CHAPTER VI. PHILOSOPHICAL ASPECTS OF DREAMS   Dreams and sleep-mentation have been discussed by several philosophers. Malcolm (1959) examined the previously stated opinions of philosophical writers on dreams and claimed they were mistaken. Essentially Malcolm challenges the idea that dreams represent mental activity in sleep and that they may be consciously experienced. The matter is raised here not only for its relevance to dream study generally but because a scientific study of lucid-dreams might resolve the philosophical issues. Descartes (see 1934) considered that a human mind is constantly conscious, even in sleep - consciousness being the 'essence' of mental substance. On dreams he says : 'all the same thoughts and conceptions which we have while awake may also come to us in sleep'. Malcolm quotes several other writers on this point : Kant : 'In deepest sleep perhaps the greatest perfection of the mind might be exercised in rational thought. For we have no reason for asserting the opposite except that we do not re­member the idea when awake. This reason however proves nothing.' Moore : 'We cease to perform (mental acts) only while we are asleep, without dreaming ; and even in sleep, so long as we dream, we are performing acts of consciousness.' Russell : 'What, in dreams, we see and hear, we do in fact see and hear, though, owing to the unusual context, what we see and hear gives rise to false beliefs. Similarly, what we remember in dreams we do really remember; that is to say, the experience called 'remembering' does occur.' Freud : 'Obviously, the dream is the life of the mind during sleep.' Malcolm approaches the problem systematically. He points out first that it is absurd and self-contradictory for a person to assert or to judge that he or she is asleep, unconscious or dead. A person might say that he or she is asleep when actually asleep, but the remark would not be taken seriously since awareness would be absent. Malcolm says : 'In order to know that when a man said 'I am asleep' he gave a true description of his own state, one would have to know that he said it while asleep and that he was aware of saying it. This is an impossible thing to know because whatever showed that he was aware of saying that sentence would also show that he was not asleep. The know­ledge required is impossible because it is self-contradictory.' He further states that 'having some conscious experience or other, no matter what, is not what is meant by being asleep.' Malcolm also considers states resembling sleep and concludes that hypnosis, for instance, is not sleep. Also, neither is a nightmare where a person is threshing about and talking. In the case of differential discrimination of external sounds (e.g. a baby’s cry may wake a person whereas continual road traffic noise does not) the lack of perception of some stimuli satisfies the criteria for sleep. Analysing further the question of judgements in sleep he considers possible ways of determining whether a person made a judgement in sleep : 1. The person might state an awareness of being asleep at the time of the judgement - but this is self-contradictory Malcolm asserts. 2. The person infers being asleep because : a. The judgement was cotemporaneous with, say, a burst of thunder -but in that case the sleeper was not fully asleep says Malcolm. b. The person knows the judgement was not made before or after sleep so it must be during sleep - but the conclusion does not follow and the truth is theoretically unverifiable. c. At the time the person was having a certain experience that only occurs in sleep - but it cannot be verified that this experience ever does occur in sleep. d. The person made the judgement whilst dreaming - but the assertion cannot be made that the judgement was made at the same physical time of dreaming - there are no grounds. Additionally, there are questions as to how the person knows dreaming occurred during sleep and how the person knows a dream was experienced. 3. No physiological phenomena can be used as evidence that a judgement was made in sleep. For instance, if ,on making a judgement a particular brain-wave occurs its presence during sleep is not proof that a judgement was made since the correlation was established in persons who were awake : it might be true or false. The arguments propounded by Malcolm also apply to other mental phenomena, he states, such as thinking, reasoning, perceiving, imagining and questioning , as well as 'passivities' such as fear, anxiety, joy and imagery. Dreaming though is an exception he says. The schema of proof against judgements in sleep does not apply. Claims of making judgements or having imagery in dreams he says are meaningless because they cannot be verified whereas what establishes that a person dreamt is the telling of a dream. Malcolm at one time held that since it is theoretically impossible to verify that someone had images, say, in his sleep, but possible to verify that he dreamt, then a dream cannot be identical with, nor composed of, images experienced during sleep. He thought it proof that dreaming is not a mental activity or a mental phenomenon or conscious experience. However, he now considers these terms vague, so instead, he opines that dreams are not composed of thoughts, feelings and so on. Malcolm states that he does not maintain that a dream is the waking impression that one dreamt. He does not know what dreaming is but we determine whether a person had a dream by receiving a dream report. Commenting on the link between REM sleep and the great number of dream reports obtained on waking from that Stage compared to others, he points out that eye-movements should not be used as a criterion for having dreamt. The assumption for instance that a person who showed a long REM period but could only report a short dream, had forgotten part of the dream does not follow. Malcolm believes that the error of psychologists and others is to assume that a dream must have a definite location and duration in physical time - an example of what Wittgenstein termed a 'prejudice' caused by 'grammatical illusions'. The dreamer's assertion that he dreamt 'just before waking' is unverifiable. Similarly, the connection between events in the dream report and external stimuli does not mean the two events were simultaneous. The link is between waking reports of dreams and physical occurrences. The 'length' of a dream, too, has no clear sense. What Malcolm says about sleep applies for most of the time and his comments on the dreaming / REM correlation are clearly sensible. However, his approach generally seems to be too simplistic. How could he explain for instance the case of a person in a state of 's1eep paralysis' where the person struggles (without actually moving) whilst being perfectly conscious of the situation? Physiologically the person is still in REM sleep, yet consciousness is present (Rechtschaffen et al, 1963). In particular, the matter of lucid-dreams could demolish his stand if the dreamer could convey information from the lucid-dream to the external world and respond intelligently to questions from the real world. This would show the presence of thought in sleep and if the dream report on waking faithfully reflected information signalled from REM sleep, it would surely be the simplest and most intelligent step to regard the experience during which the signalling occurred as a dream if that is what it was later described as. Clearly, the study of lucid-dreams could exert a profound influence on great philosophical questions. It has been the purpose of this Introduction to establish the particular area to be studied - the lucid-dream - and to illustrate how the phenomenon relates to the general backcloth of dreams and sleep-research. A thorough search of the literature indicated that no previous electro-physiological studies specifically on lucid-dreams had been published, therefore much useful information might be potentially discovered if an adequate methodology could be developed. The next section of this Thesis describes the whole course of experimentation. CHAPTER VII   THE NEW TECHNIQUE OVERVIEW The next 10 Chapters describe the experimentation performed for this study. At the beginning of the research there was no apparent previous experimental work in the area on which to base experiments. The main problem was that of marking the lucid-dream in the polygraphic record. Chapter VII describes the successful technique (involving ocular signalling) which was fortuitously hit upon. Having developed a method which effectively 'event-marked' the lucid-dream, and could also be used as a channel of communication from the lucid- dreamer, the next Chapter (VIII) describes a large-scale study which obtained and analysed electro-physiological data from 8 lucid-dream-night records (and 8 control nights) monitored from one subject. The next Chapter (XIX) presents an analysis of the psychological results of that first study. In order to consolidate the evidence for ocular signalling from Stage REM sleep, and being aware of the strength of demand-characteristics in experimental situations, the author next performed an experiment involving control subjects attempting to simulate these signals (Chapter XI). The relative scarcity of lucid-dreams in the sleep-laboratory led to an experiment to try to artificially induce lucid-dreams in subjects by an external-simulation method (Chapter XII). A further study was next performed on the original subject (Chapter XIII) to obtain information of a more diverse nature. In addition, analyses were conducted on extensive questionnaire and diary data supplied by that subject (Chapter XIV). Results of a questionnaire survey concerning lucid-dreams in a large number of persons are reported in Chapter XV, and Chapter XVI describes attempts to determine whether person­ality and intelligence factors relate to the experience of lucid-dreams. (Keith Hearne's PhD thesis, page 137) CHAPTER VII  THE NEW TECHNIQUE   VII.1 INTRODUCTION From the evidence discussed in III.8 this author did not believe that volitional motor movements (e.g. pressing a micro-switch) could occur in Stage REM sleep, in which it was assumed lucid-dreams would be found. Nevertheless, a micro-switch technique of signalling would be tried. However, in considering an alternative method it seemed a distinct possibility that eye-movements might provide the answer. The eyes are, of course, readily motile in REM sleep, whilst the rest of the body experiences a lack of muscular tonus (III.1). The question of whether REMs represent looking-at-the-picture ocular activity (Aserinsky & Kleitman, 1955 ; Dement & Kleitman, 1957a) or are simply concomitants of a change of neurophysiological. conditions (Oswald, 1962), has been a moot one in sleep-research. On the one hand, experiments such as that of Dement & Wolpert (1958a) in which last-recalled eye-movements before waking were compared with the EOG record, offer evidence for the 'scanning' hypothesis. On the other hand the fact that congenitally blind persons show REMs (Gross et al, 1965) does not support the idea. Perhaps both views are correct. Conceivably, most REMs are spontaneous and undirected (too gross to be scanning, as Oswald, 1962, commented), but that some are deliberate scanning movements. So, in the belief that directed ocular movements are possible in dreams, a sleep-lab study was begun to test the efficacy of ocular signalling from lucid-dreams, as well as the feasibility of motor responses. It was anticipated that if the subject made a sequence of 8 regular ocular movements on attaining lucidity, this would produce a distinctive EOG pattern (different from the usual random trace) which could be easily identified in the polygraphic record. This preliminary study was to continue until the S reported having had a lucid-dream in which he made ocular signals. The S would also be required to press a micro-switch, to test for any slight muscular response. If this were possible it would provide a simpler method than ocular signalling, as the micro-switch could be linked directly to the timer / marker pen on the recording apparatus. CHAPTER VII THE NEW TECHNIQUE   VII.2 METHOD A 37 year old male who reported that he experienced lucid-dreams approximately once or twice a week expressed interest in being an experimental subject. He stated that he had bad these dreams for some 20 years and reports an ability to control the content and course of action to some extent. He was unpaid.   APPARATUS This initial study was conducted in the sleep-laboratory using an Elema-Schönander Mingograf recording apparatus. Four channels recorded bipolar EEG, EOG (2 pens) and submental EMG. Control settings were : EEG : Gain : 70 microV/cm., time-constant : 0.3 secs., filter : 15 hz. EOG : Gain : 70 microV/cm., time-constant : 0.3 secs., filter : 15 hz. EMG : Gain : 20 microV/cm., time-constant : 0.015 secs., filter : 700 hz. The micro-switches (Burgess Products Co. Ltd) required a pressure of 200gm to operate. It was taped into the palm of the hand so that the fore-finger could press the button. FIGURE VII.1 ELECTRODE POSITIONS (c) Electrode emplacements 1cm diameter dome-shaped silver / silver-chloride electrodes were attached to the scalp by collodion glue and to facial skin by micro-pore surgical tape. Electrode-gel was introduced under the dome by syringe, to improve conductance. EEG was derived from electrodes positioned at approximately Pz and Fz points, according to the Inter­national 10-20 electrode system (Jasper 1958). EOG was recorded from electrodes placed above and below the outer canthus of each eye, measuring diagonally across the face, so picking-up potentials caused by shifts of the dipole moment of the electric charge between the retina and front of the eye. The EMG electrodes were situated on the jaw-bone, on either side of the chin, monit­oring muscular tonus. A ground electrode was placed on the subject's central forehead. FIGURE VII.2 LAYOUT OF THE SLEEP-LAB CHAPTER VII THE NEW TECHNIQUE   VII.3 RESULTS   The data suggest that ocular movements are a suitably effective method of signalling from the lucid-dream. On the first night of the study (night of 5th April 1975) the S woke at 8 am and reported having signalled from a lucid-dream. Unfortunately, the rec­ording equipment had been switched off shortly before. On the second night (night of 12th April 1975) the S again reported a lucid dream. He was able to make extreme horizontal eye-movements without waking*. The lucidity-onset signals are shown on page 163, whilst the whole lucid-dream is displayed on page 171. (a) Evidence for the Stage of sleep when signalling The Stage of sleep at signalling is unambiguous REM (Rechtschaffen & Kales, 1968), having : 1/ Low-voltage EEG with 'saw-toothed waves' characteristic of Stage REM sleep (Schwartz & Fischgold, 1960), and absence of alpha rhythm. (The S produces much 8-10 Hz activity when awake -see Appendix, page 369). 2/ Random, rapid-eye-movements, followed by the sequence of deliberate ocular signals. 3/ Low-voltage EMG, with heart-pulse from nearby blood-vessels showing-through. 4/ No movement artefacts present. (b) Absence of a muscular response in the hand The S reported dreaming of pressing the micro-switch, and even hearing it 'click', but in reality he did not. The switch was not malfunctioning, as it operated on test, and when the S accidentally pressed it in sleep during movement. (* The criteria of wakefulness and sleep correspond to those of Rechtschaffen & Kales, 1968) (c) Independent assessment of the results Results of this study (and several subsequent examples) were conveyed to Professor Allan Rechtschaffen at the University of Chicago, who expressed himself convinced that the signals were from Stage REM sleep. He stated that he too had used an eye-movement method for obtaining simple information from a patient in a state of narcolepsy (Private communication, 1975. See Appendix, pages 370-371). CHAPTER VII THE NEW TECHNIQUE   VII.4 CONCLUSIONS This test demonstrated that a technique of signalling from lucid-dreams using eye-signals was feasible whereas motor-movements were apparently not. The lucid-dream was seen to emerge in Stage REM sleep, as anticipated. The electro-physiological and psychological information regarding this first identified lucid-dream is stated in the next Chapter as part of the accumulated results obtained from 8 lucid-dreams. CHAPTER VIII THE 1st A.W. STUDY - ELECTRO-PHYSIOLOGICAL FINDINGS   VIII.1 INTRODUCTION   It was decided to follow up the successful demonstration of ocular signalling from lucid-dreams by a major study. The aim would be to obtain a record of 8 lucid-dreams having signals, so as to provide a satisfactory statistical comparison on several measures, with selected control nights (when the S had no lucid-dreams.) Simple manipulations of experimental conditions would be performed, and the S would be required to perform certain pre-determined tasks in the lucid-dream, on occasions (These aspects are covered in Chapter XX). Nothing was known about basic sleep parameters of lucid-dreams. Major questions selected for examination were : 1/ Do lucid-dreams always occur in REM sleep, or are some a form of hypnagogic/hypnopompic imagery happening outside of REM? 2/ What is the duration of lucid-dreams ? (This S had estimated a few minutes.) 3/ At what time of night do they occur? 4/ Is the sleep-pattern of lucid-dream nights any different from non-lucid-dream nights? 5/ If lucid-dreams do occur in Stage REM, how far into the particular REMP do they occur? 6/ Is the lucid-dream REMP disturbed in any way, indicating a lighter sleep? 7/ Are lucid-dreams different autonomically, from ordinary dreams, as reflected in heart-rate and REM amount (linked with dream activity, Dement & Wolpert, 1958b) ? It was hoped that, for the first time, answers to these questions would be objectively ascertained using the S’s ocular signals as chart-markers to indicate the onset of lucidity. It should be mentioned that this S usually roused himself after a lucid-dream. That fact alone might have identified the immediately preceding sleep-stage and time of lucid-dreams, but not reliably so, as the S could confuse dream periods if he had not actually woke after lucidity. In any case, signalling from within the dream could obviously provide much more potent information. CHAPTER VIII THE 1st A.W. STUDY - ELECTRO-PHYSIOLOGICAL FINDINGS   VIII.2 METHOD (a) Subject The same person was willing to return to the sleep-lab from time to time for further monitoring during sleep. It was decided to take advantage of this generous offer and obtain as much data as possible from the subject before attempting to generalise any findings.   (b) Apparatus The same Elema-Schönander Mingograf (see page 139) was employed at first in this study, but later a Grass polygraph (Type 7B) was used as the recording instrument. Polygraph settings were : EEG : Gain : 75 microV/cm., time-constant : 0.45 secs., Half amplitude high-frequency setting : 35 hz. EOG : Gain : 100 microV/cm., time-constant : 0.45 secs., (2 pens) Half amplitude high-frequency setting : 35 hz. EMG : Gain : 20-5O microV/cm., time-constant : 0.015 secs. Half amplitude high-frequency setting : 3Khz.   (c) Experimental design Details of manipulations and subject’s tasks are stated in the next Chapter, as only the basic para­meters of lucid-dreams are dealt with here.   (d) Instructions to subject Instructions to the subject were ess­entially the same as before (VII.2.d), with modifications for various subject tasks. The tasks are described in Chapter IX.   (e) Measures and statistics involved   18 measures were employed in the statistical analyses. These are shown in Table VIII.1, page 162. The following measures were quantified thus : REM amount : This was scored by having a naive assistant rate the amount of REM activity per page of sleep record according to 5 categories : No REMs, .25 page of REMs, .5 page of REMs, .75 page of REMs, continuous REMs. The result was divided by the number of pages to give a mean REM score. Heartrate : This was observed from the ENG trace. The pulse, from blood-vessels near the recording chin electrodes, became pronounced enough for visual counting in the low-tonus EMG during Stage REM sleep. Heart-rate was measured by a naive assistant for 1-minute periods Sleep-disturbances : Any observable increase in EMG was counted as a sleep disturbance. The naive assistant determined where this occurred. Sleep-onset : This was taken to be from the first k-complex in the sleep-record. The E observed this measure. % Stage : The percentage of each Stage of sleep was measured, up to and including the lucid-dream, but not after, in case waking and reporting created : excitement so affecting the later sleep pattern. Control data were from equivalent times in non-lucid dream nights. A number of comparisons (e.g. heart-rate before and during lucidity) were made using Sandler’s A-test (Sandler, 1955). This is a simple t-test for related samples having a normal distribution, and compares differences between pairs of scores. The sum of the square of each of these differences is divided by the square of the differences. The resultant value of A is referred to tables, using N-1 degrees of freedom. Values equal to or less than the table value are significant. Normality of the data was established by applying the Shapiro-Wilk (1965) test to the data. A computerised correlation program (N31C), (University of Liverpool Computer Laboratory), was used for cross-correlation of data. This provided a matrix of values of Pearson’s coefficient of linear correlation and tables of means and standard deviations for the equal-sized data matrix. CHAPTER VIII THE 1st A.W. STUDY - ELECTRO-PHYSIOLOGICAL FINDINGS   VIII.3 RESULTS This investigative 1st A.W. study eventually covered 45 nights, over a period of approximately 1 year, before 8 lucid-dreams were recorded. On 3 other occasions the S had a lucid-dream, but the results were not monitored for one reason or another. Table VIII.1 on page 162 displays the raw-data obtained. Six lucid-dreams occurred in complete all-night records, but there were gaps in the 2 other records. Consequently, comparisons of sleep Stages between lucid-dream and control nights were made using 6 samples each. Other comparisons used data from 8 dreams in each group. A. Basic findings for this one subject   a/ The Stage of sleep in which lucid-dreams occur All the lucid-dreams happened in unambiguous State REM sleep according to the scoring criteria of Rechtschaffen & Kales (1968). Ocular signals are shown on pages 163 to 170, and pictures of the whole sleep record for the duration of each lucid-dream are displayed on pages 171 to 177. (See also sleep-patterns : 178, 179 and 375, 376 ). b/ Duration of lucid-dreams The lucid-dreams continued for 51-354 seconds (from signals to rousing), giving a mean length of 153 seconds.   c/ Time of occurrence of lucid-dreams The dreams happened from 2.32 a.m. to 9.04 a.m., with 5 / 8 lucid-dreams occurring between 5 and 8 a.m.   d/ Sleep-patterns of lucid-dreams vs. Control nights Nine sleep-measures (% -up to and including lucid dream of : Stage 1, 2, 3, 4, REM, awake ; actual lights-out time, sleep-onset-time, and total-sleep-time) were compared, using Sandler’s A test, but no significant differences emerged (Appendix, pages 372-374). (p 152) A values were, respectively: 0.31, 0.56, 2.7, 3.4, 75.2, 0.86, 3.1, 27.3, 1.8 (5d.f.).   e/ Latency into the REMP when the lucid-dream occurs The lucid-dreams occurred from 2.12 to 51 minutes into the particular REMP, mean : 23.9 mins (decimal).   f/ Lucid-dream REMP disturbances (Figure VIII.17, page 180). In case lucid-dreams were associated with a lighter, more disturbed REMP than ordinary dreams, the number of sleep-disturbances in each lucid-dream REMP was compared to Control data. A sleep-disturbance was defined as any noticeable increase in EMG in the Polygraphic record. Such activity accomp­anies any physical movement in the sleeping subject (Rechtschaffen & Kales, 1958). No significant difference was found (A= 0.35 ; 7 df ; n.s.).   g/ Physiological correlates of arousal in lucid-dreams vs. Control dreams Mean heart-rate in the lucid-dream was not significantly different from that in Control REMPs (A= 0.56 ; 6 df ; n.s.). REM activity, however, was significantly greater in lucid-dreams than Control REMPs (A=0.15 ; 7 d.f. ; P(0.01).   h/ Physiological correlates of arousal, before and during lucidity The mean amount of REM activity during the lucid-dream was not significantly different from the mean amount before lucidity (A= O.32 ; 7 df ; n.s.). However, there was a significant increase in heart-rate during the lucid-dream compared to that before, taking a 1 minute measuring period, and allowing 20 seconds up to the start of signalling, during which time the S was achieving lucidity, (A= 0.25 ; 7 d.f. ; P < 0.05). A similar increase was not found in Control dream periods obtained from non-lucid-dream nights at equivalent times (A= 3.8 ; 6 df ; n.s) but when the Control measures were obtained after a REM burst (as occurs in the lucid-dream) a significant effect was also shown (A= 0.19 ; 7 df ; p < 0.05).   i/ A consistent effect When lucidity occurred it always followed within 5 seconds (mostly 2-3 seconds) of a REM burst or sequence - the preceding length of which varied from 6 to 55 seconds, mean : 21.6. A REM burst was defined here as a period of REM activity appearing in total to be separate from the relatively inactive background EOG record.   VII.3.B Cross-correlational findings   (a) 9 measures based on 8 lucid-dreams (Figure VIII.18, page 181) The following pairs of measures attained statistically-significant levels of association, using N-2 degrees of freedom, 2-tailed, and the 5% level of significance as a minimum. 1/ Length of pre-signal REM burst & Heart-rate during the lucid-dream (1st minute) r= 0.87 P < 0.01 The question arises as to whether this association is due to a common excitatory process independent of any dream events, or whether dream events and mentation are causing the increased REM activity and consequent elevation in heart-rate. Subjectively the S reported that he always felt excited at ach­ieving lucidity, but this feeling may have been 'rationalised' from a burst of heightened autonomic activity.   2/ Heart-rate during the lucid-dream and Heart-rate before lucidity (1 minute) r= 0.79 P < 0.02 This association tends to indicate that there is a strong autonomic influence in lucid-dreams. Order and calm, reflected in a slower heart-rate, do not generally result from the insight that one is dreaming.   3/ Latency of lucid-dream into the REMP and REM amount before lucidity, in that REMP r= -0.77 P < 0.05This finding indicates that the longer the pre-lucid REM period, the less REM occurs. This is not necess­arily characteristic only of lucid-dream REMPs. In general, the later REMPs are indeed longer, with less REM.   4/ Length of pre-signal. REM burst and Time of night of lucid-dream r= -0.70 P < 0.05 This association states that later lucid-dreams have shorter pre-signal REM bursts. This probably reflects the fact that REM is more sparse in later REMPs.   (b) 18 measures based on 6 lucid-dreams (Figure VIII.19, page 182). Pairs of measures attaining significant levels of association were : 1/ Lights-out time and Sleep-onset time (lst k-complex) r= 0.97 P < 0.01 2/ Time of night of lucid-dream and Total sleep time r= 0.90 P < 0.02 These associations are obvious and expected. 3/ REM amount in lucid-dream and Heart-rate in lucid-dream r= 090 P < 0.02 This association mirrors the link between REM and heart-rate found too in ordinary dreams.(Gassel et al, 1964). 4/ REM amount before lucidity and Disturbances in lucid-dream REMP r= -0.88 P < 0.02 Disturbances in a REMP cause interrupt­ions of REM activity. This finding supports that common observation. 5/ Heart-rate in lucid-dream (1st minute) and Length of pre-lucid REM burst r= 0.87 P < 0.05 This strong association was discussed in the previous section on measures from 6 lucid-dreams. 6/ Latency of lucid-dream into REMP and % Stage REM r= 0.84 P < 0.05 This is an expected association. 7/ Duration of lucid-dream and % Stage 4 r= -0.83 P < 0.05 This association states that the longer the lucid-dream the less the amount of Stage 4 sleep precedes it. Since the percentage of Stage 4 sleep declines in the night it might seem that this link indicates that later lucid-dreams are longer, however the association between time of lucid-dream and duration is not significant. Therefore, this link perhaps suggests that the interplay, of sleep stages can influence the duration of lucid-dreams in some way. 8/ Heart-rate before lucidity and REM amount in lucid-dream r= 0.82 P < 0.05 This is a physiological link as pre­viously discussed. CHAPTER VIII THE 1st A.W. STUDY - ELECTRO-PHYSIOLOGICAL FINDINGS   VIII.4 DISCUSSION Findings : The 8 lucid-dreams analysed with Control dreams for this study revealed important basic information about lucid-dreams - at least in this one S. It has been shown that the dreams all occurred in Stage REM sleep. Taking this to be approxim­ately 25% of total sleep time, probability of these lucid-dreams all occurring in REM by chance is : P = (0.25) ^8 (i.e. 0.000015). Lucidity was always initiated not in quiescent, tonic REM, but in a burst of 'fully blown' REM activity, so reinforcing the link between the lucid-dream and REM sleep. In addition, the lack of any sleep disturbances just before lucidity points to the S being firmly established in Stage REM. This subject's typical lucid-dream lasts some 2.5 minutes, happening at 6.30 a.m. - some 24 minutes after the beginning of the REMP. Lucidity-signals always appeared at the end of a REM burst averaging 22 seconds in length. There seemed to be no difference between lucid-dream nights and Control nights. The 9 measures compared showed no special characteristics for the lucid-dream recording nights. Similarly, lucid-dream REMPs did not differ significantly from Control REMPs. The finding that heart-rate increased significantly after lucidity from that beforehand appeared to be an interesting discovery, especially as Control periods from non-lucid-dreams did not reveal such an increase. However, it was later realised that a better Control would be to compare 1-minute samples in non-lucid-dream REMPs before and after a REM burst - since lucidity is always initiated in a REM burst. This control comparison was performed, and a similar increase in heart-rate was apparent in those measures too (A= 0.19 ; 7 df ; P < 0.05). Hence, the increase in heart-rate during lucidity is almost certainly a function of some general excitatory process associated with REM rather than due to psychological factors within the dream. A further com­parison of heart-rate in the lucid-dream and in Control REMPs after a REM burst revealed no significant difference between these measures (A= 0.5 ; 6 df ; n.s.). Similarly, the finding that REM activity in the lucid-dream, although not greater than that before lucidity, was significantly different from that measure in Control REMPs, afforded a different picture when the Control measures were after a REM burst (matched for similar time of night and duration). No effect was then observed (A= 0.613 ; 7 df ; n.s.). These findings ill­ustrate that care must be used in interpreting such results, and that suitable, comparable, Control situations must be employed. The results also demonstrate that it is possible to predict, fairly accurately, the relative heart-rate in a lucid-dream from the length of the pre-lucid REM burst. This offers compelling evidence that the emotional level of the dream is pre-determined by neuro-physiological processes that cause REM bursts and associated increase in heart-rate. It is possible that a long pre-lucid REM burst is associated with an 'active' dream, with high heart-rate. This could interfere with attempts by the dreamer to orientate and think clearly within the lucid-dream. Since later lucid-dreams tend to have shorter REM bursts these dreams may be more lucid than earlier ones. Two other results are potentially important. The first is the finding that the amount of REM in the lucid-dream REMP correlated significantly with the latency of the lucid-dream, suggesting that perhaps a certain minimum amount of REM is required before lucidity may be attained. It could be that : Physiologically : a certain level of 'cortical bombardment' from the reticular formation has to occur, or that Psychologically : enough dream-thoughts must be allowed before the topic of authenticity of the environment is touched upon. The second result is the link between % Stage 4 and duration of the lucid-dream. The consequent assumption that later lucid-dreams might be longer was shown not to be tenable. A likely explanation is that there is more room, for the lucid-dream when Stage 4 sleep is reduced.   Other points : 1/ The scanning hypothesis and eye-signals Regarding the 'scanning' hypothesis of eye-movements in dreams, the demonstration that deliberate eye-signals can be made in REM sleep must support that notion - certainly, the motor ability is present. This author is of the opinion that most REMs consist of spontaneous non-scanning movements but that some are definite object-fixating or scene-scanning movements.   2/ The pre-lucidity REM burst and non-maintenance of lucidity The consistent finding that ocular-signals were always produced after a REM burst is of importance in understanding bow lucidity is initiated. A reasonable explanation is that spontaneous neuro-physiological activity causing REMs also stimulates the cortex via the reticular formation, to a level where consciousness as reached within the dream. Once achieved, however, consciousness is not maintained indefinitely. The subject reports that either a slow 'sinking back' into ordinary dream unconsciousness occurs, or the subject experiences a feeling of waking. The author too, in his own lucid-dreams experiences an inexorable decline of lucidity despite performing such lucidity-reminding tasks such as looking at one’s own (dream) hands or thinking repeatedly 'This is a dream'. Therefore, presumably, lucidity in dreams is not a welcome phenomenon to the organism and is physiol­ogically suppressed. This is understandable since if people were always conscious in dreams it might be difficult to differentiate dream from reality. In addition, people might tarry too long in that pleasant state and neglect the real world.   3/ Is it possible to fake the signals ? The presence of REMs and low-tonus EMG of REM sleep could just conceivably be faked by a knowledgeable S, however the EEG is very characteristic and is not considered possible to emulate when awake (Schwartz & Fischgold, 1960). Never­theless, to be totally thorough, an experiment in which simulating Controls attempt to fake the signals was called for (This is reported in Chapter XI).   4/ Is a 2-way communication system possible between S and E? The study so far has demonstrated that the S can communicate to the outside world. An obvious progression would he to establish a 2-way link. This would be important in determining for instance the mental functioning of the lucid-dreamer by the S's direct involvement in various tests of memory, intelligence, personality, etc., and also to demonstrate the temporal continuity of lucidity in the dream (Experiments in 2-way communication are described in Chapter XIII). CHAPTER VIII THE 1st A.W. STUDY - ELECTRO-PHYSIOLOGICAL FINDINGS   VIII.5 CONCLUSIONS The ocular signalling method of marking lucid-dreams in the polygraphic record and reporting information to the external world, has enabled, for the first time, much basic knowledge about lucid-dreams (in this subject at least) to be ascertained. An important point is the fact that they are indeed dreams from Stage REM sleep. It is recognised though that this may not be so in all people who report lucid-dreams There may cases of waking imagery superficially masquerading as lucid-dreams. Another factor that needs to be borne in mind is that lucid-dreams might also occur in other stages of sleep, but that the musculature may be too 'sluggish' for signalling. Added to which the memory of such events might be erased on waking from that state. Nevertheless, a Stage REM lucid-dream phenomenon has been discovered in man. Having catalogued the electro-physiological data from this 1st A.W. study, the next Chapter presents the Psychological information which has been determined.   TABLE VIII.1 1st A.W. STUDY SLEEP-RECORD MEASURES - 8 LUCID DREAMS (A - H) & 8 CONTROL DREAMS ( I- P) CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   This subject tape-recorded his lucid-dreams and some ordinary dreams, on waking, over a 170 day period from August 1st 1976 to January 17th 1977. For part of that time a questionnaire was filled-in after each lucid-dream, and a diary was kept of each day’s events. The data obtained is analysed here. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.1 FREQUENCY DATA     XIV.1.1 INTRODUCTION A record of those nights on which lucid-dreams are experienced enables these questions to be answered : a. What is the subject's overall frequency of lucid-dreams during the period ? b. Are there any periodicity effects, as reflected in a recurring figure of days between lucid-dreams, or a bias for certain days of the week ? Any finding of periodicity could point to underlying causative factors concerning these dreams, in terms of regular external (environmental) and / or internal (psychological / physiological) influences. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.     XIV.1.2 RESULTS Figure XIV.1 (below) presents those days during the sampling period on which lucid-dreams occurred. Mostly they were spread out, but some bunching is apparent at the end of the sampling period. The overall frequency of lucid-dreams was one per 4 or 5 days (36 / 170). Figure XIV.2, illustrates the fre­quency of lucid-dreams for different intervals between their occurr­ence. It can be seen that one-quarter (9 / 36) of them happened 4 days after a previous one, and one sixth (6 / 36) were experienced the next day after a lucid-dream : these are the most common frequ­encies. The maximum duration between lucid-dreams was found to be 12 days. Figure XIV.3 shows that the distribution of lucid-dreams was more or less equal for the different days of the week, with a slight non-significant decrement on Wednesdays   FIGURE : XIV.1 LUCID-DREAM OCCURRENCES OVER 170 DAYS :     CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.1.3 DISCUSSION This particular subject does appear to have a degree of cyclicity in the frequency of lucid-dreams. The 4 day periodicity effect accounts for 25% of the data. It would be of interest to discover whether this is linked to any other aspect of the subject's behaviour. For instance, perhaps lucid-dreams are associated with peaks of cortical activity at 4 day intervals. This might be reflected in, say, his personality i.e. he could be more introverted on those days (Chapter XVI). A causal link too might be established by a detailed analysis of the subject's daily act­ivities. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.2 DIARY DATA   CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.2.1 INTRODUCTION The concept that the day's events can affect that night's dreams is a familiar one. It is a common observation for most people, and is an integral part of Freudian and Jungian dream theory (IV.5 ; IV.6). Two questions therefore spring to mind vis á vis lucid-dreams : 1/ Is lucidity in dreams linked to specific events in the day before ? These events may have some special psychological significance for the person and a replay in dreams may cause cortical arousal so perhaps initiating lucidity. 2/ On days before lucid-dreams, does the person experience more general stimulation (activity / emotional arousal) ? This follows on from the notion that lucidity arises from cortical excitation. Many interesting events in the day might lead to more cortical activity in dreams, hence lucidity. A study of diary data might establish whether these two factors are important or not. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.2.2 METHOD The subject (A.W.) was requested to keep a diary of each day’s events over a period of several weeks. A continuous record was not maintained, however enough data was collected for 8 days followed that night by a lucid-dream, and 8 other days which were not followed by a lucid-dream. The S was asked to put entries into the diary on an hour by hour basis, and in no circumstances to fill in the data for a previous day in case the subject's personal ideas on any possible links between events and lucid-dreams affected his diary entries. Control days were randomly selected, but were matched for day of week for obvious reasons. Measures: A measure of experienced-stimulation during the day was derived from the diary information by scoring the amount of activity and emotional arousal. Three naïve judges rated lucid and non-lucid dream days, randomly presented. The mean score was rounded off to the nearest whole number. The criteria were : 1/ Any recorded non-routine activity, 1 point. 2/ Any recorded emotional event, 1 point. Any link between references to specific activities or persons and lucid-dreams would be investigated by noting their presence or absence in the 2 sets of E and C data. For the 8 separate items of E and C data, the presence of a reference in 8 cases in one set and 0 cases in the other, 1and 7 cases, P would be approaching significance (P=0.070, 2 tailed.) CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.2.3 RESULTS (a) Diary references and lucidity. The various items compared are shown in the Appendix, page 411. No specific activity or reference to a person appears to be linked with lucidity in a dream that night. Of the 10 items compared, the greatest discrepancy concerns reference to alcoholic drink. Two of the E days and 6 of the C days have such an entry however, the trend is not significant. (b) Daytime stimulation and lucidity. An independent t-test was used to compare the B and C data scores on general daytime stimulation (See table below). There was no significant effect for E days (followed that night by a lucid-dream) compared with C days (not followed by a lucid-dream. (t= 1.77 ; df=14 ; n.s.). TABLE XIV.1 DIARY DAY-TIME STIMULATION SCORES   CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   TABLE XIV.1 DIARY DAY-TIME STIMULATION SCORES CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.2.4 DISCUSSION The amount of general stimulation was found not to be significantly higher on days before lucid-dreams, but there appears to be a trend that way. The information though concerns only this subject and is open to all the criticisms of subjective data, however it points to an association which should be further investigated. The concept, stated in the Introduction, that perhaps a more interesting (to the subject) day leads to greater cortical arousal in dreams and so makes lucidity more likely, requires further study. The other aspect of this study has shown that certain specific events in the day before do not appear to be associated with lucidity in dreams that night, although the level of analysis was not detailed. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.3 POST-LUCID-DREAM QUESTIONNAIRE CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.3.1 INTRODUCTION The purpose of the post-lucid-dream questionnaire was to obtain subjective information on various aspects of the dream, before and after lucidity, which were described in Chapter V. The items would be : imagery brightness and clarity, emotional level, clarity of thought, and bizarreness of the dream. In addition, questions would be asked on estimated duration, likelihood of the events actually happening in real life, and ability to control the dream. These were to ascertain how different exper­ientially, lucid-dreams are from their ordinary-dream 'matrix', and to investigate how the phenomenon of dream control correl­ates with any of these data. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.3.2 METHOD The subject was asked to fill in the questionnaire (See Appendix, pages 408-410) on waking after a lucid-dream. This was usually performed after the subject had recorded an account of the dream. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.3.3 RESULTS The subject completed the questionnaire after 29 lucid-dreams. The answers were in the form of numerals on a scale from 1 to 7. Raw data is shown on page 265. The time estimation data is in seconds. Since a few persons who reported experiencing lucid-dreams told the author that thinking in them could sometimes be 'clearer than when awake', the scale for items 6 and 7 was extended to cover that. 'Before and after' data was subjected to Sandler's A test, and the entire data was cross-correlated using a computer program (N3IC - of the University of Liverpool Computer Laboratory). The resultant correlational matrix is shown on page 266. The mean ratings for the first 4 items in the questionnaire (see Table XVI.2 on page 265), concerning the brightness and clarity of imagery before and after lucidity, were around the centre of the scale (1= no imagery, 7= like in strong sunshine). Mean brightness before lucidity : 3.7 Mean brightness after lucidity : 3.8 Mean clarity before lucidity : 4.3 Mean clarity after lucidity : 4.4 The mean estimated duration of the lucid-dreams was 156 seconds, which is very similar to the mean duration of 153 seconds found in the sleep-lab study. The estimated range was from 5 to 360 seconds. A t-test showed that the sleep-lab duration figures were not significantly different from the mean estimated duration of the large sample of home lucid-dreams (t=0.09 : 7 df ; n.s.). The mean level of quality of thought was found to be a little less than in waking life both before and after lucidity : Mean clarity of thought before lucidity : 2.6 Mean clarity of thought after lucidity : 3.3 (Scale : 1=very unclear ; 4= same as when awake ; 7=much more clear). The mean level of emotionality was fairly low both before and after lucidity : Mean emotionality of dream before lucidity : 2.8 Mean emotionality of dream after lucidity : 3.3 (Scale : 1=unemotional, 7= extremely emotional). The bizarreness of the dreams was, on average, was not very great before or after lucidity : Mean bizarreness before lucidity : 2.8 Mean bizarreness after lucidity : 3.6 (Scale : 1=not bizarre, 7=extremely bizarre). On average, the dream events concerned things which could possibly happen in waking life : 4.4 on scale (1= very unlikely to happen, 4= possible, 7= very unlikely). The mean level of dream-control after lucidity was 4.7 on the scale (1= no control, 7=very easy to control).   Sandler's A test findings : 1 Brightness of imagery before and after lucidity : A=0.56. Not significant. 2 Clarity of imagery before and after lucidity : A=0.33. Not significant. 3 Clarity of thought before and after lucidity : A=0.06. P<0.0001 Subjectively, thinking was significantly clearer after lucidity. 4 Emotionality before and after lucidity : A=0.14. P<0.01 Emotionality in the dream was subjectively greater after lucidity. 5 Bizarreness before and after lucidity : A=O.15. P<0.01 Dreams were more bizarre, subjectively, after lucidity. Listed below are pairs of measures which attained overall significant levels of association using N-2 degrees of freedom, 2-tailed, and the 5% level of significance as a minimum. (Table values (overall data) : 5%= 0.37 ; 2%= 0.43 ; 1%= 0.47). The data was also split (days 1-15, 16-29) and separate correlations computed to check whether any apparent overall significant correlation was consistent throughout the data. As 78 overall correlations were performed, approx­imately 4 spuriously significant correlations might be expected by chance alone (at the 5% level). (Split data table values : (13df) 5%= 0.51, 2%= 0.59. 1%= 0.64) (12df) 5%= 0.53, 2%= 0.61, 1%= 0.66).   Correlations : (SEE TABLE XIV.3,, and FIGURE XIV.4, ). 1. Brightness of imagery before and after lucidity : r= 0.91 P<0.01 Split : 13df : 0.84 ; P<0.01 12df : 0.98 ; P <0.01 2. Brightness after lucidity and clarity before lucidity : r= 0.50 P<0.01 Split : 13df : 0.56 ; P<0.05 12df : 0.42 ; n.s. 3. Brightness after lucidity and clarity after lucidity : r= 0.44 P<0.05 Split : 13df : 0.58 ; P<0.05 12df : 0.21; n.s. 4. Clarity before and after lucidity : r= 0.90 P<0.01 Split : 13df : 0.91 ; P<0.01 12df : 0.89 ; P<0.01 These 4 findings, with the Sandler's A test results essentially reflect the fact that the dream imagery does not subjectively appear to change in brightness or quality with the onset of lucidity. The non-significant correlations of some of the split data in items 2 and 3 of the list above indicates that the overall results are perhaps dubious. 5. Clarity of thought before and after lucidity : r= 0.55 P<0.01 Split : 13df : -0.26 ; n.s. 12df : 0.82 ; P<0.01 This overall finding indicates that the greater the clarity of thought before lucidity, the more so after that point. However, the split halves of the data give different results, so this finding is suspect. 6. Emotionality before and after lucidity : r= 0.60 P<0.01 Split : 13df : 0.51 ; P<0.05 12df : 0.70 ; P<0.01 The dream is subjectively more emotional after lucidity and the level of emotionality is dependent on the degree before lucidity. 7. Bizarreness before and after lucidity : r= 0.50 P<0.01 Split : 13df : 0.76 ; P<0.01 12df : 0.15 ; n.s. Overall, bizarreness of the lucid-dream appears to be linked to that measure before lucidity. However, the split analyses give different results, so the result is not reliable. 8. Bizarreness before lucidity and likelihood of events : r= -0.43 P<0.05 Split :13df : -0.50 ; n.s. 12df : -0.31 ; n.s. This overall finding is not reliable. 9. Bizarreness before lucidity and controllability of the lucid-dream : r= -0.43 P<0.02 Split : 13df : -0.33 ; n.s. 12df: -0.60 ; P<0.05 Whilst both split-data findings give negative correlations, they are different values, hence the overall finding is suspect. In 16 of the 29 dreams the dream-situation was indoors.   TABLE : XIV.3 STATISTICALLY SIGNIFICANT( OVERALL) CORRELATIONS AND SPLIT-DATA CORRELATIONS (Split-data correlations in brackets) 0.91 - · BRIGHTNESS BEFORE LUCIDITY. (.84,.98) L. “ AFTEI~ “ 10:50 1 ~ 0.90 r3’ CLARITY BEFORE LUCIDI2Y. -‘ (.58,.21) (.91,.89. 14. H AFTER “ J   5. DURATION. 0.55 r6~ CLARITY OF THOUGHT BEFOI?E LUCIDITY. (-.26,.82) 17. “ U U AFTER o.6o r8~ EMOTIONALITY BEEORE LUCI1JIi~Y. (.51,.7o) 19. U AFI1ER 10. BIZARRENESS BEFORE LUCIDITY. 1+ (.76,.1s)li~1 AFL~ER ] (-.50, -0.43 -.31) (-.33, 12. LIKEIJIHOOD OF EVENTS IN REAL LI -.6o) 13. CONTROLLABILITY OVER DREAM CONTENT AND ACTION. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.3.4 DISCUSSION The results could be a function of the subject's preconceptions, biases, responding sets and various other psychol­ogical artefacts. However, taking them at face value as being truthful observations of subjective experiences, the findings indicate that the brightness and clarity of the dream imagery does not change after lucidity, but that clarity of thought, emotionality and bizarreness do increase at that juncture. The subjectively reported increase in emotional level could be explained by the subject's excitement at achieving lucidity or it is perhaps more likely physiologically based in that the Heart-rate has been found to increase after a REM burst - which invariably preceded lucidity in the sleep-lab (Chapter VIII). The physiological state might therefore direct the emotional level of the dream. The significant increase in reported bizarreness of the dream after lucidity is somewhat unexpected, however it may be that the greater insight and observation after lucidity may expose bizarreness in greater relief. The split-data correlations indicate that some overall significant associations may be spurious. A few robust correlations survive : The subjective levels of brightness and clarity of imagery were dependent on those levels beforehand. The subjective level of emotionality in the lucid-dream was associated to the level before lucidity. This level had already been shown to increase significantly on lucidity. Thus, in these 3 measures the pre-lucid state was crucial in that it set the level for the lucid-dream. Further data from this subject and others should show whether the less reliable correlations are genuine. CHAPTER XIV ADDITIONAL DATA FROM SUBJECT A.W.   XIV.3.5 OVERALL CONCLUSIONS   1. A degree of cyclicity (of 4 days) was apparent in this subject's reported lucid dreams. 2. Lucid-dreams occurred on all days of the week, with no apparent significant bias. 3. Lucid-dreams were found to occur more often after a day of above average stimulation (i.e. more non-routine activity or emotional events). 4. Specific events the day before, did not appear to be linked with lucid-dreams that night. Subjective brightness and clarity of the dream imagery remain the same after lucidity as before. Subjective clarity of thought, emotionality and bizarreness in a dream increase significantly at lucidity. Subjective brightness and clarity of the imagery, and emotionality of the lucid-dream are significantly associated with those levels before lucidity. From the data supplied by this subject it would seem that physiological factors are important regarding lucid-dreams. They could affect their very occurrence through cyclical rhythms and the amount of cortical excitation in the day before. Also, the emotional and imagery levels appear to be dependent on those levels existing prior to lucidity. It was decided to attempt to discover whether the frequency of reported lucid-dreams (in a large group of persons) correlated with several imagery and sleep phenomena, to see if any links were apparent which might throw further light on lucid-dreams. The next Chapter describes that work.