the same day, contrary to instructions); automaticity (i.e., a percentage of any discontinuities in the consecutive order of the invisible numbers of the postcards received during Days 1 through 69); duration of behavior until first contact (i.e., number of days inclusive between Day 1 and the date of postmark of the last postcard received, not to exceed Day 69); and total and corrected number of postcards mailed between receipt of the letter from the project director and contact for the final interview (Days 72 through 84) -- a test to determine if the PHS could be effectively terminated by an authoritative figure other than the hypnotist. However, because the total and corrected number of postcards mailed both prior to and following the project director's letter were so highly correlated (rs = .950 and .999, respectively, both ps < .001), and the same was true for the association between corrected number of postcards mailed and duration of behavior (r = .944, p < .001), it was decided to limit analyses to the corrected and automaticity measures, which did not display such redundancy (r = .103, n.s.). Thus, Table 1 provides group summary statistics for each of these remaining primary dependent measures.

Planned contrasts were carried out to address the issues of interest. With regard to the extent of behavioral compliance following a PHS versus a nonhypnotic waking request, no reliable group difference was evident for the high-hypnotizable participants, t(7) = -.85, n.s., but a significant effect emerged for medium-hypnotizable participants, t(9) = -3.978, p = .006, favoring greater persistence in mailing postcards among those given the waking social request to do so. Also of interest was the effect of a PHS reinforced by a waking nonhypnotic request to mail the postcards compared to the nonhypnotic request alone. Here again, among high hypnotizables, there was no statistically significant difference in the performance of the two groups, t(6) -.062, n.s.; that is, the combination of PHS and subsequent waking request led to a level of response outside the experimental context that was similar to that derived solely from a waking social request. The same was true among medium-hypnotizable individuals, t(9) = .940, n.s.




Despite the finding that medium hypnotizables exhibited a strong compliance with the nonhypnotic request, it is clear that preceding the waking request with a PHS given during hypnosis did nothing to enhance or diminish its magnitude. Indeed, the fact that the PHS plus waking-request groups performed similarly overall to the waking­request-only groups, but not to the PHS groups, suggests that their conscious motivation to engage in the task (i.e., to honor a social agreement) might also be comparable.

In assessing the question of whether performance of the target behavior consequent to a PHS was any more "automatic" than that following a nonhypnotic request, we analyzed the percentage of discontinuities in consecutively (and invisibly) numbered mailed postcards in terms of dates of receipt. One of these analyses employed a factorial analysis of variance (ANOVA) with hypnotic ability (high vs. medium), nature of request (PHS vs. waking), and their interaction as factors. None of the main effects nor their interaction was significant (all Fs < 1.0, n.s.). A second analysis treated each experimental condition as a separate independent group in order to include the high- and medium-hypnotizable participants who received both the PHS and nonhypnotic waking request. The corresponding one-way ANOVA also yielded a nonsignificant outcome, F(5, 28) = .546, n.s. Thus, there was no evidence of any difference in automaticity of response for either medium or highly hypnotizable individuals administered a PHS versus a waking request or a combination of the two forms of instruction.

The identical analytic plan was adopted to determine the effect upon continued persistence of the target behavior brought about by instructions to terminate responding embedded in the project director's letter. Although the results of the factorial and one-way ANOVAs were uniformly nonsignificant (all Fs < 1.0, n.s.), an inspection of the summary statistics in the rightmost column of Table 1 reveals that nearly all participants had stopped mailing postcards prior to receipt of the letter, thereby rendering any test of this question inconclusive.

Given the unexpected finding that the only significant intergroup difference pertained to the level of behavioral response between medium-hypnotizable participants who received a PHS during hypnosis and those who received only a waking request to carry out the postcard task, we proceeded to study the qualitative information obtained during the final postexperimental interviews. A primary impetus for this strategy was the observation, as can be ascertained by an examination of Figure 1, that there was a greater idiosyncratic pattern of performance variability among the high-hypnotizable participants compared to their medium-hypnotizable counterparts. One striking, although certainly not solitary, example is the performance of the high- and medium-hypnotizable participants in the PHS groups. The effect of a PHS clearly produced a variable response in



the former group in the sense that some amnesic high hypnotizables performed the task for a relatively long period of time, others for a relatively short period of time, and others for a moderate length of time. In contrast, 4 out of the 5 nonamnesic medium-hypnotizable participants performed the task for a relatively short period of time, whereas only 1 individual performed in a markedly deviant manner by engaging in the task consistently for a long period of time. The tendency for medium-hypnotizable participants generally to perform in a comparable manner is evident in the other experimental conditions as well in marked contrast to the disparate performances of the high-hypnotizable participants.

The interview material reveals that 5 of 13 highly hypnotizable participants were either under the impression that they were given instructions in hypnosis that in fact had not occurred or misconstrued the actual instructions given them during hypnosis.

Among those in the high-hypnotizable category given the PHS, 1 participant stated:

I sent [postcards] for several days and then thought I was supposed to stop.... I just stopped. I never attached too much reason to any of it. I figured it was something you told me to do.... I thought I wasn't supposed to mail any more.... I was trying to figure out why I had such a big pile. I had a stronger feeling that I shouldn't send them in. I figured you'd let me know, give me more instructions.... [Experimenter: What do you think this desire to mail stems from?] I guess from something you told me when I was under hypnosis. [Experimenter: Why stop mailing?] I guess you must have told me that too.



Another individual in the same group was under the impression that there were some days on which he was not supposed to mail a postcard:

I knew I was supposed to mail one a day but thought maybe there were some days I was supposed to not send one in. Then the next day I'd mail two but I don't know why.

Yet another high-hypnotizable PHS participant was under the impression that she had been given instructions while in hypnosis to mail postcards for only 1 week:

I guess now that I didn't remember the instructions accurately -- because I thought you had said to send one card a day for just 1 week. After I left the project for the last time I couldn't figure out why you had given me so many cards.... I am pretty sure that you said to send them only for 1 week.

Some confusion also seemed to occur among highly hypnotizable participants who received amnesia suggestions for the events of the hypnosis session, even though the instructions about postcards were only delivered following hypnosis, in the waking condition.

One participant recounted:

[I was] terribly confused about the postcards -- the experiment in hypnosis was o.k., but the postcards -- I wondered if it had any bearing on hypnosis, or if you had told me in hypnosis, or if I was supposed to have this driving urge to get up and mail a postcard each day.

Similarly, another participant in the high-hypnotizable condition who received only the waking request to mail postcards following the session entertained the idea that he should not mail the postcards:

I just kept forgetting to mail them -- it annoyed me very much because I had every intention of mailing [them]. I'd look at them and intend to mail them but just didn't mail them -- intended every day to mail one and then just forget ... did not consciously decide to stop -- I just stopped thinking about it. [Experimenter: Why?] I don't know. I had the idea that possibly you had suggested to me under hypnosis that I would not mail the postcards -- but it was hard to believe that after all this time the suggestion would last, so I discounted it.

In contrast to these conjectures and misconstructions about what took place during hypnosis on the part of high-hypnotizable participants, only 1 medium-hypnotizable participant falsely surmised about what had occurred while he was experiencing hypnosis, and this supposition was a direct result of the experimenter's questions. Consistent with the interview protocol, several questions were asked regarding the participant's recall of the task instructions in the hypnotic and waking condition. (However, being in the nonhypnotic request group, the postcard task was never mentioned in hypnosis.) After stating



several times that he recalled nothing about postcards while experiencing hypnosis he said:

Now that you brought up the subject, you must have told me while I was under hypnosis unless you felt it would become instinctive ... now that I gather that you asked me under hypnosis.


This study was conceived and executed during the mid-1960s as an early attempt to understand the durability of a PHS when extended beyond the context in which it was originally administered. Other researchers have since adopted variants of this same methodology (e.g., Barnier & McConkey, 1998; Hoyt, 1990) with outcomes that encourage further investigation of the value of PHS for clinical treatment.

The data reported in this study appear to highlight two issues. One that is difficult to ignore is that a waking social request by a perceived authority or professional can be highly effective in engaging appropriate behavioral compliance. This conclusion is supported by the finding that medium hypnotizables given a simple waking request or the relevant PHS supported by a subsequent waking request responded at comparable, and generally high, levels throughout the study (a nonsignificant trend in this same direction was observed among high­hypnotizable participants). Averaging across hypnotizability categories in Table 1, the mean postcard-return rate for the waking request and PHS plus waking request conditions combined was slightly more than double the rate for PHS alone (i.e., 44 vs. 20.5). In contrast, the PHS had variable influence on the behavior of high hypnotizables who also received an instruction for PHA and, with a single exception, resulted in minimal compliance among medium-hypnotizable participants, none of whom received the PHA suggestion.

This raises the issue of the source of the overall variability in performance on the part of high-hypnotizable participants, given a PHS, to carry out the target task. Certainly, apart from differences in hypnotic capacity, a major procedural feature that distinguished the high and medium hypnotizables was the suggestion for total amnesia for the events of the hypnosis session: highly hypnotizable individuals, capable of sustaining amnesia for a minimum of 1 week, were given the amnesia suggestion, whereas medium-hypnotizable individuals were not.

We considered the possibility that, because the suggestion for PHA followed the PHS for high-hypnotizable participants during the experimental session, some individuals might have experienced difficulty responding to the PHS instructions outside the experimental context due to the countermanding influence of the amnesia suggestion. While this is certainly a reasonable hypothesis, its likelihood would seem to be



undermined by the fact that all 5 participants in this group left the experimental session with a stack of 150 postcards, despite no further discussion about them having taken place following the PHS administered during hypnosis. Furthermore, it would be difficult to explain why every member of the high-hypnotizable PHS group returned postcards following the experimental session (range = 4-60 days) if the PHA suggestion had annulled the PHS. Finally, a post hoc comparison of mean corrected postcard returns by members of the high-hypnotizable PHS group, who received the PHA suggestion during hypnosis, with their counterparts in the medium-hypnotizable PHS group, who were treated identically except for the absence of the PHA suggestion, yielded no difference in overall group performances, t(8) = .446, n.s. Thus, a direct inhibitory effect for PHA on the compulsion to carry out the PHS among highly hypnotizable participants lacks empirical support in this study.

On the other hand, the existence of amnesia for the events of hypnosis among participants in the high-hypnotizable groups may have served to obscure the demand characteristics of the study, such that a good deal was left to individual interpretation. With regard to most laboratory experiments, some of the demand characteristics are conveyed by the experimental procedures that participants undergo. The major portion of participation in the current study took place in the hypnotic condition. Given that high-hypnotizable participants were selected, in part, for their ability to sustain PHA for at least a week, they could, in the waking state, recall little of what had occurred during the course of the entire study. Hence, the demand characteristics they perceived were presumably based only on whatever cues they received in the nonhypnotic condition and on whatever recollections and vague impressions they had of what occurred during hypnosis. It is certainly possible that they might conjecture about actual events during hypnosis, and that these conjectures would not be the same for all participants. Thus, there might be a good deal of room for large individual differences in the perceived demand characteristics of the highly hypnotizable participants. If the perceived demand characteristics affected performance on the task, and there were large individual differences in the perceived demand characteristics of the high-hypnotizable participants across groups, then this may partially account for the large variability in performance on the postcard task for all three high-hypnotizable groups and might explain the lack of consistent responses to experimental conditions.

In addition to "obscuring" the demand characteristics (in the sense of permitting a good deal of speculation and the individual variability that accompanies such speculation), the existence of amnesia and the consequent conjecturing also might facilitate "rationalization" according to the predisposition of the participant. For example, participants who may have found the postcard task to be a chore and a burden could more easily



rationalize discontinuing the task by postulating that they had been given instructions in hypnosis to carry out the task for only a short period. Without the existence of amnesia it might be more difficult to rationalize and dispense with any feeling of an obligation to the experiment.

Not all high-hypnotizable participants maintained total amnesia for the events of hypnosis up to the time of the final interview. Since it might be argued that recall of the task instructions (among those participants who received these instructions in the hypnotic condition) is associated with performance on the postcard task, this question was pursued in the interview material. However, the interview data revealed no significant relation between the maintenance of PHA and the recall of the task instructions and scores obtained on the postcard measures (see Damaser, 1964, for greater detail). At the same time, no high-hypnotizable participant was able to recall the events of hypnosis in their entirety, and thus there was always room for speculation about what had actually occurred.

Our findings that moderately hypnotizable individuals tended to respond more consistently to the nonhypnotic request than did highly hypnotizable participants are consistent with earlier observations by Kellogg (1929) and Patten (1930). It is important to keep in mind, however, that these earlier studies did not control for hypnotizability among hypnotic and nonhypnotic groups and, thus, may have over­sampled moderately hypnotizable individuals for their nonhypnotic conditions, rendering their findings difficult to generalize to the higher range of hypnotic ability.

The principal findings of the present experiment suggest an important role for perceived demand characteristics in the enactment of post­hypnotic behavior. Admittedly, the unbalanced experimental design -­ in which only high-hypnotizable participants received the additional suggestion for PHA -- was instrumental in focusing our attention on postexperimental interview data. Although a number of explanations for these results may be possible, the prevailing themes derived from the interviews were (a) participants in the PHS group perceived their task as performing the suggested behavior for as long as they felt compelled to do so, whereas (b) recipients of a nonhypnotic request perceived their task as performing the requested behavior for as long as they possibly could. As Barnier and McConkey (1998) have suggested, such differences in subjective experience have important clinical implications. Whereas the overt behavioral indices appear to be similar, the advantage of a PHS may reside in its capacity to modify underlying incentives as well as the degree of effort directed towards initiating and maintaining a particular goal strategy.

No significant differences among groups were found for the measure of automaticity (i.e., sending postcards according to their scheduled dates). The absence of such an effect, however, need not imply that the execution of posthypnotic behavior does not often or



even typically occur in an automatic fashion without conscious control. The failure to find evidence for automaticity in this study may be due to inherent task features (e.g., postcards were necessarily arranged in ordered stacks) or to the insensitivity of the measure itself. Likewise, no significant intergroup differences were detected in the number of postcards sent after receipt of the letter from the project director announcing the end of the data collection phase of the study. The purpose of this manipulation was to determine if a perceived expert professional other than the hypnotist could cancel a PHS. However, by the time the letter terminating the study was scheduled to be delivered, all but 5 participants had already stopped mailing their postcards. Thus, these remain among a set of important questions whose ultimate resolution will require additional research characterized by larger sample sizes and designed to circumvent the limitations associated with the current study and others seeking to evaluate the fate of a PHS beyond the immediate hypnotic setting.

How might the present findings extrapolate to the clinical context? When the incorporation of PHS as a therapeutic adjunct is deemed the treatment of choice for a suitable client, its effectiveness may well reside in its ability to strengthen the individual's motivation to achieve self-control over specific behaviors. However, it is clear that the demand characteristics would have to be carefully explored so that the client would not rationalize a justification to give up trying if it happened that modifying the behavior began to require conscious, waking, self-initiated reminders rather than occurring effortlessly and apparently without volition. Thus, it should not be assumed that a therapeutic PHS would necessarily be helpful in bringing about a desired clinical outcome without the alliance of appropriately supportive demand characteristics.


Ascher, L. M. (1977). The role of hypnosis in behavior therapy. Annals of the New York Academy of Sciences, 296, 250-263.

Barnier, A. J., & McConkey, K. M. (1998). Posthypnotic responding away from the hypnotic setting. Psychological Science, 9, 256-262.

Damaser, E. (1964). An experimental study of long-term post-hypnotic suggestion. Unpublished doctoral dissertation, Harvard University, Cambridge, MA.

Gibson, H. B., & Heap, M. (1991). Hypnosis in therapy. Hove, UK: Lawrence Erlbaum.

Hoyt, I. P. (1990). Posthypnotic suggestion versus ordinary instruction: Compliance and attention. Unpublished doctoral dissertation, University of Wisconsin, Madison.

Kellogg, E. R. (1929). Duration of the effects of post-hypnotic suggestion. Journal of Experimental Psychology, 12, 502-514.

LeCron, L. M., & Bordeaux, J. (1947). Hypnotism today. New York: Grune and Stratton.

Orne, M. T. (1962). On the social psychology of the psychological experiment: With particular reference to demand characteristics and their implications. American Psychologist, 17, 776-783



Orne, M. T., Sheehan, P. W., & Evans, F. J. (1968). Occurrence of posthypnotic behavior outside the experimental setting. Journal of Personality and Social Psychology, 9, 189-196.

Orne, M. T., & Whitehouse, W. G. (2000). Demand characteristics. In A. E. Kazdin (Ed.), Encyclopedia of psychology (pp. 469-470). New York: Oxford University Press.

Patten, E. F. (1930). A note on the persistence of hypnotic suggestion. Journal of Abnormal and Social Psychology, 25, 319-333.

Shor, R. E., & Orne, E. C. (1962). Harvard Group Scale of Hypnotic Susceptibility, Form A. Palo Alto, CA: Consulting Psychologists Press.

Spanos, N. P., Menary, E., Brett, P. J., Cross, W., & Ahmed, Q. (1987). Failure of posthypnotic responding to occur outside the experimental setting. Journal of Abnormal Psychology, 96, 52-57.


Verhaltenspersistenz bei der Ausfuhrung posthypnotischer Suggestionen uber den Hypnosekontext hinaus: Eine Betrachtung der Rolle von wahrgenommenen Demand Characteristics

Esther Damaser, Wayne G. Whitehouse, Martin T. Orne, Emily Carota Orne und David F. Dinges

Zusammenfassung: Die Kompleanz mit der posthypnotischen Suggestion (PHS), ein spezifisches Verhalten in einer spateren, nicht-hypnotischen Situation auszufuhren, wurde bei hoch und gering suggestiblen Teilnehmern untersucht. Das Zielverhalten -- entweder unter Hypnose als posthypnotische Suggestion gegeben, im Wachzustand als soziale Aufforderung, oder beides -- sollte taglich uber einen nicht festgelegten Zeitraum hinweg ausgefuhrt werden. Die Befunde zeigen, dass allein die Aufforderung im Wachzustand ausreichte, um hohe Kompleanz herbeizufuhren. Dies galt insbesondere fur mittelgradig hypnotisierbare Teilnehmer. Demgegenuber wiesen hoch hypnotisierbare Teilnehmer, die die PHS in Verbindung mit Instruktion zur posthypnotischen Amnesie erhielten, eine hohere Variation im Verhalten auf. Hoch Hypnotisierbare, die im Wachzustand instruiert worden waren oder eine Kombination aus PHS und Aufforderung im Wachzustand erhielten, verhielten sich den mittelgradig Hypnotisierbaren vergleichbar. Postexperimentelle Interviews legen nahe, dass die wahrgenommenen Demand Characteristics zur Variation der Bestandigkeit des posthypnotischen Verhaltens auBerhalb des Hypnosekontext beitragen.


University of Konstanz, Konstanz, Germany

Persistance comportementale de l'effet d'une suggestion posthypnotique au-dela du contexte hypnotique : Un examen du role des caracteristiques d'une demande percue

Esther Damaser, Wayne G. Whitehouse, Martin T. Orne, Emily Carota Orne et David F. Dinges

Resume: Les auteurs de cet article ont etudie l'adhesion a une suggestion posthypnotique (SP), consistant a demontrer un comportement particulier dans un contexte subsequent non hypnotique par des sujets moyennement hypnotisables ou hautement hypnotisables. Le comportement cible --sollicite



soit par SP faite durant l'hypnose, soit par demande faite a l'etat d'eveil, ou les deux -- devait etre adopte chaque jour pendant une duree non precisee. Les resultats ont indique que seule la demande faite a l'etat d'eveil avait obtenu un haut degre d'observance, particulierement chez les sujets moyennement hypnotisables. En revanche, les sujets hautement hypnotisables ayant recu la SP assortie d'instructions d'amnesie post-hypnotique ont affiche de grandes variations de comportement d'un individu a un autre, alors que le comportement de leurs homologues hautement hypnotisables, qui avaient recu soit une demande faite a l'etat de veille, soit une combinaison de SP et de demande faite a l'etat de veille, etait tres semblable a celui des sujets moyennement hypnotisables ayant recu les memes instructions. Les donnees obtenues lors des entrevues postexperimentales indiquent que les caracteristiques d'une demande percue peuvent contribuer a une variation dans la persistance du comportement posthypnotique hors du contexte hypnotique.


C. Tr. (STIBC)

Persistencia conductual en seguir una sugestion posthipnotica fuera del contexto hipnotico: El papel de las caracteristicas percibidas de demanda

Esther Damaser, Wayne G. Whitehouse, Martin T. Orne, Emily Carota Orne, y David F. Dinges

Resumen: Investigamos en participantes con hipnotizabilidad media y alta el cumplimiento de una sugestion post-hipnotica (SPH) para llevar a cabo una conducta especifica en un contexto posterior no hipnotico. La conducta especifica -- solicitada ya fuera por una SPH, una solicitud social durante vigilia, o ambas -- seria llevada a cabo diariamente durante un periodo indeterminado. Los resultados indicaron que la solicitud durante vigilia sola dio un alto nivel de cumplimiento, en particular entre los participantes medianamanete hipnotizables. En contraposicion, los participantes muy hipnotizables que recibieron la SPH, junto con instrucciones de amnesia post-hipnotica, exhiberon una variacion considerable en la respuesta, en tanto que los muy hipnotizables que recibieron la solicitud de vigilia, o una combinacion de SPH y la solicitud de vigilia, tuvieron un desempeno similar a los medianamente hipnotizables. Los datos de la entrevista postexperimental sugieren que las caracteristicas de demanda percibida pueden contribuir a la variacion en la persistencia de la conducta posthipnotica fuera del contexto hipnotico.


Lund University, Lund, Sweden

The preceding paper is a reproduction of the following article (Damaser, E., Whitehouse, W.G., Orne, M.T., Orne, E.C., & Dinges, D.F. Behavioral persistence in carrying out a posthypnotic suggestion beyond the hypnotic context: A consdieration of the role of perceived demand characteristics. International Journal of Clinical and Experimental Hypnosis, 2010, 58, 1-20.). It is reproduced here with the kind permission of the Editor-in-Chief of the International Journal of Clinical and Experimental Hypnosis.