Ruth Jaffe, in a 1968 article, described one concentration camp survivor who had frequent attacks in which she relived her experience at the gates of Auschwitz, where she saw her sister led off into a group destined for death but could do nothing to save her, even though she tried to sacrifice herself instead. In her attacks, she saw people entering the gates of the camp and heard her sister's voice calling, "Katy, where are you? Why do you leave me?" Other survivors are haunted by olfactory flashbacks, suddenly smelling the gas ovens-a smell which, more than anything else, brings back the horror of the camps. Similarly, the smell of burning rubble lingered around the World Trade Center for months after 9/11-and continued as a hallucination to haunt some survivors even when the actual smell was gone.

There is a large body of literature on both acute stress reactions and delayed ones following natural disasters like tsunamis or earthquakes. (These occur in very young children too, though they may tend to reenact rather than hallucinate or reexperience the disaster.) But PTSD seems to have an even higher prevalence and greater severity following violence or disaster that is man-made; natural disasters, "acts of G.o.d," seem somehow easier to accept. This is the case with acute stress reactions, too: I see it often with my patients in hospital, who can show extraordinary courage and calmness in facing the most dreadful diseases but fly into a rage if a nurse is late with a bedpan or a medication. The amorality of nature is accepted, whether it takes the form of a monsoon, an elephant in musth, or a disease; but being subjected helplessly to the will of others is not, for human behavior always carries (or is felt to carry) a moral charge.

Following the First World War, some physicians felt that there must be an organic brain disturbance underlying what were then called war neuroses, which seemed unlike "normal" neuroses in many ways.6 The term "sh.e.l.l shock" was coined with the notion that the brains of these soldiers had been mechanically deranged by the repeated concussion of the new high-explosive sh.e.l.ls introduced in this war. There was as yet no formal recognition of the delayed effects of the severe trauma of soldiers who endured sh.e.l.ls and mustard gas for days on end, in muddy trenches that were filled with the rotting corpses of their comrades.7

Recent work by Bennet Omalu and others has shown that repeated concussion (even "mild" concussions that do not cause a loss of consciousness) can result in a chronic traumatic encephalopathy, causing memory and cognitive impairment; this may well exacerbate tendencies to depression, flashbacks, hallucination, and psychosis. Such chronic traumatic encephalopathy, along with the psychological trauma of war and injury, has been linked to the rising incidence of suicide among veterans.

That there may be biological as well as psychological determinants of PTSD would not have surprised Freud-and the treatment of these conditions may require medication as well as psychotherapy. In its worst forms, though, PTSD can be a nearly intractable disorder.

The concept of dissociation would seem crucial not only to understanding conditions like hysteria or multiple personality disorder but also to the understanding of post-traumatic syndromes. There may be an instant distancing or dissociation when a life-threatening situation occurs, as when a driver about to crash sees his car from a distance, almost like a spectacle in a theater, with a sense of being a spectator rather than a partic.i.p.ant. But the dissociations of PTSD are of a more radical kind, for the unbearable sights, sounds, smells, and emotions of the hideous experience get locked away in a separate, subterranean chamber of the mind.

Imagination is qualitatively different from hallucination. The visions of artists and scientists, the fantasies and daydreams we all have, are located in the imaginative s.p.a.ce of our own minds, our own private theaters. They do not normally appear in external s.p.a.ce, like the objects of perception. Something has to happen in the mind/brain for imagination to overleap its boundaries and be replaced by hallucination. Some dissociation or disconnection must occur, some breakdown of the mechanisms that normally allow us to recognize and take responsibility for our own thoughts and imaginings, to see them as ours and not as external in origin.

It is not clear, however, that such a dissociation can explain everything, for quite different sorts of memory may be involved. Chris Brewin and his colleagues have argued that there is a fundamental difference between the extraordinary flashback memories of PTSD and those of ordinary autobiographic memory and have provided much psychological evidence for such a difference. Brewin et al. see a radical distinction between autobiographic memories, which are verbally accessible, and flashback memories, which are not verbally or voluntarily accessible but may erupt automatically if there is any reference to the traumatic event or something (a sight, a smell, a sound) a.s.sociated with it. Autobiographic memories are not isolated-they are embedded in the context of an entire life, given a broad and deep context and perspective-and they can be revised in relation to different contexts and perspectives. This is not the case with traumatic memories. The survivors of trauma may be unable to achieve the detachment of retrospection or recollection; for them the traumatic events, in all their fearfulness and horror, all their sensorimotor vividness and concreteness, are sequestered. The events seem to be preserved in a different form of memory, isolated and unintegrated.

Given this isolation of traumatic memory, the thrust of psychotherapy must be to release the traumatic events into the light of full consciousness, to reintegrate them with autobiographic memory. This can be an exceedingly difficult and sometimes nearly impossible task.

The idea that different sorts of memory are involved gets strong support from the survivors of traumatic situations who do not get PTSD and are able to live full, unhaunted lives. One such person is my friend Ben Helfgott, who was incarcerated in a concentration camp between the ages of twelve and sixteen. Helfgott has always been able to talk fully and freely about his experiences during these years, about the killing of his parents and family and the many horrors of the camps. He can recall it all in conscious, autobiographic memory; it is an accepted, integrated part of his life. His experiences were not locked away as traumatic memories, but he knows the other side well-he has seen it in hundreds of others: "The ones who 'forget,' " he says, "they suffer later." Helfgott is one of the contributors to The Boys, a remarkable book by Martin Gilbert that relates the stories of hundreds of boys and girls who, like Helfgott, survived years in concentration camps but somehow emerged relatively undamaged and have never been subject to PTSD or hallucinations.

A deeply superst.i.tious and delusional atmosphere can also foster hallucinations arising from extreme emotional states, and these can affect entire communities. In his 1896 Lowell Lectures (collected as William James on Exceptional Mental States) James included lectures on "demoniacal possession" and witchcraft. We have very detailed descriptions of the hallucinations characteristic of both states-hallucinations which rose, at times, to epidemic proportions and were ascribed to the workings of the devil or his minions, but which we can now interpret as the effects of suggestion and even torture in societies where religion had taken on a fanatical character. In his book The Devils of Loudun, Aldous Huxley described the delusions of demonic possession that swept over the French village of Loudun in 1634, starting with a mother superior and all the nuns in an Ursuline convent. What began as Sister Jeanne's religious obsessions were magnified to a state of hallucination and hysteria, in part by the exorcists themselves, who, in effect, confirmed the entire community's fear of demons. Some of the exorcists were affected as well. Father Surin, who had been closeted for hundreds of hours with Sister Jeanne, was himself to be haunted by religious hallucinations of a terrifying nature. The madness consumed the entire village, just as it would later do in the infamous Salem witch trials.8

The conditions and pressures in Loudun or Salem may have been extraordinary, though witch-hunting and forced confession have hardly vanished from the world; they have simply taken other forms.

Severe stress accompanied by inner conflicts can readily induce in some people a splitting of consciousness, with varied sensory and motor symptoms, including hallucinations. (The old name for this condition was hysteria; it is now called conversion disorder.) This seemed to be the case with Anna O., the remarkable patient described by Freud and Breuer in their Studies on Hysteria. Anna had little outlet for her intellectual or s.e.xual energies and was strongly p.r.o.ne to daydreaming-she called it her "private theater"-even before her father's final illness and death pushed her into a splitting or dissociation of personality, an alternation between two states of consciousness. It was in her "trance" state (which Breuer and Freud called an "auto-hypnotic" state) that she had vivid and almost always frightening hallucinations. Most commonly she would see snakes, her own hair as snakes, or her father's face transformed into a death's-head. She retained no memory or consciousness of these hallucinations until she was again in a hypnotic trance, but this time induced by Breuer:

She used to hallucinate in the middle of a conversation, run off, start climbing up a tree, etc. If one caught hold of her, she would very quickly take up her interrupted sentence without knowing anything about what had happened in the interval. All these hallucinations, however, came up and were reported on in her hypnosis.

Anna's "trance" personality became more and more dominant as her illness progressed, and for long periods she would be oblivious or blind to the here and now, hallucinating herself as she was in the past. She was, at this point, living largely in a hallucinatory, almost delusional world, like the nuns of Loudun or the "witches" of Salem.

But unlike the witches, the nuns, or the tormented survivors of concentration camps and battles, Anna O. enjoyed an almost complete recovery from her symptoms, and went on to lead a full and productive life.

That Anna, who was unable to remember her hallucinations when "normal," could remember all of them when she was hypnotized, shows the similarity of her hypnotized state to her spontaneous trances.

Hypnotic suggestion, indeed, can be used to induce hallucinations.9 There is, of course, a world of difference between the long-lasting pathological state we call hysteria and the brief trance states which can be induced by a hypnotist (or by oneself). William James, in his lectures on exceptional mental states, referred to the trances of mediums who channel voices and images of the dead, and of scryers who see visions of the future in a crystal ball. Whether the voices and visions in these contexts were veridical was of less concern to James than the mental states which could produce them. Careful observation (he attended many seances) convinced him that mediums and crystal gazers were not usually conscious charlatans or liars in the ordinary sense; nor were they confabulators or phantasts. They were, he came to feel, in altered states of consciousness conducive to hallucinations-hallucinations whose content was shaped by the questions they were asked. These exceptional mental states, he thought, were achieved by self-hypnosis (no doubt facilitated by poorly lit and ambiguous surroundings and the eager expectations of their clients).

Such practices as meditation, spiritual exercises, and ecstatic drumming or dancing can also facilitate the achievement of trance states akin to that of hypnosis, with vivid hallucinations and profound physiological changes (for instance, a rigidity which allows the entire body to remain as stiff as a board while supported only at the head and feet). Meditative or contemplative techniques (often aided by sacred music, painting, or architecture) have been used in many religious traditions-sometimes to induce hallucinatory visions. Studies by Andrew Newberg and others have shown that long-term practice of meditation produces significant alterations in cerebral blood flow in parts of the brain related to attention, emotion, and some autonomic functions.

The commonest, the most sought, and (in many cultures and communities) the most "normal" of exceptional mental states is that of a spiritually attuned consciousness, in which the supernatural, the divine, is experienced as material and real. In her remarkable book When G.o.d Talks Back, the ethnologist T. M. Luhrmann provides a compelling examination of this phenomenon.

Luhrmann's earlier work, on people who practice magic in present-day Britain, involved entering their world very fully. "I did what anthropologists do," she writes. "I partic.i.p.ated in their world: I joined their groups. I read their books and novels. I practiced their techniques and performed in their rituals. For the most part, I found, the rituals depended on techniques of the imagination. You shut your eyes and saw with your mind's eye the story told by the leader of the group." She was intrigued to find that, after about a year of this practice, her own mental imagery became clearer, more detailed, and more solid; and her concentration states became "deeper and more sharply different from the everyday." One night she became immersed in a book about Arthurian Britain, "giving way," she writes, "to the story and allowing it to grip my feelings and to fill my mind." The next morning she woke up to a striking sight:

I saw six druids standing against the window, above the stirring London street below. I saw them, and they beckoned to me. I stared for a moment of stunned astonishment, and then I shot up out of bed, and they were gone. Had they been there in the flesh? I thought not. But my memory of the experience is very clear.... I remember that I saw them as clearly and distinctly and as external to me as I saw the notebook in which I recorded the moment. I remember it so clearly because it was so singular. Nothing like that had ever happened to me before.

Later, Luhrmann embarked on a study of evangelical religion. The very essence of divinity, of G.o.d, is immaterial. G.o.d cannot be seen, felt, or heard in the ordinary way. How, she wondered, in the face of this lack of evidence, does G.o.d become a real, intimate presence in the lives of so many evangelicals and other people of faith? Many evangelicals feel they have literally been touched by G.o.d, or heard his voice aloud; others speak of feeling his presence in a physical way, of knowing that he is there, walking beside them. The emphasis in evangelical Christianity, Luhrmann writes, is on prayer and other spiritual exercises as skills that must be learned and practiced. Such skills may come more easily to people who are p.r.o.ne to being completely engaged, fully absorbed, by their experiences, whether real or imaginary-the capacity, Luhrmann writes, "to focus in on the mind's object ... the mode of the novel reader and the music listener and the Sunday hiker, caught up in imagination or appreciation." Such a capacity for absorption, she feels, can be honed with practice, and this is part of what happens in prayer. Prayer techniques are often focused on attention to sensory detail:

[Congregants] practice seeing, hearing, smelling, and touching in the mind's eye. They give these imagined experiences the sensory vividness a.s.sociated with the memories of real events. What they are able to imagine becomes more real to them.

And one day the mind leaps from imagination to hallucination, and the congregant hears G.o.d, sees G.o.d.

These yearned-for voices and visions have the reality of perception. One of Luhrmann's subjects, Sarah, put it this way: "The images I see [in prayer] are very real and lucid. Different from just daydreaming. I mean, sometimes it's almost like a PowerPoint presentation." Over time, Luhrmann writes, Sarah's images "got richer and more complicated. They seemed to have sharper borders. They continued to get more complex and more distinct." Mental images become as clear and as real as the external world.

Sarah had many such experiences; some congregants might have only a single one-but even a single experience of G.o.d, imbued with the overwhelming force of actual perception, can be enough to sustain a lifetime of faith.

Even at a more modest level, all of us are susceptible to the power of suggestion, especially if it is combined with emotional arousal and ambiguous stimuli. The idea that a house is "haunted," though scoffed at by the rational mind, may nonetheless induce a watchful state of mind and even hallucination, as Leslie D. brought out in a letter to me:

Almost four years ago I started a job that is housed in one of the oldest residences in Hanover, PA. On my first day, I was told there was a resident ghost, the ghost of Mr. Gobrecht, who lived here many years ago and was a music teacher.... I suppose he died in the house. It would be almost impossible to adequately describe how much I do NOT believe in the supernatural! However, within days I started to feel something like a hand tugging on my pant leg while I sat at my desk, and once in a while a hand on my shoulder. Just a week ago we were discussing the ghost, and I felt (very p.r.o.nounced) fingers moving along my upper back, just behind my shoulder, distinct enough to make me jump. Power of suggestion, maybe?

Children not uncommonly have imaginary companions. Sometimes this may be a sort of ongoing, systematized daydreaming or storytelling, the creation of an imaginative and perhaps lonely child; in some cases it may have elements of hallucination-a hallucination that is benign and pleasant, as Hailey W. described to me:

Growing up without brothers or sisters, I created a few imaginary friends whom I played with frequently from approximately age three to six. The most memorable of these was a pair of identical twin girls named Kacey and Klacey. They were my age and size, and we would often do things together like play on the swings in the backyard or have tea parties. Kacey and Klacey also had a little sister named Milky. I had a strong image of them all in my mind's eye, and they seemed very real to me at the time. My parents were mostly amused by it, though they did question whether it was natural for my imaginary friends to be so detailed and plentiful. They recall me having long conversations at the table with "no one," and when asked, I would always say I was talking with Kacey and Klacey. Often when playing (with toys, or games) I would say I was playing with Kacey and Klacey or Milky. I would talk about them often as well, and for a period of time I remember being fixated on the idea of a seeing-eye dog, begging my mother to let me have one. Rather taken aback, my mother asked where I got the idea; I replied that Kacey and Klacey's mother was blind, and that I wanted a seeing-eye dog like hers. As an adult, I am still surprised when someone tells me that they never had imaginary friends growing up, as they were such an important-and enjoyable-part of my childhood.

And yet "imagination" may not be an adequate term here, for imaginary companions may seem intensely real, as no other products of fantasy or imagination do. Perhaps the difficulty of fitting our adult categories of "reality" and "imagination" to the thoughts and play of children is not surprising; for, if Piaget is right, children cannot consistently and confidently distinguish fantasy from reality, inner from outer worlds, until the age of seven or so. It is usually at this age, or a little later, that imaginary companions tend to disappear.

Children may also be more accepting of their hallucinations, having not yet learned that hallucinations are considered (in our culture) "abnormal." Tom W. wrote to me about his "intended" childhood hallucinations, hypnagogic visions he would bring on as entertainment from the ages of four to seven:

I used to entertain myself while falling asleep by hallucinating. I would lie in bed and stare up at the ceiling in the halflight.... I would stare at a fixed point, and by holding my eyes very still, the ceiling would neutralize and gradually become swarming pixels, which would become patterns: waves and grids and paisleys. Then, in the midst of that, figures would start to appear and interact. I remember quite a few-[and] I remember the exceptional visual clarity of them. Once the vision was present, I could look around at things the way you would a film.

There was another way I used to do this. There was a family portrait that hung at the foot of my bed, a cla.s.sical staged photo of my grandparents, cousins, an aunt and uncle, my parents, my brother, and me. Behind us was a huge privet hedge. Again, in the evening, I would gaze at the portrait. Very quickly, strange and delightfully silly things would start to happen: apples would grow out of the privet hedge, my cousins would begin to chatter and chase each other around the group. My grandmother's head would "pop off" and attach to her two calves, which would then start to dance about. Grim as that seems now, I found it hilarious then.

At the other end of life, there is a special sort of hallucination that may attend death or the antic.i.p.ation of death. Working in old-age homes and nursing homes, I have been struck and moved by how often patients who are lucid, sane, and fully conscious may have hallucinations when they feel that death is near.

When Rosalie-the very old blind lady I described in the chapter on Charles Bonnet syndrome-became ill and thought she was dying, she had visions of her mother and heard her mother's voice welcoming her into heaven. These hallucinations were completely different in character from her usual CBS hallucinations-they were multisensory, personal, addressed to her, and steeped in warmth and tenderness. Her CBS hallucinations, by contrast, had no apparent relation to her and aroused no emotion. I have known other patients (who did not have CBS or any other special condition facilitating hallucinations) to have similar deathbed hallucinations-sometimes the first and last hallucinations in their lives.

1. Many of H. G. Wells's short stories also involve guilt hallucinations. In "The Moth," a zoologist who feels himself responsible for the death of his lifelong rival is haunted and finally driven mad by a giant moth that no one else can see, a moth of a genus unknown to science; but in his lucid moments, he jokes that it is the ghost of his deceased rival.

d.i.c.kens, a haunted man himself, wrote five books on this theme, the best known of these being A Christmas Carol. And in Great Expectations, he provides a dramatic account of Pip's vision after his first, horrified encounter with Miss Havisham:

I thought it a strange thing then, and I thought it a stranger thing long afterwards. I turned my eyes-a little dimmed by looking up at the frosty light-towards a great wooden beam in a low nook of the building near me on my right hand, and I saw a figure hanging there by the neck. A figure all in yellow white, with but one shoe to the feet; and it hung so that I could see that the faded tr.i.m.m.i.n.gs of the dress were like earthy paper, and that the face was Miss Havisham's, with a movement going over the whole countenance as if she were trying to call me. In the terror of seeing the figure, and in the terror of being certain that it had not been there a moment before, I at first ran from it, and then ran towards it. And my terror was greatest of all when I found no figure there.