Perry, C. W., Evans, F. J., O'Connell, D. N., Orne, E. C., & Orne, M. T. Behavioral response to verbal stimuli administered and tested during REM sleep: A further investigation. Waking and Sleeping, 1978, 2, 35-42.

Behavioral Response to Verbal Stimuli Administered and Tested during REM Sleep : A Further Investigation"

Campbell W. Perry 3, Frederick J. Evans, Donald N. O'Connell 4, Emily Carota Orne, Martin T. Orne

Institute of Pennsylvania Hospital and University of Pennsylvania

Key words: Behavioral response, EEG, Hypnotic Susceptibility, REM Sleep, Response Latencies, "Sleep-Sleep" Model, SPR, Verbal Stimuli.

Abstract : The present study sought to determine whether a sleeping subject can differentially respond behaviorally to meaningfully related stimulation and irrelevant stimuli. It attempted to extend the findings of previous studies by evaluating the sleeping person's ability to discriminate as opposed to merely recognizing the presence of verbal stimuli.

Previous studies reported a relationship between sleep response and hypnotizability, but suggested that the apparent relationship might be an artifact of differential rapport between subjects of differing susceptibilities and the laboratory. The present study, using subjects of different levels of hypnotic susceptibility who all had considerable laboratory experience, found no relationship between hypnotizability and sleep response, thus lending strong support to the rapport hypothesis.

It was found that the latencies of correct responses to meaningful stimuli were appreciably slower than both the latencies of responses to irrelevant stimuli and of incorrect responses to relevant stimuli, suggesting, as in previous studies, slow processing of information and/or slow mobilization of response mechanisms. The responses that were observed were less frequent than in previous studies; as before, they occurred in the absence of alpha activity during both the suggestion and the cue word presentation.

Upon awakening, subjects had no recall of the verbally presented material, showed no SPR responses to the cue words when presented in a word association test, and could not remember responding. The evidence further indicated that responses were causally related to the verbal instructions and could be distinguished empirically from random responsivity.


Until quite recently, sleep was thought of as a unitary state, in which the individual is cut off from external stimuli and in which his cognitive activities are almost exclusively self-generating (e.g., dreams). There was some recognition that external stimuli might affect sleep, and particularly dreaming, but these were considered to be negligible influences, and the cognitive implications were, for the most part, minimized. Following the work of Aserinsky and Kleitman (1953), the multiphasic nature of sleep became more generally recognized. Gradually there developed an awareness that there might be a very subtle and complex interaction between the sleeping person and his environment -- an interaction which might affect, significantly, the form and content of otherwise spontaneous sleep mentation.

Studies seeking to clarify the nature of such transactions between a sleeping subject and the environment have either used the wake-sleep model (stimuli administered while awake and tested during sleep) (Berger, 1963; Oswald, Taylor and Treisman, 1960; Williams, Morlock and Morlock, 1966; Zung and Wilson, 1961) or the sleep-wake model (stimuli administered during sleep and tested during subsequent morning awakening) (Beh and Barratt, 1965; Emmons and Simon, 1956; Levy, Coolidge and Staab, 1972; Simon and Emmons, 1955; 1956 (a); 1956 (b)). The sleep-sleep model (stimuli administered and tested during sleep) has been less frequently employed (Beigel, 1959; Dement and Wolpert, 1958). Most studies have used simple tonal or verbal stimuli and

1 This study was supported in part by a grant from the Institute for Experimental Psychiatry and by grant # AF-AFOSR-707-67 from the Air Force Office of Scientific Research (AFSC), United States Air Force. The authors wish to thank Eugene Cogan, Charles H. Holland, Edgar P. Nace, Ulric Neisser, David A. Paskewitz, and JoAnne L. Withington for their invaluable assistance in preparing this manuscript. Special appreciation is due to Mary McElroy, Luise Miller, William Mitchell, Del Schmeidler, Stanley Silverstein, George Smiltens, and Calvin Stafford for technical assistance during various stages of the investigation

2 Reprint requests to Unit for Experimental Psychiatry, The Institute of Pennsylvania Hospital, 111 North 49th Street, Philadelphia, PA 19139 (Dr. Evans)

3 Now at Concordia University, Montreal

4 Deceased




36 C.W. Perry et al.: Responsivity in REM Sleep

have measured responsivity by changes in electrodermal activity, appearance of evoked k-complexes in the EEG record, or have examined various measures of dream incorporation. These studies have been reviewed elsewhere. (Cobb, Evans, Gustafson, O'Connell, Orne and Shor, 1965).

There has been relatively little experimental investigation of the possibility that complex information processing can occur during EEG-defined sleep and be reflected in behavior. A recent series of studies (Cobb et al., 1965; Evans, Gustafson, O'Connell, Orne and Shor, 1966, 1969, 1970; Evans, 1972) suggests that sleeping subjects can respond behaviorally to meaningful verbal instructions. Studies in this series have all employed a similar approach. During EEG-defined stage REM sleep, verbal instructions or "suggestions" were administered to the subject, requiring him to react in a specified manner when a particular cue word was presented. A typical suggestion was: "Whenever I say the word 'itch,' your nose will feel itchy until you scratch it." Subsequently, only the cue word "itch" was presented during the same and again during later stage REM epochs, and the presence or absence of the appropriate behavioral response to the cue (e.g., nose scratching) was noted.

In the first study using this "sleep-sleep" paradigm (Cobb et al., 1965), four subjects who were highly susceptible to hypnosis responded consistently to such verbal instructions, whereas four insusceptible subjects did not respond. A subsequent study (Evans et al., 1966, 1969, 1970) attempted to eliminate the influence of some extraneous presleep variables present in the original study. The potential effects of differential rapport or prior hypnotic experience were controlled by conducting independent hypnotic evaluations only after the subject had completed the study. These controls resulted in a somewhat diminished rate of response to cues, but it was still found that subjects responded on many occasions without showing any of the traditional signs of arousal in the period between stimulus presentation and response. Upon awakening, they did not recall their behavior during sleep. However, subjects were able to respond to the cue word on subsequent nights without readministration of the suggestion and without any evidence of intervening waking memory of the suggestion.

As in the original study (Cobb et al., 1965) a relationship between sleep response and hypnotic susceptibility was observed, though it proved to be complex. Hypnotic susceptibility was unrelated to the percentage of correct responses but was related to response frequency which, in turn, was a function of how well the subject had slept. Insusceptible subjects tended to be awakened more often by cue stimulations; hence they received fewer cues and consequently responded less frequently than susceptible subjects. This finding suggests the possibility that rapport with the laboratory, rather than hypnotizability itself, was the crucial mediating variable.

The Present Study

The responsivity in sleep to meaningful verbal stimuli observed in previous investigations may be influenced by a number of inter-related factors. Accordingly certain variables were given particular emphasis in the present study.

(a) Presleep Set. -- Notwithstanding the effort to eliminate the influence of presleep set in the previous study, (Evans et al., 1969; 1970) it is possible that a set that "something would occur during sleep" was established and could have been communicated among subjects, especially as subjects in the previous study had all been student nurses who resided in the same dormitory. Further, the experimenter administered suggestions and cues from the bedside. To prevent alarming subjects who might be inadvertently awakened by this procedure, they were informed before retiring that the experimenter might enter the bedroom to check on their sleep. For the present study, therefore, subjects were selected who had little possibility of knowing each other. Suggestions and cues were administered from the adjoining room through an intercom system and responses were observed in total darkness using an infra-red vidicon system.

(b) Random Behavior.-- Subjects make many spontaneous body movements in the course of a night's sleep, (Oswald, Berger, Jaramillo, Keddie, Olley and Plunkett, 1963) some of which are similar to the criterion responses. In the present study, responses were compared on occasions where they were appropriate and on occasions where they were inappropriate to a cue that had been administered. This provided a means of testing whether the sleep response is an artifact of body movements which occur spontaneously throughout the night, independent of suggestion and cue words.

(c) Relationship of Cue Words to Suggestions. -- In all studies that employ a sleep-sleep model, stimulus registration is confounded with responsivity. There is no independent way of determining whether a cue word has actually registered in relationship to an antecedent suggestion to respond other than the presence of appropriate responses. To test the possibility that appropriate responses might simply be the result of the cue words themselves, rather than the instructions to respond to cue words, all cue words in the present study were first presented during stage REM sleep prior to the administration of any suggestions. In other words, would the cue word "itch" lead to scratching of the nose in preference to other parts of the body even prior to the suggestion that the subject scratch his nose whenever he heard the word


37 C.W. Perry et al.: Responsivity in REM Sleep

"itch?" Appropriate responses would clearly imply that the suggestions were unnecessary and that the cue words themselves provided sufficient information to elicit the appropriate response.

(d) The Recognition of Appropriate Cues.-- In the previous studies of this series, responsivity in sleep was assessed in a simple recognition situation. It is possible that once a suggestion had been administered, any verbal stimulus that is similar to the cue word would elicit the appropriate response. To test this possibility in the present study, "dummy" cue words were interspersed among real cue word testings following the administration of suggestions. If appropriate responses were to occur to "dummy" cue words, the same responses to actual cue words could not be attributed to the registration of the suggestions.

(e) The Effect of Hypnotizability. -- Some of the previous studies have suggested that differential rapport established with high and low susceptible subjects, rather than hypnotic susceptibility itself, may account for greater sleep responsivity among highly susceptible subjects. The present study sought to equalize rapport effects by using only subjects -- of all levels of hypnotizability -- who had considerable previous experimental laboratory experience. The sleep experimenter and technician, however, had never worked with the subjects before and had no knowledge of their degree of hypnotizability. 5

Accordingly the present study addressed itself specifically to the following questions:

(i) Can it be shown that a suggestion, given while the subject is asleep without eliciting EEG evidence of arousal, will affect subsequent sleep behavior despite tight control over the physical, experimenterand subject-effect parameters?

(ii) Can a sleeping subject discriminate between meaningfully related and extraneous cue words?

(iii) Will the relationship between hypnotizability and sleep responsivity disappear when subjects are matched for the amount of previous experience with the laboratory?



Fifteen male college students were invited to participate in a psychophysiological study of sleep, paying $10.50 per night. Their degree of hypnotic susceptibility had been assessed in advance by several different experimenters who used at least two standardized and two diagnostic hypnotic rating procedures. The standardized scales employed were the Harvard Group Scale of Hypnotic Susceptibility (HGSHS : A) of Shor and E. Orne (1962) and The Stanford Hypnotic Susceptibility Scale, Forms A, B, and C of Weitzenhoffer and Hilgard, 1959, 1962. The diagnostic rating procedure was that of Orne and O'Connell, 1967. Both the experimenter and technician were kept blind as to the subjects' previous hypnotic ratings.

Further, in order to evaluate arousal from stage REM during sleep, it was necessary to ensure that in the wake state subjects had sufficient alpha to be able to use its occurrence in sleep as an arousal criterion. Therefore each subject's alpha density was determined in advance. In an independent session with another experimenter a 2-minute period of EEG was recorded while the subject relaxed with his eyes closed. Only those subjects who had manifested a waking EEG alpha density of at least 40 % during this period were asked to participate in the present study.

Laboratory Setting

The subject slept in a quiet, comfortably furnished, air-conditioned room with minimal illumination. A special effort was made to provide a pleasant home bedroom atmosphere. Each subject was requested to limit his sleep on the preceding night to 5 h, and to refrain from alcohol, coffee, or stimulants during the preceding day. Upon arrival, he was questioned about his sleep patterns in a standard fashion. The experiment was then described to the subject as basic sleep research, which sought to explore the interrelationships between a number of electrophysiological variables. To minimize situation anxiety, the measures being employed were explained to the subject while the various recording electrodes were attached to his body. After the technician attached the electrodes, the experimenter conducted a word association test of 30 items (which included the cue words and "dummies" to be used subsequently) and left the room. Upon awakening the following morning, the subject was interviewed about the events of the night and on how well he had slept. The word association test was re-administered. During this period, skin potential responses (SPR) were recorded on the polygraphs and verbal responses on tape. Following this, the technician removed the electrodes. Subjects who did not respond on night 1 to any cue words, or who slept poorly were not recruited for subsequent nights. (In one case only, a subject who responded on night 1 was not available for further participation.)

Sleep Monitoring

Recordings were made with an Offner Type-R 8channel Dynograph, situated in the adjacent control

5 A Note on Terminology. This is one of a series of investigations which have used identical procedures for investigating cognitive processing during sleep. A change in terminology has been made, however, to conform with recent recommendations on labeling of sleep stages. (Rechtschaffen and Kales, 1968). Previously we called that part of stage REM sleep which was accompanied by bursts of rapid eye movements Stage 1 REM and that part which was not, Stage 1 NREM. In the present paper, stage REM refers to what was hitherto labeled Stage 1 and includes both REM and NREM segments.


38 C.W. Perry et al.: Responsivity in REM Sleep

room. The following measures were taken: monopolar occipital, mid-central and frontal EEG, sub-mental EMG (chin), bilateral eye movements, skin potential response (SPR-AC), skin potential level (SP-DC), and gross body movement. Two sets of electrodes were used for the EEG and electrodermal measures to ensure against electrode failures during the night (Van Kirk and Austin, 1964). The measurement of SPR followed the procedures described by O'Connell and Tursky (1960; 1962). The critical stage REM testing periods were recorded using a paper speed of 25 mm/sec; intervening stages of sleep were recorded on a schedule of 2 pages at 2.5 mm/sec alternated with 2 pages at 25 mm/sec.

Responses to stimulation during stage REM were recorded using a Panasonic television camera equipped with infra-red vidicon so that filming could take place in darkness. The camera and infra-red lights were concealed in the ceiling immediately above the subject's bed.

Administering Sleep Suggestions and Testing Cues

The polygraph technician signaled the presence of artifact-free stage REM sleep with a red light. Whenever this occurred the experimenter administered suggestions and cues over the microphone in a slow, quiet voice. The technician determined the presence of stage REM by visual on-line reading of the record using the criteria proposed by O'Connell, Gustafson, Evans, Orne and Shor (1963) in which the presence of three or more cycles of alpha are taken to indicate "arousal." Whenever such transient activity was observed in either the occipital or mid-central channels, the record was disqualified for 12 seconds. The presence of rapid eye movements in the bilateral eye movement channel was not considered in the on-line determination of alpha-free stage REM sleep.

All testings were conducted in stage REM sleep; stage 1 following awakening was not utilized. The subject was considered awake if alpha activity of a frequency, amplitude, and density similar to his waking pattern was evident for a minimum of 2 min. The experimenter stopped administering suggestions immediately whenever the technician signaled the presence of alpha activity. Two minutes of alpha-free stage REM sleep were allowed to pass before the experimenter attempted a further administration.

Upon successful administration of a suggestion, the experimenter proceeded to administer its appropriate cue word. For each cue word, a standard interval of 3 min was allowed between testings. When alpha was elicited during a cue presentation, 3 min of alpha-free activity had to elapse before it was re-administered.

It was not always possible to deliver the cue word in the same stage REM period in which the appropriate suggestion had been administered. At times, a stage REM period would terminate contiguously with the completion of suggestion administration. The cue word was always presented in the succeeding stage REM period. If the subject returned for a second or third experimental night, the cue word was retested without re-administration of the suggestion. For each of the two suggestions, the experimenter sought to deliver a minimum of five alpha-free testings of the appropriate cue word. The suggestions were counterbalanced for order of presentation. These procedures are identical to the ones used in the previous investigation (Evans et al., 1966, 1969, 1970).

Plan of the Experiment

Two suggestions which had been employed in previous studies were chosen to explore the dimensions of responsivity during sleep. They were:

Suggestion One: Whenever I say the word "itch," your nose will feel itchy until you scratch your nose. Cue Word: itch.

Suggestion Two: Whenever I say the word "pillow," your pillow will feel uncomfortable until you move your pillow with your hand. Cue Word: pillow.

The suggestions were repeated until two uninterrupted alpha-free administrations had been successfully completed.

Dummy Cues. -- To evaluate whether the subjects could discriminate between meaningful cue words and extraneous, irrelevant verbal stimuli, dummy cue words and pretests of cue words were instituted. For each suggestion there existed a dummy cue word in addition to the meaningfully related cue word. For "itch,'"' the dummy was "leg"; for "pillow," it was "blanket." The dummy cue words were selected to have the same number of syllables as the meaningfully related cue words. An equal number of dummy and real cue word testings were interspersed randomly.

Pretesting of Cues. -- As a test of the possibility that cue words in themselves elicit appropriate behavior, each of the four cue words in this study ("itch," "leg," "pillow," and "blanket") was tested once during the first stage of REM epoch of each night's sleep, prior to the administration of any suggestions.



Subjects and Rating Procedures

Subjects. -- Of the 15 subjects initially recruited, 5 were eliminated for failing to respond on night 1. The remaining 10 subjects slept for a total of 23 nights, including the initial selection night. Of these, 1 subject participated for 1 night, 5 subjects for 2 nights, and 4 subjects for 3 nights. The subjects slept, on the


39 C.W. Perry et al.: Responsivity in REM Sleep

average, from 7 hours 17 minutes (night 1) to 7 hours 26 minutes (night 3). Standard deviations ranged from 53-60 min.

Rating of Stage REM Sleep. -- Three independent judges rated the 65 suggestion administrations and 306 cue testings for the presence of alpha-free stage REM sleep in the period between stimulation and response. A cue was eliminated from further consideration whenever two of the three judges rated it as having three or more cycles of alpha either in the 5-second period before or during its presentation. Of the 306 (real and dummy) cues presented, 37 (or 12.1 %) were eliminated by this method. Of the 22 real cue presentations eliminated, 5 elicited appropriate responses (22.7 %) while none of the 15 dummy cues elicited such responses.

Rating of Responses. -- Responses to cue stimulation were rated as either appropriate or inappropriate by the experimenter at the time of their occurrence on the basis of predetermined response criteria. A nose scratch to "itch" or its dummy ("leg") was taken as the criterion of appropriate response to cue testings of Suggestion One. All movements consonant with an uncomfortable pillow -- lifting the head, moving the head or the pillow to another position, putting the hand under the head, putting the hand on the pillow -- were taken as appropriate responses to the cue words "pillow" and "blanket."

The experimenter's ratings were validated by randomly recording several nights of cue testing on video tape. Two observers were required to make independent ratings of 71 responses -- appropriate and inappropriate -- from the videotapes, using the above response criteria. They were blind as to whether or not a suggestion had been administered and to the experimenter's ratings of responses, though they were aware of which cue had been delivered on any particular trial. All of the 47 videotaped responses rated by the experimenter as inappropriate were also rated in this manner by the two judges. Of the 24 responses rated by the experimenter as appropriate, the two judges disagreed on only 2 occasions. One of these was a disagreement about a head movement ("pillow"); the other was a disagreement as to whether the nose or the upper lip had been scratched in response to "leg."

Incidence of the Sleep Responses

Frequency of Response to Cues. -- Of the 269 alpha-free presentations of cue words, 18 were of pretests and will be discussed separately. The 251 test cue presentations elicited 117 behavioral responses of some sort. The subjects did not react behaviorally to the remaining 134 cue words.

The 117 cue presentations were subdivided into presentations which elicited appropriate and inappropriate responses respectively. The results are presented in Table 1.

An analysis of variance comparing the percentages of appropriate response to real and dummy cues revealed no significant differences, either on night 1 or on all nights combined. (F =0.36, night 1; F=0.82, all nights). Further, there were no significant differences in amount of appropriate response to "itch" and "pillow" (F=1.16, night 1; F=0.19, all nights). For all analyses, p > 0.05, df =1, 9, N =10.

Causal Relationships Between Suggestions, Cues and Responses

Pretesting of Cues. -- It is possible that the responses observed were to cue words themselves, rather than to instructions to respond to the cue word; the cue words "itch" and "pillow" may have been sufficient to elicit appropriate behavior. Six of the subjects in the sample provided data to assess this possibility. The cue words "itch," "pillow," "leg," and "blanket" were presented at the beginning of their first night of participation, prior to the administration of any suggestions. On all of the 18 alpha-free pretests (4 presentations of "itch" and "leg," 5 of "pillow" and "blanket") the criterion behavior was not elicited. Subsequently all of these subjects responded to cue words once the appropriate suggestion had been administered. In addition, the cue words were tested 14 times at the beginning of night 2, and 11 times at the commencement of night 3. These 25 early testings elicited only one appropriate response. It is possible that there is less responsivity earlier in the night than later, so that response rates to pretests on all 3 nights may be confounded with a time of night effect.

Randomness of Sleep Responses.-- The lack of sig-


40 C.W. Perry et al.: Responsivity in REM Sleep

nificant differences in rate of response to real and dummy cues may be due to the response being a coincidental random body movement. This possibility was examined by comparing the percentages of correct responses to their corresponding cues (i.e., nose scratches to "itch," pillow responses to "pillow") as against percentages of such responses to non-correspondence cues (i.e., pillow responses to "itch," nose scratches to "pillow"). Equal amounts of appropriate response could be expected for both if responsivity is merely a coincidental random body movement. An analysis of variance comparing rates of appropriate response on these two occasions, over all nights, indicated significantly greater amounts of correct response to appropriate cue words than to inappropriate cue words (F=5.12; p<0.05). By contrast, when this comparison was made with dummy cues, comparing nose scratches to "leg" and pillow responses to "blanket" (i.e., appropriate cue-response pairings) with nose scratches to "blanket" and pillow responses to "leg" (i.e., inappropriate cue-response pairings), there were no significant effects (F=0.32; p>0.05). This finding was repeated when only responsivity to the first suggestion administered was examined (F=5.68; p<0.05 for real cues; F=1.06; p>0.05 for dummy cues). These results suggest that observed responses are not random body movements, but are specific responses to the suggested content associated with the cues.

Appropriateness of Responses to Cues. -- Foregoing sections indicate that responses appropriate for one cue word were, at times, emitted inappropriately to another cue. That is, the cue word "itch" sometimes elicited a "pillow" response, and vice versa. In all, this occurred on eight occasions. In seven of these instances (mostly nose scratches in response to "pillow"), the suggestion and cue word for "itch" had been administered and tested earlier in the night. Similar effects were found with dummy cue words. Occasionally, for instance, a nose scratch occurred in response to "blanket," the dummy cue word for "pillow." In four cases out of six the appropriate cue word for "itch" had been administered earlier in the night. Thus in only a minority of cases, criterion responses occurred prior to the administration of their appropriate suggestions.

Latency of Sleep Responses. -- Latencies for night 1 and for all nights combined were divided separately into "fast" and "slow" by means of a median split. For both the 74 latencies on night 1 and the 117 latencies of all nights combined, the median was found to be 110 sec. For each subject the percentage of responses falling above 110 sec (that is, "slow" latencies) within each of the three response categories was calculated. The response classes were (a) appropriate responses to real cues, (b) inappropriate responses to real cues, and (c) all responses (appropriate and inappropriate) to dummy cues. This latter category was made necessary by the small number of appropriate

responses to dummy cues, which did not permit independent analysis.

On night 1 there was no significant difference in percentage of slow latencies among the three response categories (F=2.01; p>0.05). By contrast, over all nights combined, there was a significant effect (F=4.62; p<0.05, df=2, 18). Slow latencies were associated more frequently with appropriate responses to real cues than with the other two response categories. Figure 1 indicates that modal appropriate response latency to real cues occurred between 150-180 sec. By contrast, latencies of all other response categories were not related to any particular time interval. The mean latency of appropriate responses to real cues was slower than for appropriate responses to dummy cues (140.5 sec vs. 112.1 sec).

Walking Recall of Events During Sleep.-- Immediately upon morning awakening, subjects were questioned about their nocturnal experience. Although the interview was open-ended, certain questions were always. asked: how well had the subject slept, had he had any awakenings, had it been quiet in the bedroom, had he heard any sounds or noises, had anything been going through his mind prior to the experimenter's awakening him, and had he been dreaming.

Four subjects provided seven reports of experiment-related material during the 23 nights of participation. Three of these subjects said they thought they had heard the experimenter say various words over the microphone. In most cases, they could


41 C.W. Perry et al.: Responsivity in REM Sleep

not recall what words they thought the experimenter said. One subject thought he had heard his name spoken. Another suggested that it might have been a dream. In all, there was slight and fragmentary waking recall of the sleep stimulation. The word association test revealed no significant differences, either before or after sleep, between cue words, dummy cue words, or neutral words; this held for both verbal latency and SPR latency and amplitude. Further, none of the subjects emitted any of the appropriate behavioral responses to the cue words "itch" and "pillow" during these testings of word associations. Upon completion of the experiment, subjects underwent a postexperimental inquiry by an independent experimenter in order to ascertain their perception of the experimenter's purpose. In every instance, despite extensive prodding, the subjects appeared to be oblivious to what had occurred during the night.

Hypnotizability and Sleep Response

Hypnotic Rating and Sleep Response. -- Hypnotic susceptibility was assessed in advance of this study by independent experimenters. These ratings were unknown to the experimenter and technician. All subjects were used to examine this relationship between hypnotizability and sleep response -- the ten responders plus five subjects who did not respond on night 1. No relationship was found between diagnostic hypnotic rating and percentage of correct responses to real cues (H=0.41; p>0.05; N=6, 5, 4). Further, no significant effects were found when hypnotic susceptibility was related to the frequency of appropriate response, the number of real cue presentations, and the number of real plus dummy cue presentations. Response frequency was also unrelated to the experimenter's blind guesses of each subject's level of hypnotic susceptibility.


The present study attempted to demonstrate that the behavior elicited is a function of instructions to respond to cue words rather than to cue words themselves. Accordingly cue words were pretested during stage REM sleep prior to the administration of suggestions. Appropriate responses to these pretests were never elicited on night 1. Early testings on later nights of the study likewise elicited little responsivity -- one correct response in 25 testings. Thus the findings of the pretests may have been confounded inadvertently with a time of night effect, but are in accord with the other evidence of causal relationship between sleep responsivity A and cue word presentation.

Despite differences in procedure from previous investigations, the present findings support earlier evidence of a sleep response for which subjects have, at best, fragmentary recall on the following morning. Responsivity in sleep continued to occur despite the elimination of possible presleep set effects through the selection of subjects unknown to each other, and through the infra-red vidicon system of observation.

It was found, as previously, that responses were slower for appropriate than for inappropriate responses. However, in the earlier studies, the latencies were considerably faster than those recorded on the present occasion. This discrepancy appears to reflect one of the main differences between this and prior experiments. The introduction of "dummy" cues imposed a more difficult task upon the sleeping subject by requiring him to make discriminations between verbal stimuli instead of mere recognition of a single stimulus. The slower response latency may also reflect success in eliminating presleep set.

Responsivity in the present study was distinguishable as behavior elicited by suggestion related cues. Nearly all of the incidents of nose scratching occurred in response to the cue word "itch" whereas on the basis of a hypothesis of random responding, one would expect equal amounts of nose scratching to "pillow." The same finding held for the cue word "pillow." Further, there were no differences between rates of criterion responses (nose scratching and pillow responses) to the non-meaningfully related dummy cue words "leg" and "blanket."

On the small number of occasions when criterion behavior was elicited to an inappropriate cue word (e.g., a nose scratch to "pillow"), the appropriate suggestion and cue word (e.g.,"itch ") had been administered and tested earlier in the night, a finding that is consistent with evidence of poor discrimination during sleep. It was rare to find appropriate responses occurring prior to the administration of the relevant suggestion and cue.

The diminution of response rate in the present study (14.1 % among subjects selected on the basis of their responsivity on the first night in the present discrimination situation vs. 18.2% among unselected volunteers in the previous study which had been a recognition situation) appears to duplicate parameters influencing sleep responsivity that have been isolated in "wake-sleep" studies. In these studies, where subjects are required to respond while asleep to stimuli presented when awake, higher response rates are reported characteristically to recognition than to discrimination tasks. (Oswald et al., 1960; Williams et al., 1966; Zung et al. 1961). Previous "wake-sleep" and the present "sleep-sleep" studies concur in indicating that cognitive processing during sleep is relatively primitive, in that discriminatory powers are reduced, though not, by any means, ablated.

The present study did not confirm the previously obtained relationship between sleep responsivity and hypnotic susceptibility. Earlier studies have inter-


42 C.W. Perry et al.: Responsivity in REM Sleep

preted the finding of a relationship as possibly due to differential rapport with the experimenter of high and low susceptible subjects rather than to hypnotizability itself. In the present study, a major effort was made to eliminate differential rapport effects by using highly experienced subjects of all hypnotic susceptibility levels who had equal familiarity with the laboratory. The performance of this specially selected subject group is consistent with a rapport hypothesis, though further experimentation is needed to demonstrate the effect conclusively.


Aserinsky, E., and Kleitman, N.: Regularly occuring periods of eye motility and concomitant phenomena during sleep. Science (1953), 118: 273-274

Beh, H.C., and Barratt, P.E.H.: Discrimination and Conditioning during sleep as indicated by the electroencephalogram. Science (1965), 147: 1470-1471

Beigel, H.: Mental processes during the production of dreams. J. Psychol. (1959), 43: 171-187

Berger, R.J.: Experimental modification of dream content by meaningful verbal stimuli. Brit. J. Psychiat. (1963), 109: 722-740

Cobb, J.C., Evans, F.J., Gustafson, L.A., O'Connell, D.N., Orne, M.T., and Shor, R.E.: Specific motor response during sleep to sleep-administered meaningful suggestions: An exploratory investigation. Percept Mot. Skills (1965), 20: 629-636

Dement, W.C., and Wolpert, E.A.: The relation of eye movements, body motility and external stimuli to dream content. J. Exp. Psychol. (1958) 55: 543-553

Emmons, W.H., and Simon, C.W.: The non-recall of material presented during sleep. Amer. J. Psychol. (1956), 69: 76-81

Evans, F.J.: Hypnosis and sleep: Techniques for exploring cognitive activity during sleep. In: Fromm, E., Shor, R.E. (eds): Hypnosis: Research Developments and Perspectives. Chicago, Aldine Press, 1972, pp. 43-83

Evans, F.J., Gustafson, L.A., O'Connell, D.N., Orne, M.T., and Shor, R.E.: Response during sleep with intervening waking amnesia. Science (1966), 152: 666-667

Evans, F.J., Gustafson, L.A., O'Connell, D.N., Orne, M.T., and Shor, R.E.: Sleep-induced behavioral response: Relationship to susceptibility to hypnosis and laboratory sleep patterns. J. Nerv. Ment. Dis. (1969) 148: 467-476

Evans, F.J., Gustafson, L.A., O'Connell, D.N., Orne, M.T., and Shor, R.E.: Verbally induced behavioral responses during sleep. J. Nerv. Ment. Dis. (1970), 150: 171-187

Levy, C.M., Coolidge, F.L., and Staab, L.C.: Paired Associate learning during EEG-defined sleep: A preliminary study. Aust. J. Psychol. (1972), 24: 219-226

O'Connell, D.N., and Tursky, B.: Silver-silver chloride sponge electrodes for skin potential recording. Amer. J. Psychol. (1960) 73: 302-304

O'Connell, D.N., Tursky, B.: Special modifications of the silversilver chloride sponge electrode for skin recording. Psychophysiol. Newsltr. (1962), 8: 31-37

O'Connell, D.N., Gustafson, L.A., Evans, F.J., Orne, M.T., and Shor, R.E.: Can waking and stage 1 sleep always be told apart by EEG criteria alone? Read before the Assoc Psychophysiol Study Sleep, Washington, DC March, 1963

Orne, M.T., and O'Connell, D.N.: Diagnostic ratings of hypnotizability. Int. J. Clin. Exp. Hyp. (1967), 15: 125-133

Oswald, I., Berger, R.J., Jaramillo, R.A., Keddie, K.M.G., Olley, P.C., Plunkett, G.B.: Melancholia and barbiturates: A controlled EEG, body and eye movement study of sleep. Brit. J. Psychiat. (1963), 109: 66-78

Oswald, I., Taylor, A.M., and Treisman, M.: Discriminative responses to stimulation during human sleep. Brain (1960), 83: 440-453

Rechtschaffen, A., and Kales, A. (eds.) A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. Washington, DC, Public Health Service, U.S. Government Printing Office, 1968

Shor, R.E., and Orne, E.C.: The Harvard Group Scale of Hypnotic Susceptibility, Form A. Palo Alto, Calif, Consulting Psychologists Press, 1962

Simon, C.W., and Emmons, W.H.: Learning during sleep? Psychol Bull (1955), 52: 328-342

Simon, C.W., Emmons, W.H.: EEG, consciousness, and sleep. Science (1956), 124: 1066-1069

Simon, C.W., Emmons, W.H.: Responses to material presented during previous levels of sleep. J. Exp. Psychol. (1956), 51: 89-97

Van Kirk, K., and Austin, M.T.: The electroencephalogram during all night recording: Technique. Amer. J. EEG Technol. (1964), 4: 53-61

Weitzenhoffer, A.M., and 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

Williams, H.L., Morlock, H.C., and Morlock, J.V.: Instrumental behavior during sleep. Psychophysiology (1966), 2: 208-216

Zung, W.W.K., and Wilson, W.P.: Responses to auditory stimulation during sleep. Arch. Gen. Psychiatry (1961), 4: 548-552

The preceding paper is a reproduction of the following article (Perry, C. W., Evans, F. J., O'Connell, D. N., Orne, E. C., & Orne, M. T. Behavioral response to verbal stimuli administered and tested during REM sleep: A further investigation. Waking and Sleeping, 1978, 2, 35-42.). The journal Waking and Sleeping was published between 1976 and 1980 by Wurzburg Editio Asklepion.