Five years ago, I developed a simple technique for maintaining conscious awareness during the transition from waking to sleep. The method is to count to yourself ("one, I"m dreaming; two, I"m dreaming," and so on) while drifting off to sleep, maintaining a certain level of vigilance as you do so. The result is that at some point-say, "forty-eight, I"m dreaming"-you will find that you are dreaming! The "I"m dreaming" phrase helps to remind you of what you intend to do, but it is not strictly necessary. Simply focusing your attention on counting probably allows you to retain sufficient alertness to recognize dream images for what they are, when they appear. This and similar techniques apparently work best for people who tend to fall asleep rapidly, and frequently experience sleep-onset (hypnagogic) dreaming.
I gained the most information about this technique through tutoring one of our oneironauts in its practice. Laurie C. regularly experiences vivid dreaming at sleep onset. Some of her sleep-onset dreams are like the "snapshots" people typically think of as characteristic hypnagogic imagery. Others seem more like full-blown dream films, with her vividly felt presence and active partic.i.p.ation in what are sometimes quite extensive sequences of scenes.
Knowing that Ouspensky had been able to maintain awareness in the hypnagogic state, I suggested to Laurie that she too ought to be able to be lucid during sleep-onset dreaming. I helped her practice by watching her fall asleep while she made the effort to retain consciousness. She would often fall asleep within fifteen to thirty seconds, as I could tell by her rolling or rapid eye movements and body twitches. At that point, I would wake her with the question, "What was just happening?" She almost always recalled a vivid dream scene, but rarely became lucid at the time. It seemed to me she needed a reminder. So I asked her to silently repeat the phrase "I"m dreaming" while counting to herself as she fell asleep. And it worked! She would be repeating "one, I"m dreaming; two, I"m dreaming; ... forty-four, I"m dreaming," until suddenly realizing she was dreaming.
After about a week of practice, Laurie was able to dispense with the "I"m dreaming" phrase, and found that simple counting to herself worked as well. In a surprisingly short time, she was able to become lucid in her sleep-onset dreams almost at will.
At this point, we decided to observe her in the sleep lab for three consecutive nights. On these nights, Laurie made an effort to retain consciousness while entering sleep-onset dream states.
On each of these experimental sessions, she repeatedly rested quietly, but vigilantly, and counted to herself ("one, two, three ...", until she began to dream. She usually awakened five to ten seconds later and tape-recorded a description of what she had just experienced. Laurie reported having been lucid in twenty-five of the forty-two resulting dreams. Visual inspection of the polygraph records showed that all of these "dreamlets" (I use the term because none lasted more than several seconds) occurred, whether they were lucid or not, during NREM Stage 1 sleep, with slow eye movements. The following is a sample: "I was in the grocery store, going down an aisle; only I was standing on a cart. It was whizzing real fast. As I went by the c.o.ke and Pepsi bottles, I realized that I was dreaming. I remembered to look at my hands, but they wouldn"t move up to eye level."
Laurie"s inability to look at her dream hands may point to a significant difference between these sleep-onset lucid "dreamlets" and REM lucid dreams. During REM lucid dreams, the dreamer normally has complete volitional control of his or her dream body. I had wondered if the same would be true at sleep onset, so I asked Laurie to carry out a prearranged dream action of putting her hands before her face when she found herself conscious of dreaming. This seemingly simple task proved impossible, at least in this one case; obviously, more research is needed in order to settle the issue.
An important factor influencing the kind of results you are likely to obtain with the foregoing technique is timing. Rather than attempting to enter the lucid dream state at the beginning of your sleep cycle, you would do much better to try later in the night, toward morning, especially after already awakening from a dream. This is because the lucid dreams characteristic of sleep onset at the beginning of the night typically take place in NREM Stage 1 sleep, and hence rarely seem to last more than a few seconds. In contrast, when these techniques are practiced upon awakening from REM periods later in the night, it is often possible to return to REM sleep and experience extended, full-blown lucid dreams. However, the optimal time for entering REM lucid dreams directly from the waking state may be in the afternoon. I base this suggestion on the fact that in my personal record, the proportion of waking-initiated lucid dreams to dream-initiated ones is six times higher during my afternoon naps than in nocturnal sleep.
The various techniques described above for entering the dream state from the waking state work on the same fundamental principle: you lie in bed deeply relaxed but vigilant, and perform a repet.i.tive or continuous mental activity upon which you focus your attention. Keeping this task going maintains your inner focus of attention, and with it your wakeful inner consciousness, while your drowsy external awareness diminishes and finally vanishes altogether as you fall asleep. In essence, the idea is to let your body fall asleep while you keep your mind awake.
There is another, and for most people far easier, way to become lucid in a dream: become very familiar with your dreams, get to know what is dreamlike about them, and simply intend to recognize that they are dreams while they are happening. Evidently, simply intending to recognize that one is dreaming is enough to increase the frequency of occurrence of lucid dreams.
In the West, the earliest account of learning lucid dreaming was provided by the Marquis d"Hervey de Saint-Denys, who taught himself through the following method. By the age of thirteen, he had become fascinated with his dreams and devoted a great deal of time to recording and sketching them. He explained that:
as a result of thinking about my dreams during the day, and a.n.a.lyzing and describing them, these activities became part of the store of memories of waking life on which my mind drew during sleep. Thus one night I dreamt that I was writing up my dreams, some of which were particularly unusual. On waking, I thought what a great pity it was that I had not been aware of this exceptional situation while still asleep. What a golden opportunity lost-I thought. I would have been able to note so many interesting details. I was obsessed by this idea for several days, and the mere fact that I kept thinking about it during the day soon resulted in my having the same dream again. There was one modification, however: this time the original ideas summoned up by a.s.sociation the idea I was dreaming, and I became perfectly aware of this fact. I was able to concentrate particularly on the details of the dream that interested me, so as to fix them in my mind all the more clearly on waking.9
After another year of obsessive preoccupation with his dreams, the Marquis reported that he was having lucid dreams almost every night.
R. T. Browne, one of the contributors to Volume II of Narayana"s Dream Problem, described essentially this same method. a.s.serting that "under certain conditions" it is "possible for the dreamer to remain cognizant of the fact that he is dreaming during the dream state ... one of the means by which he may accomplish this is by placing in his so-called subconscious mind the suggestion, clear-cut and positive, that he will be conscious of the fact, under a given set of circ.u.mstances, that he is dreaming."
This method was tested in the early 1970s by Patricia Garfield. She merely told herself, before going to sleep: "Tonight, I will have a lucid dream." Garfield said that with this method she obtained a "cla.s.sical learning curve, increasing the frequency of prolonged lucid dreams from a baseline of zero to a high of three per week."10She used the method for five or six years, during which time she had an average of four or five lucid dreams per month. These results indicate that merely telling oneself to have a lucid dream can provide a starting point, at least, for deliberately inducing lucid dreams.
Before beginning my own efforts to learn to dream lucidly, I had had occasional spontaneous lucid dreams over the years, especially as a young child. At about the age of five, I had a series of dreams that I would intentionally redream on successive nights. I vividly recall a scene from one of them, as described in Chapter 4, in which I experienced momentary anxiety about having been underwater for a long time-but then I remembered that in dreams I could breathe underwater. How long I continued these practices I do not know, but the next lucid dream that I can recall took place almost twenty years later. For several years I experienced occasional lucid dreams (about one per month) that were sufficiently intriguing to persuade me to undertake a careful study of the phenomenon.
During the first year and a half of my research, I used self-suggestion for lucid-dream induction, with results equivalent to Garfield"s.
By the end of phase one, I had observed two factors that seemed to be a.s.sociated with the occurrence of my lucid dreams. The first, and most obvious, was motivation. During this period, there were two months during which I reported, respectively, two and three times more lucid dreams than the average for the rest of this period. During the first month, September 1977, I was preparing a Ph.D. dissertation proposal in which I claimed that I ought to be able to learn to have lucid dreams at will. During the second month, January 1978, I was attempting (successfully) to have lucid dreams in the sleep laboratory. During both of these months I was highly motivated, since I felt challenged to demonstrate the practicality of a laboratory study of lucid dreaming. However, I found it impossible to maintain this high level of motivation, as is seen by the decline in lucid-dreaming frequency following these two months.
Gradually, more self-observation led to the realization that a second psychological factor was involved: the intention to remember to be lucid during the next dream. This clarification of intention was accompanied by an immediate increase in the monthly frequency of my lucid dreams. Once I discovered that memory was the key to lucid dreaming, further practice and methodological refinements allowed me to arrive, within a year, at my goal: a method by which I could reliably induce lucid dreams.
Mnemonic Induction of Lucid Dreams (MILD)
MILD is based on nothing more complex or esoteric than our ability to remember that there are actions we wish to perform in the future. Aside from writing ourselves memos (a device of little use here, for obvious reasons!) we do this by forming a mental connection between what we want to do and the future circ.u.mstances in which we intend to do it. Making this connection is greatly facilitated by the mnemonic device-the memory aid-of visualizing yourself doing what it is you intend to remember. It is also helpful to verbalize the intention: "When such-and-such happens, I want to remember to do so-and-so." For example: "When I pa.s.s the bank, I want to remember to draw out some cash."
The verbalization that I use to organize my intended effort is: "Next time I"m dreaming, I want to remember to recognize I"m dreaming." The "when" and "what" of the intended action must be clearly specified.
I generate this intention either immediately after awakening from an earlier REM period, or following a period of full wakefulness, as detailed below. An important point is that in order to produce the desired effect, it is necessary to do more than just mindlessly recite the phrase. You must really intend to have a lucid dream. Here is the recommended procedure spelled out step by step:
During the early morning, when you awaken spontaneously from a dream, go over the dream several times until you have memorized it.
Then, while lying in bed and returning to sleep, say to yourself, "Next time I"m dreaming, I want to remember to recognize I"m dreaming."
Visualize yourself as being back in the dream just rehea.r.s.ed; only this time, see yourself realizing that you are, in fact, dreaming.
Repeat steps two and three until you feel your intention is clearly fixed or you fall asleep.
If all goes well, in a short time you will find yourself lucid in another dream (which need not closely resemble the one you have rehea.r.s.ed).
The mental set involved in this procedure is much like the one you adopt when you decide to awaken at a certain hour, and go to sleep after setting your mental alarm clock. The ability to awaken in your dreams may be regarded as a sort of refinement of the ability to awaken from your dreams.
The reason for the "early morning" specified in step one is that lucid dreamers from van Eeden to Garfield have reported that such dreams occur almost exclusively during the morning hours. Our research at Stanford indicates that lucid dreaming occurs during REM periods, and since most REM sleep takes place in the later part of a night"s sleep, this is likely to be the most favorable time for lucid dreaming. Although some dreamers have successfully induced lucid dreams using MILD during the first REM period of the night, the technique seems to be most effective when practiced during the early morning after awakening from a dream.
If you find yourself just too drowsy to follow the procedure as described above, you might try to wake yourself up by engaging in several minutes of any activity that demands full wakefulness, such as writing down your dream, reading, or simply getting out of bed. This is because certain activities have been observed to promote lucid dreaming upon a subsequent return to sleep. Garfield, for example, found that "s.e.xual intercourse during the middle of the night was often followed by a lucid dream." Scott Sparrow, in contrast, reported that early-morning meditation favored lucid dreaming (but only if he did the meditation for itself and not to have lucid dreams). Other lucid dreamers indicate early-morning reading or writing to be favorable. The diversity of these activities suggests that it is not the particular activity, but the wakefulness that facilitates lucid dreaming during subsequent sleep.
However, wakefulness presents the would-be lucid dreamer with disadvantages as well as advantages. Often it is impossible to return to sleep after waking up fully. Another problem is that since proximity to REM sleep while doing MILD probably favors successful lucid-dream induction, the longer you wait before returning to sleep, the less likely it is for lucid dreams to occur. When you awaken from REM dreaming, your brain persists in a REM-like state for several minutes afterwards. Scientists have demonstrated this by awakening people from both REM and NREM sleep and having them devise stories. The stories told after awakening from REM were more dreamlike than the NREM stories. Thus, something of the REM state carries over for a few minutes after waking from a dream. If we reenter our dreams immediately after waking from REM, which is the point at which our dreaming and waking worlds are closest, it is the optimal time for carrying over our lucid-dream intentions from the waking mind into the dreaming mind.
People are likely to differ as to which of these two factors-wakefulness and REM carryover-are more effective for them, and I recommend experimenting with both when using MILD to induce lucid dreams. Once I myself learned how to use MILD, I experienced as many as four lucid dreams in a single night, and indeed seemed able to attain lucidity on any night that I tried it. I see no reason why the same shouldn"t be true for others.
Since motivation is an important factor in inducing lucid dreams, how can we be sure increased motivation doesn"t account for the improvements I have attributed to MILD? During all the nights I attempted to induce lucid dreams while being physiologically monitored, my motivation was very high. Using self-suggestion, I had only one lucid dream in seven nights of laboratory recordings, but when I began practicing MILD, I had fifteen lucid dreams in thirteen recording nights. It should be clear that it is the method, and not merely the motivation, that accounted for these results.
MILD also seems to work well for others, especially those who meet the requirements of high motivation and excellent dream recall. "High motivation" means having a strong desire to develop the skill of lucid dreaming, and by "excellent dream recall" I mean being able to awaken from (and remember) dreams two to three times per night or more. Students in my workshops and courses have almost always succeeded with MILD if they met these two conditions. Two of my students increased their lucid-dream frequency from less than one per month to about twenty per month during an eight-week course. Even the average student had three or four lucid dreams in the same two-month period. All this should make it clear that it is possible to learn to have lucid dreams. What one dreamer can do, others can do as well.
Undoubtedly, the future will see the development of much more effective techniques for lucid-dream induction that promise to make this world available to anyone who needs or desires it. Who knows-perhaps entry into lucid dreams will one day be no more difficult than falling asleep. (Although many will still say that that is no easy matter!) Future Access to the Lucid Dream State
At present, it is possible to point to several techniques that could be developed to induce lucid dreams. One is hypnosis. "Auto-suggestion," or implanting in oneself the command to do something, is a form of hypnosis, and it is involved in practicing MILD. Many people have found it at least moderately effective in inducing lucid dreams. For the fortunate minority who are easily hypnotizable, post-hypnotic suggestions to have lucid dreams may be more effective when given by a hypnotist, rather than when given by oneself. Although I am only moderately hypnotizable, I have been hypnotized three times and given post-hypnotic suggestions to have lucid dreams and on two of the three occasions they worked.
Along the same lines, Charles Tart has mentioned some preliminary work suggesting "that post-hypnotic suggestions may have some potential in inducing lucid dreaming."11 I think one could change Tart"s "may" to "do." In my view, this is an area greatly deserving of cultivation, since hypnosis could help provide access to lucid dreaming for those hypnotizable individuals who might find the state very useful, but who may not be able to get there on their own steam. Such a technique would probably greatly further the therapeutic applications of lucid dreaming (see chapter 7).
A Ph.D. dissertation by Joe Dane, recently of the University of Virginia at Charlottesville, provides the strongest support yet for the possible usefulness of post-hypnotic suggestion as a lucid-dream induction method. Dane"s fifteen female subjects-all scoring above the median level on a measure of hypnotic susceptibility, all recalling at least one dream per month, and all claiming no prior experience of lucidity-were given post-hypnotic suggestions to dream lucidly when they were recorded on a single night. Fourteen of the fifteen reported at least a minor episode of lucid dreaming. Contrary to most other studies, Dane"s turned up more reports of lucid dreaming from non-REM sleep than from REM sleep. Some of the NREM lucid dreams resembled the "dreamlets" described earlier in this chapter, so it is difficult to evaluate just how significant these results actually are. Counting only the five unambiguous REM lucid dreams verified by eye-movement signals gives a considerably more modest success rate of thirty-three percent. However, even that seems most impressive when it is taken into account that the subjects were all previously non-lucid dreamers. I believe that Dane has significantly advanced the state of the art with respect to lucid dream induction by means of post-hypnotic suggestion.
The use of some particular element of dream content (for instance, my "multiplying contact lenses") as a lucidity cue was mentioned in Chapter 5. A closely related technique involves presenting, during REM sleep, an external stimulus as a lucidity cue, since it is well known that environmental stimuli are at times incorporated into dreams. Almost everyone has had the experience of hearing an annoying sound, such as a neighbor"s buzz saw, only to awaken a moment later to find that what had really been making the noise was the equally annoying alarm clock. If we have a kinder system for awakening, music or the news may find its way into our dreams. Stimuli in the rest of the sensory modes (smell, touch, sight, temperature, and possibly other senses as well) are also occasionally incorporated in dreams, and any of these sensory pathways could carry a reminder to the dreamer that he or she is dreaming.
This all might seem contradictory to the reader accustomed to thinking of sleep as being "dead to the world." But the fact is that the sleeping brain maintains a degree of contact with the environment, a.n.a.lyzing for meaning the information about external events that is received through the senses. After all, we are able to awaken when our own name is called, but remain asleep when someone else"s name is called or an airplane flies overhead. And consider the mother who sleeps through her husband"s loud snoring, hut awakens to her baby"s faint cries coming from another room. If we are monitoring the environment for the occurrence of significant events such as these, why couldn"t we sleep with the intention of noting some prearranged external sensory cue as a reminder that we are dreaming?
The idea occurred to me that the most direct approach would be to use as a cue a sentence stating what the dreamer wishes to become aware of: "This is a dream." I first tried this out in 1978 at the Stanford Sleep Laboratory, in collaboration with Dr. Lynn Nagel. Lynn got the short end of the deal, staying up all night monitoring my brain waves and REMs while I slept. When he observed me in REM sleep, he turned on a tape recording I had made earlier, playing it at a moderate level from a speaker next to my bed. The recorded message said, in my own voice, "Stephen, you"re dreaming" and after a few seconds added the suggestion that I continue to sleep but realize that I was dreaming. At the time, I had not been sleeping too well, still being a newcomer to the sleep lab, and it seemed to me that I was lying in bed awake. Then from the next room, I heard the voice of a doctor commenting in Germanic accents, "Amazing! Ze subject has had no REM sleep all night!" Hearing this, I was not surprised. As far as I knew, I had had no sleep of any kind. But the next moment, I was astonished to hear my own voice coming over the PA. system, announcing, "You"re dreaming!" I became lucid immediately. It had worked! I was very excited. In a dream world suddenly beautiful and more vivid than waking life, I was awake in my sleep! But a few seconds later, the recording continued with a voice now loud enough to wake the dead, to say nothing of the sleeping, "Continue to sleep"-and I awoke!
This first experiment showed us that lucid dreams could indeed be induced by direct verbal suggestion during REM sleep. The fact that in the dream I heard "You"re dreaming" loud and clear, but did not hear my name at all, is interesting. Perhaps unconsciously hearing my name stimulated my attention, allowing me to hear the rest of the message consciously.
We used my own voice to record the message for two reasons. First, we hoped that being reminded by one"s own voice would seem more like reminding oneself mentally, and second, because an earlier study found that when subjects heard tape recordings of their own voices during REM, the result was dreams in which the subjects were more active, a.s.sertive, and independent than when they heard recordings of other peoples" voices.12 Since these qualities are a.s.sociated with lucid dreaming, we hoped that hearing my own voice would reinforce these qualities and facilitate the realization I was dreaming.
This was the beginning of a series of explorations that are still continuing. We asked four people interested in lucid dreaming to spend one or two nights in the sleep lab. They each made a recording that repeated the phrase, "This is a dream" every four to eight seconds. This was played, at a gradually increasing volume, five to ten minutes after the beginning of each REM period. The subjects were instructed to signal by means of a pair of left and right eye movements whenever they heard the tape or recognized they were dreaming. The technician turned off the tape recorder immediately upon observing this eye-movement signal on the polygraph. If the subjects did not awaken by themselves within two minutes of the signal, the technician awakened them and asked for dream reports.
The tape stimulus was applied a total of fifteen times, producing lucidity in a third of the cases and, in general, one of four results:
Awakening. In the majority of cases (fifty-three percent), the subjects reported hearing the tape only after they had been awakened by it.
Incorporation with lucidity. In twenty percent of the cases, the subjects reported hearing the tape in the dream and signaling while still dreaming (now lucidly). This is the same sort of result we obtained in our first experiments with taped stimuli.
Incorporation without lucidity. The subjects twice reported dream content obviously related to the taped stimulus-without, however, becoming lucid before awakening. The most curious example of this was when a subject awoke after the tape had been played and wrote a report of his dream. I asked him at that point whether he had heard the tape, and he replied he was sure he had not. I was most surprised when I later read his written report. Near the end of his dream, he complained that someone was trying to tell him something, but he wouldn"t listen to them. What were they saying? "You"re dreaming!" Remarkably, this subject had not even recognized the phrase while writing a report of it after awakening!
Lucidity without incorporation. On two occasions, our subjects attained lucidity and signaled (while the tape was being played) without consciously hearing the stimulus in their dreams at all. This is exactly the opposite of the preceding situation. In one of these cases, having been awakened from my first REM period by the tape, I was frustrated at having had my sleep disturbed for nothing, and decided I would try to induce a lucid dream on my own during my next REM period before the tape had a chance to wake me up. So I performed the MILD technique while returning to sleep. The next thing I knew I was in a violent struggle with my father. I recognized I was dreaming and thought, "So this is the first ten minutes of REM," since I hadn"t heard any tape. This thought seemed to lead to a fading of the dream, and I soon woke up. Shortly after I wrote my dream report, the technician entered my room and asked if I had heard the tape. "What?" I said, confused. "When? What tape?" It turned out the tape had been turned on about twenty seconds before I had signaled-which was, by the way, after twenty, not ten, minutes of REM. Apparently, my unconscious awareness of the message helped me realize I was dreaming.
The results of this study also gave us a sense of the complexity and the multiplicity of variables involved. First there is the question of when is the best time to apply the stimulus, since not every moment of REM sleep seems equally suited to lucid dreaming. As for the message itself, what is its optimal form? First person-"I"m dreaming"? Second person-"You"re dreaming"? Or objective-"This is a dream"? Our research at Stanford is aimed at finding our way out of this tangle of questions, to gain a reliable means of inducing lucid dreams in people who have had no prior experience with the phenomenon.
Another significant question is whether or not a verbal cue is best. In principle, any stimulus in any sensory mode could be used as a cue to remind a dreamer that he or she is dreaming. Perhaps a melody (say, Bach"s Sleepers Awake!) might be more effective than speech. Or-since smell is the only sense that does not pa.s.s through the relay station in the brain called the thalamus, and thus may not be as inhibited as the other senses during sleep-it may be that scent would function as an especially effective cue.
A cla.s.sic study by Dement and Wolpert13 examined dream incorporation in several sensory modes. They found that tactile stimuli were more frequently incorporated into dreams than either light or sound. This suggests that tactile stimulation might provide an effective lucidity cue. In fact, in Great Britain a few years ago, Keith Hearne was promoting a "dream machine" operating on this principle. Hearne"s device measured breathing rate with a nasal temperature sensor, and applied a series of electric shocks to the dreamer"s wrist when an increase in breathing rate was detected. The problem was that if the dreamer was lucky, the shocks would occur during REM sleep, but they could just as easily come during NREM sleep or even waking. If the shocks did not awaken the sleeper, but were instead incorporated into the dream, and if the dreamer recognized the incorporated stimuli as lucidity cues, a lucid dream could have been produced. These are big "ifs" and the scant scientific research available on this machine suggests it to be only marginally effective at best. Though the idea seems a good one, the execution was apparently less than optimal for the intended purpose. I believe it is probably only a matter of time before someone perfects and markets an effective lucid-dream induction device; this is currently one of the top priorities of my own research.