The Science of Dreams

What the Laboratory Found

Dreaming entered the laboratory in 1953, when Eugene Aserinsky, a graduate student in Nathaniel Kleitman’s laboratory at Chicago, noticed that sleepers’ eyes moved rapidly at intervals through the night, and that those woken during the movements usually reported a dream (Aserinsky and Kleitman, 1953). Dement and Kleitman (1957) then woken sleepers systematically: about four awakenings in five from rapid eye movement (REM) sleep yielded a dream, fewer than one in ten from the other stages, and the length of the dream as told matched the minutes of REM that had passed. The night was found to run in cycles of roughly ninety minutes, with the REM periods lengthening towards morning, so that most of the dreaming an adult does falls in the last two hours of sleep.

The equation of dreaming with REM did not last. Foulkes (1962) asked sleepers not “did you dream?” but “what was going through your mind?” and found some report from most awakenings in non-REM sleep as well, shorter and more like thought. Michel Jouvet (1959) showed in cats that REM sleep is a state of its own, with the muscles paralysed by the brainstem; when that paralysis fails, in REM sleep behaviour disorder, the sleeper acts the dream out (Schenck et al., 1986), and a majority of those patients later develop Parkinson’s disease or a related disorder, so that the acted dream has become one of the earliest known signs of it (Schenck, Bundlie, and Mahowald, 1996).

An average adult remembers about one dream a week, women somewhat more than men, and the young more than the old (Schredl, 2008); the number rises sharply as soon as a journal is kept by the bed, which is the first thing every researcher asks of a subject.

Where Dreams Are Made

Hobson and McCarley (1977) proposed that the dream is the forebrain making what sense it can of random signals sent up from the pons during REM, the “activation-synthesis” account, which took the meaning out of dreams for a generation of neurologists. Mark Solms answered it from the clinic. In 361 neurological patients (Solms, 1997, 2000) he found that dreaming stopped altogether after damage to two regions, the junction of the parietal, temporal, and occipital lobes, and the white matter beneath the frontal lobes that carries the dopamine pathways of wanting and seeking, while damage to the pons that abolished REM did not abolish dreaming. Dreaming and REM, he concluded, are run by different mechanisms, and the dream is driven by the same circuits as appetite and motivation.

Siclari and her colleagues (2017) settled the question with 256-channel recordings of the sleeping scalp: whether a sleeper would report a dream, in REM or non-REM alike, was predicted by a drop in slow-wave activity in a “posterior hot zone” at the back of the brain, the same regions Solms had found from lesions, and the content of the dream, faces, speech, movement, lit up the parts of that zone that handle them awake. Horikawa et al. (2013) decoded the objects in sleepers’ imagery at sleep onset from brain scans; Dresler et al. (2011) had lucid dreamers clench their fists in the dream and watched the hand area of the motor cortex answer. The dream, on this evidence, is perception and action run from the inside, with the gates to the senses and the muscles shut.

What Dreams Are Made Of

Calvin Hall began collecting dream reports in the 1940s and by 1966 he and Robert Van de Castle had a coding system and a thousand dreams, five each from a hundred men and a hundred women, coded by trained judges for characters, settings, emotions, social interactions, and outcomes (Hall and Van de Castle, 1966). The norms it produced have been replicated, with small variations, in samples across decades and countries (Domhoff, 1996), and they are the yardstick the Journal’s coding is set against. Their broad findings: men’s dreams have more male characters, more physical aggression, and more strangers; women’s have more familiar people, more friendliness, and more indoor settings; four emotions in five, in both, are unpleasant; and misfortune falls on the dreamer six times more often than good fortune.

The dream is almost never a replay. Fosse et al. (2003) checked 299 dream reports against the sleepers’ diaries and found a day’s episode reproduced whole in fewer than two in a hundred: dreams take pieces of recent experience and recombine them. Pieces of the day before appear most, then again, oddly, pieces from five to seven days before, the “dream-lag” (Nielsen and Powell, 1989; Blagrove et al., 2011). Asked which dreams they had ever had, 1,181 students listed being chased, falling, school and examinations, arriving too late, and sex at the head (Nielsen et al., 2003), and the list is close to the same in every country where it has been asked.

The strongest finding of fifty years of content analysis is continuity: dreams carry on the waking concerns, people, and moods of the dreamer (Schredl and Hofmann, 2003; Domhoff, 2003). What a person does and worries about by day is what they dream of by night, with the proportions shifted, personal relationships and emotions over, say, reading and arithmetic, which hardly appear at all.

What Dreams May Be For

No theory of the function of dreaming is settled, and some researchers hold that the dream has none, being a by-product of sleep’s housekeeping. The candidates with evidence behind them: Revonsuo (2000) argued from the prevalence of threat and pursuit that dreaming evolved to rehearse danger, the “threat simulation” theory. Walker and van der Helm (2009) proposed that REM sleep strips the emotional charge from the memories of the day, “overnight therapy,” and Cartwright (2006) found that people going through divorce who dreamed of the former spouse, with feeling, were more often recovered from depression a year on than those who did not. Wamsley et al. (2010) had subjects learn a virtual maze and nap: those who dreamed of the maze, however obliquely, improved far more on waking than those who did not, as if the dream were the memory being worked on.

Hartmann (1998) read the dream as emotion given a picture, a widening of the mind’s connections under the pressure of feeling, strongest after trauma, which is why the survivor dreams of tidal waves before dreaming of the event. These accounts are not exclusive, and the evidence for each is suggestive rather than conclusive; what they share, and what separates them from the activation-synthesis view, is that the dream’s content is doing something.

The Nightmare

The bad dream is common and the nightmare, the one that wakes you, less so: most adults have one or more a year, and between two and six in a hundred have one every week (Levin and Nielsen, 2007). Levin and Nielsen’s model holds that nightmares are the ordinary fear-extinction work of dreaming failing under load, in a person disposed to distress, and they are a core symptom of post-traumatic stress. They can be treated without drugs: in imagery rehearsal therapy the dreamer writes the nightmare down, rewrites its ending while awake, and rehearses the new version for a few minutes a day, and Krakow et al. (2001), in a controlled trial of 168 women with assault-related post-traumatic stress, found nightmares and the stress symptoms both fell by that alone. The nightmare that sits on the chest and cannot be moved from is a different thing, sleep paralysis, and has its own reading in the Clinical Readings.

Lucid Dreaming

That one can know one is dreaming, while dreaming, was reported by Aristotle and named by van Eeden in 1913, and dismissed by most laboratories until Keith Hearne (1978) and then Stephen LaBerge (LaBerge et al., 1981) had sleepers signal from inside REM sleep with an agreed pattern of eye movements, which the recording caught while every other muscle was paralysed. Voss et al. (2009) found the lucid dream a hybrid state, with the fast frontal activity of waking laid over the REM pattern, and in 2014 induced lucidity in sleepers by stimulating the frontal scalp at the same frequency. About half of adults have had a lucid dream at least once and a quarter have them monthly (Saunders et al., 2016); the techniques that produce them, reality-testing by day and waking in the small hours to set an intention, work modestly and reliably (Baird, Mota-Rolim, and Dresler, 2019). The lucid dreamer is, for now, the only witness who can report from the inside while the instruments watch from outside.

Jung and the Laboratory

The Symbols of the Psyche are drawn from analysts who held that the dream means something, and the laboratory has returned a mixed verdict on them. Jung’s central claim was compensation: that the dream presents what the waking attitude leaves out (Jung, 1948). Hall tested it in the 1950s and Domhoff after him, and the content studies found continuity instead, the dream carrying on the day’s concerns, not its opposite (Hall, 1953; Domhoff, 1996). Freud’s wish-fulfilment and disguise have fared no better, and the recurrence of the same few themes in every population has not required a collective unconscious to explain it.

What the analysts got right, on the same evidence, is that the dream is personal and continuous with the life, that it is emotional before it is anything else, that it takes recent experience and makes images of it, and that a dream recorded and read over months shows patterns the dreamer did not know they had, which is the premise of the Journal. The coding there is Hall and Van de Castle’s, not Jung’s; the catalogue of symbols is offered as what the traditions have said, to be read against the figures and not instead of them.

Keeping a Journal the Science Can Use

Record the dream before anything else, in the present tense, without interpretation; the memory decays within minutes, and rises again with the habit. Record every dream, the dull ones included, since the norms are percentages and the selection of the memorable ones distorts them. Code it the same day: the characters, the setting, the feelings, what happened. A hundred dreams give figures worth setting beside the norms; twenty give a sketch (Domhoff, 1996). Note what the day held, because the continuity is what you are looking for, and read the series back by the month rather than the dream by itself.