Activation of a h.o.m.ologous area in the left hemisphere may produce lexical hallucinations-of letters, numbers, musical notation, sometimes words or pseudowords, or even sentences. One of Mavromatis's subjects put it this way: "When dozing or before going to sleep ... I appear to be reading a book. I see the print clearly and distinguish the words, but the words rarely seem to have any particular significance. The books I appear to be reading are never books with which I am familiar, but frequently deal with whatever subject I have been reading during the day."

(While hypnagogic images of faces and places are usually unrecognizable, there is a distinct category of hypnagogia which McKellar and Simpson call "perseverative": hallucinations or recurrent images of something one has been exposed to earlier in the day. If, for example, one has been driving all day, one may "see" a hedgerow or line of trees continually unfurling before one's closed eyes.)

Hypnagogic imagery may be faint or colorless, but it often has brilliant and highly saturated color. Ardis and McKellar, in a 1956 paper, cited a case in which the subject described "colors of the spectrum intensified as though bathed in the fiercest sunlight." They compared this, as others have, to the exaggeration of color with mescaline. In hypnagogic hallucinations, luminosity or outlines may also seem to be abnormally distinct, with shadows or furrows exaggerated-sometimes such exaggerations go with cartoonlike figures or scenes. Many people speak of an "impossible" clarity or a "microscopic" detail in their hypnagogic visions. Images may seem finer-grained than perception itself, as if the inner eye has an acuity of 20/5 rather than 20/20 (this hyperacuity is a feature common to many types of visual hallucination).

One may "see" a constellation of images in hypnagogia-a landscape in the middle, a face erupting in the upper left corner, a complex geometric pattern around the edge-all present simultaneously and all evolving or metamorphosing in their own ways, a sort of multifocal hallucination. Many people describe hallucinatory polyopia, multiplications of objects or figures (one of McKellar's subjects saw a pink c.o.c.katoo, then hundreds of pink c.o.c.katoos talking to each other).

Figures or objects may suddenly zoom towards one, getting larger and more detailed, then retreat. Hypnagogic images, often compared to snapshots or slides, flash into consciousness, hold for a second or two, then disappear; they may be replaced by other images that seem to have no connection or apparent a.s.sociation to one another.

Hypnagogic visions may seem like something from "another world"-this phrase is used again and again by people describing their visions. Edgar Allan Poe stressed the fact that his own hypnagogic images were not only unfamiliar but unlike anything he had ever seen before; they had "the absoluteness of novelty."2

Most hypnagogic images are not like true hallucinations: they are not felt as real, and they are not projected into external s.p.a.ce. And yet they have many of the special features of hallucinations-they are involuntary, uncontrollable, autonomous; they may have preternatural colors and detail and undergo rapid and bizarre transformations unlike those of normal mental imagery.

There is something about the rapid and spontaneous transformations specific to hypnagogic imagery that suggests the brain is "idling," as my correspondent Mr. Utter suggested. Neuroscientists now tend to speak of "default networks" in the brain, which generate their own images. Perhaps one may also venture the term "play" and think of the visual cortex playing with every permutation, playing with no goal, no focus, no meanings-a random activity or perhaps an activity with so many microdeterminants that no pattern is ever repeated. Few phenomena give such a sense of the brain's creativity and computational power as the almost infinitely varied, ever-changing torrent of patterns and forms which may be seen in hypnagogic states.

Although Mavromatis writes of hypnagogia as "the unique state of consciousness between wakefulness and sleep," he sees affinities with other states of consciousness-those of dreams, meditation, trance, and creativity-as well as the altered modes of consciousness in schizophrenia, hysteria, and some drug-induced states. Although hypnagogic hallucinations are sensory (and thus cortical, being produced by the visual cortex, auditory cortex, etc.), he feels that the initiating processes may be in the more primitive, subcortical parts of the brain, and this, too, is something that hypnagogia may share with dreams.

And yet the two are quite distinct. Dreams come in episodes, not flashes; they have a continuity, a coherence, a narrative, a theme. One is a partic.i.p.ant or a partic.i.p.ant-observer in one's dreams, whereas with hypnagogia, one is merely a spectator. Dreams call on one's wishes and fears, and they often replay experiences from the previous day or two, a.s.sisting in the consolidation of memory. They sometimes seem to suggest the solution to a problem; they have a strongly personal quality and are determined mostly from above-they are largely "top-down" creations (although, as Allan Hobson argues, with a wealth of supporting evidence, they also employ "bottom-up" processes). In contrast, hypnagogic imagery or hallucination, with its largely sensory qualities-enhanced or exaggerated color and detail and outlines, luminosity, distortions, multiplications, and zoomings-and its detachment from personal experience, is overwhelmingly a "bottom-up" process. (But this is a simplification, for given the two-way traffic at every level in the nervous system, most processes are both top-down and bottom-up.) Hypnagogia and dreaming are both extraordinary states of consciousness, as different from each other as they are different from waking consciousness.

Hypnopompic hallucinations-those that may come upon waking-are often profoundly different in character from hypnagogic hallucinations.3 Hypnagogic hallucinations, seen with closed eyes or in darkness, proceed quietly and fleetingly in their own imaginative s.p.a.ce and are not usually felt to be physically present in one's room. Hypnopompic hallucinations are often seen with open eyes, in bright illumination; they are frequently projected into external s.p.a.ce and seem to be totally solid and real. They sometimes give amus.e.m.e.nt or pleasure, but they often cause distress or even terror, for they may seem charged with intentionality and ready to attack the just-wakened hallucinator. There is no such intentionality with hypnagogic hallucinations, which are experienced as spectacles unrelated to the hallucinator.

While hypnopompic hallucinations are only occasional with most people, they may occur frequently in some, as is the case with Donald Fish, an Australian man whom I met in Sydney after he wrote to me about his vivid hallucinations:

I wake from a calm sleep and perhaps a fairly normal dream with a shock, and there before me is a creature that even Hollywood couldn't create. The hallucinations fade in about ten seconds, and I can move when I have them. In fact I usually jump about a foot into the air and scream.... The hallucinations are becoming worse-now about four a night-I am now becoming terrified of going to bed. The following are some examples of what I see:

A huge figure of an angel standing over me next to a figure of death in black.

A rotting corpse lying next to me.

A huge crocodile at my throat.

A dead baby on the floor covered in blood.

Hideous faces laughing at me.

Giant spiders-very frequent.

Huge hand over my face. Also one on the floor five feet across.

Drifting spider webs.

Birds and insects flying into my face.

Two faces looking at me from under a rock.

Image of myself-only looking older-standing by the bed in a suit.

Two rats eating a potato.

A ma.s.s of different colored flags descending onto me.

Ugly-looking primitive man lying on floor covered in tufts of orange hair.

Shards of gla.s.s falling on me.

Two wire lobster pots.

Dots of red, increasing to thousands like spattered blood.

Ma.s.ses of logs falling on me.

It is often said that hypnagogic and hypnopompic hallucinations are more vivid and most easily remembered in childhood, but Mr. Fish's hallucinations have been lifelong-they started when he was eight, and he is now over eighty. Why he should be so p.r.o.ne to hypnopompic hallucinations is a mystery. Although he has had thousands of hypnopompic hallucinations, he has been able to live a full life and function consistently at a high creative level. A graphic designer and visual artist with a brilliant imagination, he sometimes finds inspiration in his surreal hallucinations.

While Mr. Fish's hypnopompic imagery is extreme in its frequency (and very distressing to him), it is not atypical in character. Elyn S. wrote to me about her own hypnopompic images:

The most typical one would involve me sitting up in bed and seeing a person-often an old lady-staring at me at some distance from the foot of my bed. (I imagine that such hallucinations are thought to be ghosts by some people-but not by me.) Other examples are seeing a foot-wide spider crawling up my wall; seeing fireworks; and seeing a little devil at the foot of my bed riding a bicycle in place.

A powerfully persuasive form of hallucination, not explicitly sensory at all, is the feeling of the "presence" of someone or something nearby, a presence that may be felt as malevolent or benign. The sense of conviction that someone is there can be irresistible at such times.

For me, hypnopompic experiences are usually more auditory than visual, and they take a variety of forms. Sometimes they are persistences of dreams or nightmares. On one occasion I heard a scratching sound in the corner of the room. I paid little attention at first, thinking it was just a mouse in the walls. But the scratching grew louder and louder and began to frighten me. Alarmed, I flung a pillow into the corner. But the action (or, rather, the imagined action) of flinging fully awakened me, and I opened my eyes to find that I was in my own bedroom, not the hospital-like room of my dream. But the scratching sound continued, loud and utterly "real," for several seconds after I woke.

I have had musical hallucinations (when taking chloral hydrate as a sleeping aid) which were continuations of dream music into the waking state-once with a Mozart quintet. My normal musical memory and imagery is not that strong-I am quite incapable of hearing every instrument in a quintet, let alone an orchestra-so the experience of hearing the Mozart, hearing every instrument, was a startling (and beautiful) one. Under more normal conditions I experience a hypnopompic state of heightened (and somewhat uncritical) musical sensibility; whatever music I hear in this state delights me. This happens almost every morning when I am awoken by my clock radio, which is tuned to a cla.s.sical station. (An artist friend describes a similar enhancement of color and texture when he lies in bed after first opening his eyes in the morning.)

Recently, I had a startling and rather moving visual hallucination. I cannot recollect what I was dreaming, if indeed I was dreaming, but when I awoke I saw my own face-or, rather, my face as it was when I was forty, black-bearded, smiling rather shyly. The face was about two feet away, life-sized, in faint, unsaturated pastel color, poised in midair; it seemed to look at me with curiosity and affection, and then, after about five seconds, it faded out. It gave me an odd, nostalgic sense of continuity with my younger self. As I lay in bed, I wondered whether, when young, I had ever had a vision of my present, almost eighty-year-old face, a hypnopompic "h.e.l.lo" across forty years.

While we may have the most fantastical and surreal experiences in our dreams, we accept these because we are enveloped in our dream consciousness, and there is no critical consciousness outside it (the rare phenomenon of lucid dreaming is an exception). When we awake we can remember only fragments, a tiny fraction of our dreams, and can easily dismiss these as "just a dream."

Hallucinations, in contrast, are startling and apt to be remembered in great detail-this is one of the central contrasts between sleep-related hallucinations and dreaming. My colleague Dr. D. has had only one hypnopompic hallucination in his life, and it occurred thirty years ago. But he retains the most vivid memory of it:

It was a relaxed summer night. I awoke around 2 AM, as I sometimes do in the middle of the night, and next to me, standing six and a half feet tall, was an imposing Native American. A huge man, muscles chiseled, black hair and black eyes. I realized, seemingly simultaneously, that if he wanted to kill me there was nothing I could do, and that he must not be real. Yet he was standing there, like a statue yet very alive. My mind flashed about-how could he have gotten into the house?... Why was he motionless?... This can't be real. Yet his presence frightened me. He became diaphanous after five or ten seconds, gently vaporizing into invisibility.4

Given the outlandish quality of some hypnopompic images, their often terrifying emotional resonance, and perhaps the heightened suggestibility that may go with such states, it is very understandable that hypnopompic visions of angels and devils may engender not only wonder or horror but belief in their physical reality. Indeed, one must wonder to what degree the very idea of monsters, ghostly spirits, or phantoms originated with such hallucinations. One can easily imagine that, coupled with a personal or cultural disposition to believe in a disembodied, spiritual realm, these hallucinations, though they have a real physiological basis, might reinforce a belief in the supernatural.

The term "hypnopompic" was introduced in 1901 by F. W. H. Myers, an English poet and cla.s.sicist who was fascinated by the emerging study of psychology. He was a friend of William James's and a founding member of the Society for Psychical Research, where he sought to connect the abnormal and paranormal with normal psychological function. Myers's work was highly influential.

Living in the late nineteenth century, a time in which seances and mediums were all the rage, Myers wrote extensively of ghosts, apparitions, and phantoms. Like many of his contemporaries, he believed in the idea of life after death, but he tried to place it in a scientific context. Although he felt that experiences likely to be interpreted as supernatural visitations were especially apt to occur in hypnopompic states, he also believed in the objective reality of a spiritual or supernatural realm, to which the mind might be given brief access in various physiological states, such as dreaming, hypnopompic states, trance states, and certain forms of epilepsy. But at the same time, he thought that hypnopompic hallucinations might be fragments of dreams or nightmares persisting into the waking state-in effect, waking dreams.

Yet reading Myers's 1903 two-volume Human Personality and Its Survival After Bodily Death, as well as Phantasms of the Living, the compilation of case histories he and his colleagues (Gurney et al.) published in 1886, one feels that the majority of "psychical" or "paranormal" experiences described are, in fact, hallucinations-hallucinations arising in states of bereavement, social isolation, or sensory deprivation, and above all in drowsy or trancelike states.

My colleague Dr. B., a psychotherapist, related the following story to me, about a ten-year-old boy who woke one morning "to find a woman dressed in blue hovering at the foot of his bed, surrounded by radiant light":

She introduced herself as his "guardian angel," speaking in a soft, gentle voice. The child was terrified, and turned on the light beside his bed, expecting the image to disappear. The woman remained suspended in the air, however, and he ran from the room, awakening his parents.