Feeling ill and unsteady, and with his memory occasionally playing tricks on him, Pinfold decides that a cruise to India might be restorative. His sleeping mixture runs out after two or three days, but his drinking stays at a high level. Barely has the ship got under way than he starts to have auditory hallucinations; most are of voices, but on occasion he hears music, a dog barking, the sound of a murderous beating administered by the captain of the ship and his doxy, and the sound of a huge ma.s.s of metal being thrown overboard. Visually, everything and everyone seems normal-a quiet ship with unremarkable crew and pa.s.sengers, steaming quietly past Gibraltar into the Mediterranean. But complex and sometimes preposterous delusions are engendered by his auditory hallucinations: he understands, for example, that Spain has claimed sovereignty over Gibraltar and will be taking possession of the vessel, and that his persecutors possess thought-reading and thought-broadcasting machines.

Some of the voices address him directly-tauntingly, hatefully, accusingly; they often suggest that he commit suicide-although there is a sweet voice, too (the sister of one of his tormentors, he understands), who says she is in love with him, and asks if he loves her. Pinfold says he must see her, as well as hear her, but she says that this is impossible, that it is "against the Rules." Pinfold's hallucinations are exclusively auditory, and he is not "allowed" to see the speaker-for this might shatter the delusion.

Such elaborate deliria and psychoses have a top-down as well as a bottom-up quality, like dreams. They are volcano-like eruptions from the "lower" levels in the brain-the sensory a.s.sociation cortex, hippocampal circuits, and the limbic system-but they are also shaped by the intellectual, emotional, and imaginative powers of the individual, and by the beliefs and style of the culture in which he is embedded.

A great many medical and neurological conditions, as well as all sorts of drugs (whether taken for therapeutic purposes or for recreation), can produce such temporary, "organic" psychoses. One patient who stays most vividly in my mind was a postencephalitic man, a man of much cultivation and charm, Seymour L. (I refer to him and his hallucinations briefly in Awakenings). When given a very modest dose of L-dopa for his parkinsonism, Seymour became pathologically excited and, in particular, started to hear voices. One day he came up to me. I was a kind man, he said, and he had been shocked to hear me say, "Take your hat and your coat, Seymour, go up to the roof of the hospital, and jump off."

I replied that I would not dream of saying anything like that to him, and that he must be hallucinating. "Did you see me?" I continued.

"No," Seymour answered, "I just heard you."

"If you hear the voice again," I said to him, "look round and see if I am there. If you cannot see me, you will know it is a hallucination." Seymour pondered this briefly, then shook his head.

"It won't work," he said.

The next day he again heard my voice telling him to take his hat and his coat, go up to the roof of the hospital, and jump off, but now the voice added, "And you don't need to turn round, because I am really here." Fortunately, Mr. L. was able to resist jumping, and when we stopped his L-dopa, the voices stopped, too. (Three years later, Seymour tried L-dopa again, and this time he responded beautifully, without a hint of delirium or psychosis.)

1. In addition to the overt delirium that may be a.s.sociated with life-threatening medical problems, it is not uncommon for people to have slight delirium, so mild that it would not occur to them to consult a physician, and which they themselves may disregard or forget. Gowers, in 1907, wrote that migraine is "often attended by quiet delirium of which nothing can be subsequently recalled."

There has always been inconsistency in defining delirium, and as Dimitrios Adamis and his colleagues pointed out in their review of the subject, it has frequently been confused with dementia and other conditions. Hippocrates, they wrote, "used about sixteen words to refer to and name the clinical syndrome which we now call delirium." There was additional confusion with the medicalization of insanity in the nineteenth century, as German Berrios has noted, so that insanity was referred to as delire chronique. Even now the terminology is ambiguous, so that delirium is sometimes called "toxic psychosis."

2. Just such an appearance of delirious images when closing the eyes, and their disappearance when the eyes are opened, is described by John Maynard Keynes in his memoir "Dr. Melchior":

By the time we were back in Paris, I was feeling extremely unwell and took to my bed two days later. High fever followed.... I lay in my suite in the Majestic, nearly delirious, and the image of the raised pattern on the nouveau art wall-paper so preyed on my sensibilities in the dark that it was a relief to switch on the light and, by perceiving the reality, to be relieved for a moment from the yet more hideous pressure of its imagined outlines.

3. In a prefatory note to a later edition, Waugh wrote: "Three years ago Mr. Waugh suffered a brief bout of hallucinations resembling what is here described.... Mr. Waugh does not deny that 'Mr. Pinfold' is largely based on himself." Thus we may accept The Ordeal as an autobiographic "case history" of a psychosis, an organic psychosis, albeit one written with a mastery of observation and description-and a sense of plot and suspense-that no purely medical case history has.

W. H. Auden once said that Waugh had "learned nothing" from his ordeal, but it at least enabled him to write a richly comic memoir, a new departure quite unlike anything he had written previously.

11

On the Threshold of Sleep

In 1992, I received a letter from Robert Utter, an Australian man who had heard me speak about migraine aura on television. He wrote, "You described how some migraine sufferers see elaborate patterns before their eyes ... and speculated that they might be a manifestation of some deep pattern-generating function in the brain." This reminded him of the experience that he routinely had upon going to bed:

This usually occurs at the moment when my head hits the pillow at night; my eyes close and ... I see imagery. I do not mean pictures; more usually they are patterns or textures, such as repeated shapes, or shadows of shapes, or an item from an image, such as gra.s.s from a landscape or wood grain, wavelets or raindrops ... transformed in the most extraordinary ways at a great speed. Shapes are replicated, multiplied, reversed in negative, etc. Color is added, tinted, subtracted. Textures are the most fascinating; gra.s.s becomes fur becomes hair follicles becomes waving, dancing lines of light, and a hundred other variations and all the subtle gradients between them that my words are too coa.r.s.e to describe.

These images and their subsequent changes appear and fade without my control. The experience is fugitive, sometimes lasting a few seconds, sometimes minutes. I cannot predict their appearance. They appear to take place not in my eye, but in some dimension of s.p.a.ce before me. The strength of the imagery varies from barely perceptible to vivid, like a dream image. But unlike dreams, there are absolutely no emotional overtones. Though they are fascinating, I do not feel moved by them.... The whole experience seems to be devoid of meaning.

He wondered whether this imagery represented a sort of "idling" in the visual part of the brain, in the absence of perception.

What Mr. Utter described so vividly are not dreams but involuntary images or quasi-hallucinations appearing just before sleep-hypnagogic hallucinations, to use the term coined by the French psychologist Alfred Maury in 1848. They are estimated to occur in a majority of people, at least occasionally, although they may be so subtle as to go unnoticed.

While Maury's original observations were all of his own imagery, Francis Galton provided one of the first systematic investigations of hypnagogic hallucinations, gathering information from a number of subjects. In his 1883 book Inquiries into Human Faculty, he observed that very few people might at first admit to having such imagery. It was only when he sent out questionnaires stressing the common and benign qualities of these hallucinations that some of his subjects felt free to speak about them.

Galton was struck by the fact that he, too, had hypnagogic hallucinations, even though it had taken time and patience for him to realize this. "Had I been asked, before I thought of carefully trying, I should have emphatically declared that my field of view in the dark was essentially of a uniform black, subject to an occasional light-purple cloudiness and other small variations," he wrote. Once he began observing more closely, however, he saw that

a kaleidoscopic change of patterns and forms is continually going on, but they are too fugitive and elaborate for me to draw with any approach to truth. I am astonished at their variety.... They disappear out of sight and memory the instant I begin to think about anything, and it is curious to me that they should often be so certainly present and yet be habitually overlooked.

Among the scores of people who responded to Galton's questionnaire was the Reverend George Henslow ("whose visions," Galton wrote, "are far more vivid than mine").1 One of Henslow's hallucinations started with a vision of a crossbow, then of an arrow, then a flight of arrows, which changed into falling stars and then into flakes of snow. This was followed by a finely detailed vision of a rectory and then of a bed of red tulips. There were quickly changing images in which he reported visual a.s.sociation (for instance, arrows became stars, then snowflakes) but no narrative continuity. Henslow's imagery was extremely vivid, but it had no quality of a dream or story.

Henslow emphasized how greatly these hallucinations differed from voluntary images; the latter were a.s.sembled slowly, bit by bit, like a painting, and seemed to be in the realm of everyday experience, while the former appeared spontaneously, unbidden and full-blown. His hypnagogic hallucinations were "very frequently of great beauty and highly brilliant. Cut gla.s.s (far more elaborate than I am conscious of ever having seen), highly chased gold and silver filigree ornaments; gold and silver flower-stands, etc.; elaborate colored patterns of carpets in brilliant tints."

While Galton singled out this description for its clarity and detail, Henslow was only one of many who described essentially similar visions when they were in a quiet, darkened room, ready for sleep. These visions varied in vividness, from faint imagery such as Galton himself had to virtual hallucination, though such hallucinations were never mistaken for reality.

Galton did not regard the disposition to hypnagogic visions as pathological; he thought that while a few people might experience them frequently and vividly whenever they went to sleep, most (if not all) people experienced them at least on occasion. It was a normal phenomenon, although special conditions-darkness or closing the eyes, a pa.s.sive state of mind, the imminence of sleep-were needed to bring it out.

Few other scientists paid much attention to hypnagogic visions until the 1950s, when Peter McKellar and his colleagues started what was to be a decades-long investigation of near-sleep hallucinations, making detailed observations of their content and prevalence in a large population (the student body at the University of Aberdeen) and comparing them with other forms of hallucination, especially those induced by mescaline. In the 1960s, they were able to complement their phenomenological observations with EEG studies as their subjects pa.s.sed from full wakefulness to a hypnagogic state.

More than half of McKellar's subjects reported hypnagogic imagery, and auditory hallucinations (of voices, bells, or animal or other noises) were just as common as visual ones. Many of my own correspondents also describe simple auditory hallucinations: dogs barking, telephones ringing, a name being called.

In his book Upstate, Edmund Wilson described a hypnagogic hallucination of a sort that many people share:

I seem to hear the telephone ringing just before I am completely awake in the morning. At first, I would go to answer it, but find that it was not ringing. Now I simply lie in bed, and if the sound is not repeated, I know that it is imaginary and don't get up.

Antonella B. hears music as she is falling asleep. The first time it happened, she wrote, "I heard a really nice cla.s.sical piece, played by a big orchestra, very complex and unknown." Usually, no images accompany her music, "just beautiful sounds that fill my brain up."

Susan F., a librarian, had more elaborate auditory hallucinations, as she wrote in a letter:

For several decades, just as I am drifting off to sleep, I have heard sentences uttered. They are always grammatically correct, usually in English, and usually spoken by a man. (On a few occasions they were spoken by a woman and once in a language I could not understand. I can recognize the differences between the Romance languages, Chinese, Korean, j.a.panese, Russian, and Polish, but it was none of these.) Sometimes the sentences are commands, such as "Go get me a gla.s.s of water," but at other times they are just statements or questions. During the summer of 1993, I kept a log of what I heard. Here are some of the sentences: "Once he was walking in front of me"; "This is yours, perhaps"; "Do you know what the photo looks like?"; "Mama wants some cookies"; "I smell the unicorn"; "Go get a shampoo."

What I hear bears no relationship to what I have read, seen, experienced or remembered on that day, previous day, week or year. Frequently when my husband is driving and we are on a long trip, I will nod off in the car. The sentences come very rapidly then. I will nod off for a second, hear a sentence in the twilight of waking, repeat the sentence to my husband, and then nod off again, hear another sentence in the twilight and so on, until I decide to wake up and stay awake.

In Speak, Memory, Nabokov provided an eloquent description of his own hypnagogic imagery, both auditory and visual:

As far back as I remember ... I have been subject to mild hallucinations.... Just before falling asleep, I often become aware of a kind of one-sided conversation going on in an adjacent section of my mind, quite independently from the actual trend of my thoughts. It is a neutral, detached, anonymous voice, which I catch saying words of no importance to me whatever-an English or a Russian sentence, not even addressed to me, and so trivial that I hardly dare give samples.... This silly phenomenon seems to be the auditory counterpart of certain praedormitary visions, which I also know well.... They come and go, without the drowsy observer's partic.i.p.ation, but are essentially different from dream pictures for he is still master of his senses. They are often grotesque. I am pestered by roguish profiles, by some coa.r.s.e-featured and florid dwarf with a swelling nostril or ear. At times, however, my photisms take on a rather soothing flou quality, and then I see-projected, as it were, upon the inside of the eyelid-gray figures walking between beehives, or small black parrots gradually vanishing among mountain snows, or a mauve remoteness melting beyond moving masts.

Faces are especially common in hypnagogic hallucinations, as Andreas Mavromatis emphasizes in his encyclopedic book Hypnagogia: The Unique State of Consciousness Between Wakefulness and Sleep. He cites one man who described this in 1886; the faces, he wrote,

seem to come up out of the darkness, as a mist, and rapidly develop into sharp delineation, a.s.suming roundness, vividness, and living reality. They fade off only to give place to others, which succeed with surprising rapidity and in enormous mult.i.tude. Formerly the faces were wonderfully ugly. They were human, but resembling animals, yet such animals as have no fellows in the creation, diabolical-looking.... Latterly the faces have become exquisitely beautiful. Forms and features of faultless perfection now succeed each other in infinite variety and number.

Many other descriptions stress how common it is to see faces, sometimes cl.u.s.ters of faces, with each face highly individuated but unrecognizable. F. E. Leaning, in her 1925 paper on hypnagogia, speculated that such an emphasis on faces "almost suggests that there is a special 'face-seeing' propensity in the mind." Leaning's "propensity," we now know, has its anatomical substrate in a specialized portion of the visual cortex, the fusiform face area. Dominic ffytche and his colleagues have shown in fMRI studies that it is precisely this area in the right hemisphere which is activated when faces are hallucinated.