INDIVIDUAL DIFFERENCES IN MENTAL ACTIVITY AT SLEEP ONSET 1

Transcription

1 Journal of Abnormal Psychology 1966, Vol. 71, No. 4, INDIVIDUAL DIFFERENCES IN MENTAL ACTIVITY AT SLEEP ONSET 1 DAVID FOULKES, PAUL S. SPEAR, AND JOHN D. SYMONDS University of Wyoming Reports of sleep-onset and nocturnal (EEG Stage 1, rapid eye-movement [REM] period) mentation, CPIs, and thematic fantasy responses were collected from 32 young-adult Ss. In confirmation of recent findings by Foulkes and Vogel, dreamlike mental activity was found to occur with fairly substantial frequency at sleep onset, albeit with wide individual differences. Patterning of personality correlates with hypnagogic and nocturnal dream recall suggests that the former varies directly with waking ego-strength and adaptive flexibility, while the latter varies directly with indications of waking maladaptive symptomatology. A psychodynamic interpretation of these results is offered. Considerable evidence (e.g., Dement, 1965) has accumulated in recent years relating nocturnal dreaming to a particular phase of EEG sleep, ascending EEG Stage 1 accompanied by intermittent rapid eye movements (REMs). Periods of this stage of sleep recur in a cyclic fashion throughout the night. They are not normally observed at the onset of sleep, however, the first such period generally occurring only after an hour or more of non- REM sleep. Foulkes and Vogel (196S) have recently demonstrated the occurrence of dreamlike mentation at sleep onset in the absence of an ascending Stage 1 EEG or REMs. Although the content of such mentation may ultimately prove discriminable from content elicited during REM sleep, leaving undisturbed the association of REM sleep with a unique class of mental fantasy, these authors state that, by superficial appearance, it is not always easy to discriminate the sleep-onset dream from the REM-sleep dream. One of the most striking findings of the Foulkes and Vogel study was the extent of individual differences in the reporting of dreaming at sleep onset and in the association 1 Su-pport for the research herein reported has been received from the Department of Psychology, University of Wyoming, and from the National Science Foundation under Grant GS-860 to David Foulkes. The authors also wish to acknowledge the substantial and generous assistance of Allan Rechtschaffen in the establishment of a laboratory for sleep research at the University of Wyoming, and the helpful comments of Gerald Vogel upon an earlier draft of this report. 280 of dream reports with particular phases of EEG and EOG activity. This inter- and intrasubject variability in the reporting of at sleep onset contrasts with the practically universal high frequency of dream-reporting upon awakenings from nocturnal periods of REM sleep. These authors also noted their "clinical impressions" that the subjects who frequently reported dreamlike mentation at sleep onset were less anxious and constricted than the subjects characterized by better controlled and less hallucinatory mental content during this period. The present investigation was undertaken (a) to attempt to confirm, with a much larger subject sample than that of the Foulkes and Vogel study, the frequent occurrence of at sleep onset; (b) to determine whether personality variables associated with the hypnagogic occurrence of are in accord with Foulkes and Vogel's speculations on personality differences between sleep-onset dreamers and sleep-onset nondreamers; and (c) to examine personality variables differentially associated with hypnagogic and nocturnal (REM) dreaming, in an attempt to assess the possibly different psychodynamic functions each kind of dream might serve. METHOD Thirty-two subjects, 16 young-adult males and 16 young-adult females, were recruited from a predominantly freshman course in introductory psychology, receiving class-credit points for their service. Each subject spent 1 night in the laboratory, during which continuous EEG and EOG recordings were

2 MENTAL ACTIVITY AT SLEEP ONSET 281 -made according to the method of Dement and Kleltman (1957). EOG recordings were taken in monopolar fashion from electrodes placed at the outer canthus of each eye. Experimental awakenings were made both during the sleep-onset period and during REM sleep. A total of seven awakenings was scheduled for each night in the following order: Hypnagogic awakenings 1. Sleep-onset descending Stage 1 EEG 2 2. Sleep-onset alpha EEG with slow eye movements (SEMs) 3. Sleep-onset alpha EEG with SEMs 4. Sleep-onset descending Stage 1 EEG Nocturnal awakenings 5. REM sleep, 10 minutes after onset of second REM period of the night 6. REM sleep, 10 minutes after onset of third REM period of the night 7. REM sleep, 10 minutes after onset of fourth REM period of the night The sleep-onset awakenings were made in categories found by Foulkes and Vogel to produce an intermediate level of hypnagogic recall of dreamlike fantasy, hence, in categories especially favorable for a study of individual differences in such recall. Not all subjects completed the entire schedule because of their failure to complete four full cycles of sleep following the initial series of sleep-onset awakenings, but a total number of 78 REM-sleep awakenings were made with 31 subjects as well as the entire number of 128 scheduled sleep-onset awakenings. Awakenings were made by the experimenter calling the subject by name. When alert, the subject was asked what, if anything, had been going through his mind immediately before the experimenter's call. Upon completion of a spontaneous report, subjects were then asked if the content they experienced was visual ("Was there any visual imagery? What did you see?") and hallucinatory ("Were you, during the experience you've reported, aware that you were observing the contents of your own mind, or did you feel that you were participating in or observing events out in the real world?") and were also asked to classify their state of wakefulness or sleep immediately prior to the experimenter's call ("Were you awake and alert, awake but drowsy, drifting off to sleep, in light sleep, or in deep sleep?"). If the subject had no mental content to report, he was asked to "think on it" for a moment or so, and, if his answer was still negative, he classified his state of wakefulness or sleep immediately prior to the awakening and was then allowed to attempt to return to sleep. All interviews were tape-recorded for later transcription. 2 Descending Stage 1 EEG differs from ascending Stage 1 EEG in several respects: (a) it follows wakefulness rather than non-rem sleep; (b) it is unaccompanied by REMs; and (c) it has not hitherto been thought to be associated with the occurrence of dreaming. The transcriptions of the sleep-onset and REMsleep interviews were transferred to cards, one awakening protocol to each card. These cards, arranged in a random order, were given to two judges who knew neither the subjects nor the awakening categories associated with the reports contained therein. The judges were to rate the reports on 7-point scales for the appearance of Aggressive and Sexual content and for Hedonic Tone (1-very pleasant to 7-very unpleasant). Judges also coded the subjects' responses indicating their Depth of Sleep (0, awake and alert; 1, awake but drowsy; etc.) prior to awakening, and, employing the subjects' responses to the questions on the visual and hallucinatory aspects of their experiences, rated all reports on the following scale of Dreamlike Fantasy: 0. No content reported, feels mind was blank 1. No content reported, feels something was going through his mind, but forgets what 2. Conceptual content, everydayish 3. Conceptual content, bizarre or unusual topics 4. Perceptual content, nonhallucinatory, everydayish, undramatic 5. Perceptual content, nonhallucinatory, bizarre or unusual, dramatic 6. Perceptual content, hallucinatory, everydayish, undramatic 7. Perceptual content, hallucinatory, bizarre or unusual, dramatic Reliabilities of the judges' ratings were as follows (averaged subject-ratings for all reports): Hypnagogic reports REM reports Aggression Sex Hedonic Tone Depth of Sleep Dreamlike Fantasy The relatively low reliability for Sex for REM reports reflects the almost total absence of this category of content in our sample of REM reports, while the relatively low REM reliability for Hedonic Tone must reflect the relatively greater ambiguity of hedonic tone in REM content than in hypnagogic content or in wakeful fantasy, since the same judges yielded considerably higher reliabilities in their ratings of hedonic tone for these latter kinds of content than they did for REM content. Subjects contributed three kinds of information in addition to their reports of hypnagogic and nocturnal mentation. Prior to their night of experimental service, they estimated their usual frequency of dream recall on Antrobus' (1962) scale. "Recaller" scores were computed by converting the subject's response (e.g., a check at "once every night") into a weekly frequency (e.g., 7.00). Subsequent to their night of experimental service, all subjects completed the California Psychological Inventory (CPI) and wrote responses to a "Picture-Story Test" which consisted of cards 3BM and 11 of the Thematic Apperception Test (TAT). CPIs were scored for the standard

3 282 D. FOULKES, P. S. SPEAR, AND J. D. SYMONDS TABLE l MEAN RATINGS OF HYPNAGOGIC AND NOCTURNAL MENTATION Alpha SEM Hypnagogic reports Stage 1 Total Alpha SEM versus Stage 1 P REM reports REM versus hypnagogic t Aggression (0-6) Sex (0-6) Hedonic tone (1-7) Depth of sleep (0-4) Dreamlike fantasy (0-7) Word count (mean number) OS ' scales, while TAT responses were rated by the two judges for Aggression, Sex, Hedonic Tone, and Imaginativeness. Reliabilities of these ratings (summed subject-ratings for the two stories) were: Aggression,.92; Sex,.70; Hedonic Tone,.85; and Imagination,.91. RESULTS Incidence and Quality of Mental Content Reported at Sleep Onset Table 1 presents the means of subject means for each of the rating variables applied to alpha-sem and Stage 1 hypnagogic reports and to REM-sleep reports and for substantive word counts made of each class of report. All intrahypnagogic and hypnagogic- REM mean differences were tested with the Wilcoxen matched-pairs signed-ranks test (two-tail). From the data in Table 1 it is apparent that there is very little manifest instinctual (aggressive, sexual) content in either hypnagogic or REM reports, that mean hedonic tone does not vary significantly between reports of hypnagogic mentation and reports of nocturnal mentation, and that neither quantity (Word Count) nor quality (Dreamlike Fantasy) dimensions successfully discriminate hypnagogic from REM mentation. The only rating dimension which does discriminate descending Stage 1 reports from alpha-sem reports, and REM reports from hypnagogic reports, is subjectively estimated Depth of Sleep before "awakening," and this is essentially a validation of the experimenter's choice of points at which the subject was awakened. A more refined analysis, however, does reveal some differences between hypnagogic and REM-sleep reports. Considering reports which meet a relatively strict criterion of dreamlikeness (rated either 6 or 7 by both raters on the scale of Dreamlike Fantasy), a difference emerges between the two classes of awakening. Of 104 such "dreams," 54 were from REM sleep (comprising 69.2% of all REM awakenings) and SO (29 from descending Stage 1 and 21 from alpha-sem) from hypnagogic awakenings (comprising 39.1% of all hypnagogic awakenings). This difference was significant (by the Wilcoxen test) at the.01 level. When consideration is further restricted to reports rated 7 by both raters on the scale of Dreamlike Fantasy (i.e., reports unequivocally perceptual-hallucinatory and also either bizarre or highly dramatic), 21.8% of REM reports are so classified, compared to of descending Stage 1 reports and of alpha-sem reports. This difference was significant (again by the Wilcoxen test) at the.05 level. "Dreams," in terms of either of two relatively strict sets of criteria, thus occur significantly more often during REM sleep than during the sleep-onset period. They are by no means absent, however, at sleep onset, occurring more than half as often as during REM sleep. In this respect, the data of the present study confirm Foulkes and Vogel's observations of the occurrence of dreaming at sleep onset. Further analysis also reveals why mean scores on Dreamlike Fantasy do not discriminate REM reports from hypnagogic reports while the frequency of 6 or 7 ratings does so discriminate. REM ratings on this scale tend to constitute a bimodal distribution, with scores clustering either at the high end of the scale, that is, strongly dreamlike recall (69.2% of the judgments of both raters are 6 or 7), or at the low end of the scale, that is, no recall at all (17.9% of the reports are judged by both raters as 0 or 1). While hypnagogic reports are not so often rated

4 MENTAL ACTIVITY AT SLEEP ONSET or 7 as are REM reports, neither, on the other hand, are they so often rated 0 or 1. Only 3.1% of alpha-sem reports failed to produce any recall, while the comparable figure for descending Stage 1 was 7.8%. The difference between hypnagogic and REM awakenings in producing no-content reports was significant, by the Wilcoxen test, at the.01 level. Thus, REM awakenings produce both significantly more "dream" reports and significantly more no-content reports than do hypnagogic awakenings. Either recall or the content to be recalled appears to be more of an all-ornothing phenomenon during REM sleep, while during the sleep-onset period either recall or content is more of a graded process. The essentially rectangular distribution of sleep-onset ratings on the Dreamlike Fantasy Scale confirms Foulkes and Vogel's observation of substantial individual differences in the occurrence of at sleep onset. Considering reports rated 6 or 7 by both raters on this scale, for example, 8 subjects had a 0% frequency of such reports, 8 subjects had 25% frequency, 8 subjects had 50% frequency, 6 subjects had 75% frequency, and 2 subjects had 100% frequency. Such a distribution, of course, is favorable to the investigation of personality factors associated with individual differences in the occurrence of at sleep onset. Personality Variables Associated with Sleep- Onset Dreaming Significant positive Pearson productmoment correlations were observed between the CPI Social Presence and Self-Acceptance Scales and hypnagogic Dreamlike Fantasy ratings, averaged over the two raters and summed over all four hypnagogic reports (.35 and.37, respectively, p.05). A significant negative correlation was observed between such ratings of hypnagogic Dreamlike Fantasy and the CPI Socialization Scale (r =.57, p=.001). Hypnagogic dreamers thus seem to have greater social poise, to be more self-accepting, and to be less rigidly conforming to social standards than do hypnagogic nondreamers. Although ratings of sleep-onset mentation on the Dreamlike Fantasy scale correlated significantly with REM Word Counts (r =.47, p =.01), they did not correlate significantly with ratings of nocturnal mentation on the Dreamlike Fantasy scale (r =.20) nor with other qualitative features (aggression, sex, hedonic tone) of reported REM mentation. It does not appear, then, that hypnagogic and nocturnal Dreamlike Fantasy are highly intercorrelated with one another. The picture is quite different, however, when hypnagogic and waking fantasy are compared with one another. Hypnagogic Dreamlike Fantasy correlated positively and significantly with all five scores available for Thematic Apperception Test content: with TAT Word Counts, r =.42 (p =.05); with TAT Aggression, r =.51 (p=.01); with TAT Sex, r-.39 (p-.05); with TAT Hedonic Tone, r.44 (p.05); and with TAT Imagination, r -.40 (p -.05). Hypnagogic mentation and controlled waking fantasy production are, evidently, strongly intercorrelated with one another. By inference, one might conclude that hypnagogic fantasizing is under the same kinds of ego-controls as predominate in the thematic fantasy test, but that neither share as much in common with true nocturnal fantasy as they do with one another. And, in fact, TAT imagination correlates only.22, a value not significantly different from zero, with REM-sleep ratings of Dreamlike Fantasy. Thematic Apperception Test Aggression and Hedonic Tone did correlate significantly, however, with REMsleep Dreamlike Fantasy (r.45 and r =.42, respectively; p =.01 and p =.05, respectively). Also supporting the notion that waking ego-functions are more strongly related to hypnagogic fantasizing than to nocturnal fantasizing is the evidence with respect to the CPI Psychological Mindedness Scale. This scale correlated positively (r =.17), but insignificantly, with hypnagogic Dreamlike Fantasy, while it was the only one of 18 CPI scales to correlate significantly with nocturnal Dreamlike Fantasy, the direction of the relation being negative (r =.40, p =.05) in this case. The CPI Socialization Scale which correlated so significantly with hypnagogic Dreamlike Fantasy correlated.01 with nocturnal Dreamlike Fantasy. The CPI Social Presence and Self-Acceptance Scales, which also related significantly with hypnagogic

5 284 D. FOULKES, P. S. SPEAR, AND J. D. SYMONDS Dreamlike Fantasy, correlated.16 and.10, respectively, with nocturnal Dreamlike Fantasy, neither of these values being significantly different from zero. Recaller scores, preexperimental impressions of frequency of dream occurrence, correlated significantly (r =.35, p =.05) with ratings of nocturnal Dreamlike Fantasy, but.00 with ratings of hypnagogic Dreamlike Fantasy. Evidently, of the two varieties of dream, it is the nocturnal REM-period dream which provides whatever accurate basis there is for these everyday impressions of dream occurrence. The low magnitude of the significant correlation indicates, however, that factors other than variations in the quality of nocturnal fantasy determine everyday impressions of dream occurrence. The significant correlation of Recaller scores with certain CPI scales (Socialization,.37 [p =.05]; Self-Control, -.45 [p =.01]; and Good Impression [a fake-good scale comparable to L or K on the MMPI], -.36 [p =.05]) which were not significantly associated with nocturnal Dreamlike Fantasy suggests that, apart from any differences in the actual quality of reportable REM-sleep mental content, it is the highly socialized, self-controlled, and defensive person who is least likely to be aware of the nature and extent of such content. Psychological Mindedness, which correlated negatively with the Dreamlike Fantasy of experimentally elicited REM reports, also correlated negatively (r =.55, p =.01) with everyday impressions of dream recall. Differential Personality Variables Associated with Sleep-Onset and Nocturnal Dreaming From the CPI and TAT variables associated with hypnagogic dreaming and nocturnal dreaming, it appears that the former is more of an ego-function and more directly related to waking ego-strength and flexibility than is the latter. But perhaps the best method for illuminating personality characteristics differentially associated with hypnagogic and nocturnal (REM) dreaming is the examination and comparison of individual personality test items which best discriminate hypnagogic "dreamers" from hypnagogic "nondreamers" and nocturnal "dreamers" from nocturnal "nondreamers." To this end, an analysis was conducted to determine the CPI items most successful in discriminating responses of subjects rated high in hypnagogic Dreamlike Fantasy from responses of subjects rated low in hypnagogic Dreamlike Fantasy and in discriminating responses of subjects rated high in nocturnal Dreamlike Fantasy from responses of subjects rated low in nocturnal Dreamlike Fantasy. The groups selected for this analysis comprised the top and bottom quarter of the distributions of ratings of Dreamlike Fantasy in each period of sleep. There was little overlap between the REMsleep and the hypnagogic comparison groups: of the 8 subjects high in REM fantasy, 2 were in the corresponding group for hypnagogic fantasy, 3 in the low group for hypnagogic fantasy, and 3 in neither of the hypnagogic comparison groups; of the 8 subjects low in REM fantasy, 2 were in the corresponding group for hypnagogic fantasy, 2 in the high group for hypnagogic fantasy, and 4 in neither of the hypnagogic comparison groups. The sex-composition of comparison groups was well-balanced for each sleep period (in the high hypnagogic fantasy group there were 5 males and 3 females, in the low hypnagogic fantasy group there were 4 males and 4 females, and both the high and low REM fantasy groups were comprised of 5 females and 3 males). Table 2 presents the results of the CPI item analyses. Looking first at the items most successful in discriminating high and low hypnagogic "dreamers," it is apparent that several characteristics of the "authoritarian personality" syndrome pervade the entire list (rigidity, conventionality, intolerance, antiintraception) with the hypnagogic nondreamers' responses closely approximating the standard conception of the authoritarian syndrome (Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950). An inspection of the items most successful in discriminating high and low nocturnal "dreamers" reveals considerably less immediate homogeneity of content than was the case for the items discriminating hypnagogic fantasy groups. It is readily apparent, however, that somewhat different dimensions of personality are operative in the two cases. Whilf the item response "false" characterized the high hypnagogic fantasy group on 7 of the 8

6 MENTAL ACTIVITY AT SLEEP ONSET 285 TABLE 2 MOST DISCRIMINATING CPI ITEMS group high in at sleep onset 87.5% 37.5% 37.5% group high in during REM sleep 75.0% 87.5% 75.0% Item 112. I set a high standard for myself and I feel others should do the same 148. I believe women should have as much sexual freedom as men 367. My home life was always very pleasant 67. I feel sure that there is only one true religion 69. I would disapprove of anyone's drinking to the point of intoxication at a party 131. One of my aims in life is to accomplish something that would make my mother proud of me 371. I would rather be a steady and dependable worker than a brilliant but unstable one 404. I am in favor of a very strict enforcement of all laws, no matter what the consequences Item 105. I am fascinated by fire 128. It takes a lot of argument to convince most people of the truth 151. I have strange and peculiar thoughts 155. A person should adapt his ideas and his behavior to the group that happens to be with him at the time 300. Police cars should be especially marked so that you can always see them coming group low in at sleep onset % 75.0% group low in during REM sleep 25.0% most discriminating items for fantasy during that period of sleep, the item response "true" characterized the high fantasy group for all 5 of the most discriminating items for fantasy reported from REM sleep. Perhaps the simplest common denominator of the S items to which nocturnal dreamers tended to reply "true" is that all are keyed false on several scales of social adjustment or personal growth (105 and 300 on the CPI Responsibility Scale, 128 on the CPI Capacity for Status Scale, 151 on the CPI Self-Control and Tolerance Scales, and 155 on the Psychological Mindedness Scale). These responses of the high fantasizer during REM sleep thus seem indicative of something less than optimal social or personal adjustment. Indeed, in one case ("I have strange and peculiar thoughts"), the manifest content of the discriminating item which is endorsed by the REM-period fantasizer suggests considerable disturbance. Such results and interpretations are consistent with Foulkes and Rechtschaffen's (1964) data demonstrating a pattern of positive correlation between scores on MMPI pathology scales and intensity of recalled REM-sleep mentation. Shifting from the more obvious characteristics of the discriminating items for REM-sleep fantasy to their possible latent meaning, it appears that while the low hypnagogic fantasizer exerts rigidly successful control over his impulse life, the high nocturnal fantasizer shows fascination with impulse life (as symbolized, for example, by fire) in conjunction with weakened egocontrol mechanisms (as indicated by the presence of strange and peculiar thoughts and by the concern with evading detection evident in the police-car response and in the strategy of adopting the protective coloring of conformity to the environment in which one finds himself). DISCUSSION Dreaming, sometimes difficult to distinguish from REM-sleep dreaming, does occur with fairly substantial frequency during sleep onset. Although, in the present study, "dreams" (i.e., reports which judges agree are of perceptual, hallucinatory, and dramatic content) occurred significantly more often during REM sleep than at sleep onset, they

7 286 D. FOULKES, P. S. SPEAR, AND J. D. SYMONDS did occur at sleep onset, and more than half as often as they did during REM sleep. As Foulkes and Vogel (1965) have noted, such findings imply that research on the physiological basis of, or on mental activity occurring during, nocturnal periods of REM sleep cannot provide completely satisfactory answers to questions on the physiology and psychology of dreaming. But there clearly are certain respects in which sleep-onset dream activity differs from nocturnal dream activity. One of these is the susceptibility of the association of dream reporting with EEG/EOG activity to variation from subject to subject. The results of the present study indicated, in confirmation of Foulkes and Vogel's earlier but less systematic observations, that healthy and expansive ego functioning during wakefulness is positively associated with the occurrence of during sleep onset: the hypnagogic dreamers tended to have greater social poise, to be more self-accepting, to be less rigidly conforming to social standards, and to be more adept at producing voluntary waking fantasy than did the hypnagogic nondreamers. These findings may be contrasted with the minimal indications in the present study, and the more substantial evidence collected by Foulkes and Rechtschaffen (1964), that signs of maladaptive or pathological waking response are positively associated with dreamlike qualities of REM-sleep fantasy. Elsewhere, Vogel et al. (1966) have presented a psychodynamic view of sleep onset. They note that successful sleep onset implies increasing withdrawal of ego cathexes from the external world and propose that this withdrawal has certain unintended consequences, viz., narcissistic, regressive mental content. It is also suggested that individual differences in ego-strength are directly associated with tolerance for such content and, hence, with variations in the dreamlike quality of mentation at sleep onset. The findings of our study are entirely in accord with this view of hypnagogic dreams as ego-processes, initiated and controlled by the self, and as correlates of the adaptiveness of waking egofunctioning. Vogel et al. also propose that the REM dream is the ego's defensive response to pressures (increased instinctual tensions, etc.) arising outside the ego, rather than a process initiated or directly controlled by the ego itself. The less adequate the ego, the less wellcontrolled will be the ensuing dream. On this view, it is to be expected that ego-strength will be inversely related with the dreamlike properties of mental experience during REM sleep. The evidence of the Foulkes and Rechtschaffen (1964) study, and to a lesser degree, of the present study, is also supportive of this proposition. It appears, then, that the hypnagogic dream might profitably be viewed as an egocontrolled excursion into inner thoughts and feelings following the ego's voluntary decathexis of sensory input from the external world. Subjects with rigid defenses against impulse-life tend to resist any encroachment of such regressive mental content upon wakeful levels of ego-functioning, and thus experience little dreamlike content at sleep onset. The nocturnal dream, on the other hand, may be viewed as the ego's involuntary response to unconscious impulses and anxieties which become active during REM sleep. Subjects with egos lacking adequate defenses against impulse-life tend to be overwhelmed by it during REM sleep, hence experience especially vivid REM-sleep dreams. REFERENCES ADORNO, T. W., FRENKEL-BRUNSWIK, E., LEVINSON, D. J., & SANFORD, R. N. The authoritarian personality. New York: Harper, ANTROBUS, J. S. Patterns of dreaming and dream recall. Unpublished doctoral dissertation, Columbia University, DEMENT, W. C. An essay on dreams. In T. M. Newcomb (Ed.), New directions in psychology II. New York: Holt, Rinehart & Winston, Pp DEMENT, W., & KLEITMAN, N. Cyclic variations in EEC during sleep and their relation to eye movements, body motility, and dreaming. Electroencephalography and Clinical Neurophysiology, 19S7, 9, FOULKES, D., & RECHTSCHAFFEN, A. Presleep determinants of dream content: Effects of two films. Perceptual & Motor Skills, 1964, 19, FOULKES, D., & VOGEL, G. Mental activity at sleep onset. Journal of Abnormal Psychology, 1965, 70, VOGEL, G., FOULKES, D., & TROSMAN, H. Ego functions and dreaming during sleep onset. Archives of General Psychiatry, 1966, 14, (Received June 30, 1965)