One leading neuroscientist explains that we’re all capable of hallucinations—and that, to better understand the brain, these weird experiences are worth listening to.
The persistence of hallucinations causes great suffering for some. In “Hallucinations,” neuroscientist Oliver Sacks explores the many kinds of unreal experiences people have.
Hallucinations don’t belong wholly to the insane. Much more commonly, they are linked to sensory deprivation, intoxication, illness, or injury. In Hallucinations (Knopf, 2012), Oliver Sacks weaves together stories of his patients, of his own mind-altering experiences and of his lifelong fascination with hallucinations. The book examines perceived reality and illuminates what “seeing things” or “hearing things” tells us about the organization and structure of our brains, how “visions” have influenced every culture’s folklore and art, and why the potential for hallucination is present in us all. In this excerpt from chapter four, “Hearing Things,” Sacks dives into the fascinating realm of auditory hallucinations, from the malicious and psychotic to the puzzling constant private concert of “earworms”—non-existent music that just won’t go away.
In 1973 the journal Science published an article that caused an immediate furor. It was entitled "On Being Sane in Insane Places," and it described how, as an experiment, eight "pseudopatients" with no history of mental illness presented themselves at a variety of hospitals across the United States. Their single complaint was that they "heard voices." They told hospital staff that they could not really make out what the voices said but that they heard the words "empty," "hollow," and "thud." Apart from this fabrication, they behaved normally and recounted their own (normal) past experiences and medical histories. Nonetheless, all of them were diagnosed as schizophrenic (except one, who was diagnosed with "manic-depressive psychosis"), hospitalized for up to two months, and prescribed antipsychotic medications (which they did not swallow). Once admitted to the mental wards, they continued to speak and behave normally; they reported to the medical staff that their hallucinated voices had disappeared and that they felt fine. They even kept notes on their experiment, quite openly (this was registered in the nursing notes for one pseudopatient as "writing behavior"), but none of the pseudopatients were identified as such by the staff.
This experiment, designed by David Rosenhan, a Stanford psychologist (and himself a pseudopatient), emphasized, among other things, that the single symptom of "hearing voices" could suffice for an immediate, categorical diagnosis of schizophrenia even in the absence of any other symptoms or abnormalities of behavior. Psychiatry, and society in general, had been subverted by the almost axiomatic belief that "hearing voices" spelled madness and never occurred except in the context of severe mental disturbance.
This belief is a fairly recent one, as the careful and humane reservations of early researchers on schizophrenia made clear. But by the 1970s, antipsychotic drugs and tranquilizers had begun to replace other treatments, and careful history taking, looking at the whole life of the patient, had largely been replaced by the use of DSM criteria to make snap diagnoses.
Eugen Bleuler, who directed the huge Burghölzli asylum near Zurich from 1898 to 1927, paid close and sympathetic attention to the many hundreds of schizophrenic people under his care. He recognized that the "voices" his patients heard, however outlandish they might seem, were closely associated with their mental states and delusions. The voices, he wrote, embodied "all their strivings and fears ... their entire transformed relationship to the external world ... above all ... [to] the pathological or hostile powers" that beset them. He described these in vivid detail in his great 1911 monograph, Dementia Praecox; or, The Group of Schizophrenias:
The voices not only speak to the patient, but they pass electricity through the body, beat him, paralyse him, take his thoughts away. They are often hypostasized as people, or in other very bizarre ways. For example, a patient claims that a "voice" is perched above each of his ears. One voice is a little larger than the other but both are about the size of a walnut, and they consist of nothing but a large ugly mouth.
Threats or curses form the main and most common content of the "voices." Day and night they come from everywhere, from the walls, from above and below, from the cellar and the roof, from heaven and from hell, from near and far ... When the patient is eating, he hears a voice saying, "Each mouthful is stolen." If he drops something, he hears, "If only your foot had been chopped off."
The voices are often very contradictory. At one time they may be against the patient ... then they may contradict themselves ... The roles of pro and con are often taken over by voices of different people ... The voice of a daughter tells a patient: "He is going to be burned alive," while his mother's voice says, "He will not be burned." Besides their persecutors the patients often hear the voice of some protector.
The voices are often localized in the body ... A polyp may be the occasion for localizing the voices in the nose. An intestinal disturbance brings them into connection with the abdomen ... In cases of sexual complexes, the penis, the urine in the bladder, or the nose utter obscene words ... A really or imaginarily gravid patient will hear her child or children speaking inside her womb ... Inanimate objects may speak. The lemonade speaks, the patient's name is heard to be coming from a glass of milk. The furniture speaks to him.
Bleuler wrote, "Almost every schizophrenic who is hospitalized hears 'voices.'" But he emphasized that the reverse did not hold — that hearing voices did not necessarily denote schizophrenia. In the popular imagination, though, hallucinatory voices are almost synonymous with schizophrenia — a great misconception, for most people who do hear voices are not schizophrenic.
Many people report hearing voices which are not particularly directed at them, as Nancy C. wrote:
I hallucinate conversations on a regular basis, often as I am falling asleep at night. It seems to me that these conversations are real and are actually taking place between real people, at the very time I'm hearing them, but are occurring somewhere else. I hear couples arguing, all kinds of things. They are not voices I can identify, they are not people I know. I feel like I'm a radio, tuned into someone else's world. (Though always an American-English-speaking world.) I can't think of any way to regard these experiences except as hallucinations. I am never a participant; I am never addressed. I am just listening in.
"Hallucinations in the sane" were well recognized in the nineteenth century, and with the rise of neurology, people sought to understand more clearly what caused them. In England in the 1880s, the Society for Psychical Research was founded to collect and investigate reports of apparitions or hallucinations, especially those of the bereaved, and many eminent scientists — physicists as well as physiologists and psychologists — joined the society (William James was active in the American branch). Telepathy, clairvoyance, communication with the dead, and the nature of a spirit world became the subjects of systematic investigation.
These early researchers found that hallucinations were not uncommon in the general population. Their 1894 "International Census of Waking Hallucinations in the Sane" examined the occurrence and nature of hallucinations experienced by normal people in normal circumstances (they took care to exclude anyone with obvious medical or psychiatric problems). Seventeen thousand people were sent a single question:
Have you ever, when believing yourself to be completely awake, had a vivid impression of seeing or being touched by a living being or inanimate object, or of hearing a voice, which impression, as far as you could discover, was not due to an external physical cause?
More than 10 percent responded in the affirmative, and of those, more than a third heard voices. As John Watkins noted in his book Hearing Voices, hallucinated voices "having some kind of religious or supernatural content represented a small but significant minority of these reports." Most of the hallucinations, however, were of a more quotidian character.
Perhaps the commonest auditory hallucination is hearing one's own name spoken — either by a familiar voice or an anonymous one. Freud, writing in The Psychopathology of Everyday Life, remarked on this:
During the days when I was living alone in a foreign city — I was a young man at the time — I quite often heard my name suddenly called by an unmistakable and beloved voice; I then noted down the exact moment of the hallucination and made anxious enquiries of those at home about what had happened at that time. Nothing had happened.
The voices that are sometimes heard by people with schizophrenia tend to be accusing, threatening, jeering, or persecuting. By contrast, the voices hallucinated by the "normal" are often quite unremarkable, as Daniel Smith brings out in his book Muses, Madmen, and Prophets: Hearing Voices and the Borders of Sanity. Smith's own father and grandfather heard such voices, and they had very different reactions. His father started hearing voices at the age of thirteen, Smith writes:
These voices weren't elaborate, and they weren't disturbing in content. They issued simple commands. They instructed him, for instance, to move a glass from one side of the table to another or to use a particular subway turnstile. Yet in listening to them and obeying them his interior life became, by all reports, unendurable.
Smith's grandfather, by contrast, was nonchalant, even playful, in regard to his hallucinatory voices. He described how he tried to use them in betting at the racetrack. ("It didn't work, my mind was clouded with voices telling me that this horse could win or maybe this one is ready to win.") It was much more successful when he played cards with his friends. Neither the grandfather or the father had strong supernatural inclinations; nor did they have any significant mental illness. They just heard unremarkable voices concerned with everyday things — as do millions of others.
Smith's father and grandfather rarely spoke of their voices. They listened to them in secrecy and silence, perhaps feeling that admitting to hearing voices would be seen as an indication of madness or at least serious psychiatric turmoil. Yet many recent studies confirm that it is not that uncommon to hear voices and that the majority of those who do are not schizophrenic; they are more like Smith's father and grandfather.
It is clear that attitudes to hearing voices are critically important. One can be tortured by voices, as Daniel Smith's father was, or accepting and easygoing, like his grandfather. Behind these personal attitudes are the attitudes of society, attitudes which have differed profoundly in different times and places.
Hearing voices occurs in every culture and has often been accorded great importance — the gods of Greek myth often spoke to mortals, and the gods of the great monotheistic traditions, too. Voices have been significant in this regard, perhaps more so than visions, for voices, language, can convey an explicit message or command as images alone cannot.
Until the eighteenth century, voices — like visions — were ascribed to supernatural agencies: gods or demons, angels or djinns. No doubt there was sometimes an overlap between such voices and those of psychosis or hysteria, but for the most part, voices were not regarded as pathological; if they stayed inconspicuous and private, they were simply accepted as part of human nature, part of the way it was with some people.
Around the middle of the eighteenth century, a new secular philosophy started to gain ground with the philosophers and scientists of the Enlightenment, and hallucinatory visions and voices came to be seen as having a physiological basis in the overactivity of certain centers in the brain.
But the romantic idea of "inspiration" still held, too — the artist, especially the writer, was seen or saw himself as the transcriber, the amanuensis, of a Voice, and sometimes had to wait years (as Rilke did) for the Voice to speak.
Talking to oneself is basic to human beings, for we are a linguistic species; the great Russian psychologist Lev Vygotsky thought that "inner speech" was a prerequisite of all voluntary activity. I talk to myself, as many of us do, for much of the day — admonishing myself ("You fool! Where did you leave your glasses?"), encouraging myself ("You can do it!"), complaining ("Why is that car in my lane?"), and, more rarely, congratulating myself ("It's done!"). Those voices are not externalized; I would never mistake them for the voice of God, or anyone else.
But when I was in great danger once, trying to descend a mountain with a badly injured leg, I heard an inner voice that was wholly unlike my normal babble of inner speech. I had a great struggle crossing a stream with a buckled and dislocating knee. The effort left me stunned, motionless for a couple of minutes, and then a delicious languor came over me, and I thought to myself, Why not rest here? A nap maybe? This was immediately countered by a strong, clear, commanding voice, which said, "You can't rest here — you can't rest anywhere. You've got to go on. Find a pace you can keep up and go on steadily." This good voice, this Life voice, braced and resolved me. I stopped trembling and did not falter again.
Joe Simpson, climbing in the Andes, also had a catastrophic accident, falling off an ice ledge and ending up in a deep crevasse with a broken leg. He struggled to survive, as he recounted in Touching the Void — and a voice was crucial in encouraging and directing him:
There was silence, and snow, and a clear sky empty of life, and me, sitting there, taking it all in, accepting what I must try to achieve. There were no dark forces acting against me. A voice in my head told me that this was true, cutting through the jumble in my mind with its coldly rational sound.
It was as if there were two minds within me arguing the toss. The voice was clean and sharp and commanding. It was always right, and I listened to it when it spoke and acted on its decisions. The other mind rambled out a disconnected series of images, and memories and hopes, which I attended to in a daydream state as I set about obeying the orders of the voice. I had to get to the glacier ... The voice told me exactly how to go about it, and I obeyed while my other mind jumped abstractly from one idea to another ... The voice, and the watch, urged me into motion whenever the heat from the glacier halted me in a drowsy exhausted daze. It was three o'clock — only three and a half hours of daylight left. I kept moving but soon realized that I was making ponderously slow headway. It didn't seem to concern me that I was moving like a snail. So long as I obeyed the voice, then I would be all right.
Such voices may occur with anyone in situations of extreme threat or danger. Freud heard voices on two such occasions, as he mentioned in his book On Aphasia:
I remember having twice been in danger of my life, and each time the awareness of the danger occurred to me quite suddenly. On both occasions I felt "this was the end," and while otherwise my inner language proceeded with only indistinct sound images and slight lip movements, in these situations of danger I heard the words as if somebody was shouting them into my ear, and at the same time I saw them as if they were printed on a piece of paper floating in the air.
The threat to life may also come from within, and although we cannot know how many attempts at suicide have been prevented by a voice, I suspect this is not uncommon. My friend Liz, following the collapse of a love affair, found herself heartbroken and despondent. About to swallow a handful of sleeping tablets and wash them down with a tumbler of whiskey, she was startled to hear a voice say, "No. You don't want to do that," and then "Remember that what you are feeling now you will not be feeling later." The voice seemed to come from the outside; it was a man's voice, though whose she did not know. She said, faintly, "Who said that?" There was no answer, but a "granular" figure (as she put it) materialized in the chair opposite her — a young man in eighteenth-century dress who glimmered for a few seconds and then disappeared. A feeling of immense relief and joy came over her. Although Liz knew that the voice must have come from the deepest part of herself, she speaks of it, playfully, as her "guardian angel."
Various explanations have been offered for why people hear voices, and different ones may apply in different circumstances. It seems likely, for example, that the predominantly hostile or persecuting voices of psychosis have a very different basis from the hearing of one's own name called in an empty house; and that this again is different in origin from the voices which come in emergencies or desperate situations.
Auditory hallucinations may be associated with abnormal activation of the primary auditory cortex; this is a subject which needs much more investigation not only in those with psychosis but in the population at large — the vast majority of studies so far have examined only auditory hallucinations in psychiatric patients.
Some researchers have proposed that auditory hallucinations result from a failure to recognize internally generated speech as one's own (or perhaps it stems from a cross-activation with the auditory areas so that what most of us experience as our own thoughts becomes "voiced").
Perhaps there is some sort of physiological barrier or inhibition that normally prevents most of us from "hearing" such inner voices as external. Perhaps that barrier is somehow breached or undeveloped in those who do hear constant voices. Perhaps, however, one should invert the question — and ask why most of us do not hear voices. Julian Jaynes, in his influential 1976 book The Origin of Consciousness in the Breakdown of the Bicameral Mind, speculated that, not so long ago, all humans heard voices — generated internally, from the right hemisphere of the brain, but perceived (by the left hemisphere) as if external, and taken as direct communications from the gods. Sometime around 1000 B.C., Jaynes proposed, with the rise of modern consciousness, the voices became internalized and recognized as our own.
Others have proposed that auditory hallucinations may come from an abnormal attention to the subvocal stream which accompanies verbal thinking. It is clear that "hearing voices" and "auditory hallucinations" are terms that cover a variety of different phenomena.
While voices carry meaning — whether this is trivial or portentous — some auditory hallucinations consist of little more than odd noises. Probably the most common of these are classified as tinnitus, an almost nonstop hissing or ringing sound that often goes with hearing loss, and may be intolerably loud at times.
Hearing noises — hummings, mutterings, twitterings, rappings, rustlings, ringings, muffled voices — is commonly associated with hearing problems, and this may be aggravated by many factors, including delirium, dementia, toxins, or stress. When medical residents, for example, are on call for long periods, sleep deprivation may produce a variety of hallucinations involving any sensory modality. One young neurologist wrote to me that after being on call for more than thirty hours, he would hear the hospital's telemetry and ventilator alarms, and sometimes after arriving home he kept hallucinating the phone ringing.
Although musical phrases or songs may be heard along with voices or other noises, a great many people "hear" only music or musical phrases. Musical hallucinations may arise from a stroke, a tumor, an aneurysm, an infectious disease, a neurodegenerative process, or toxic or metabolic disturbances. Hallucinations in such situations usually disappear as soon as the provocative cause is treated or subsides.
Sometimes it is difficult to pinpoint a particular cause for musical hallucinations, but in the predominantly geriatric population I work with, by far the commonest cause of musical hallucination is hearing loss or deafness — and here the hallucinations may be stubbornly persistent, even if the hearing is improved by hearing aids or cochlear implants. Diane G. wrote to me:
I have had tinnitus as far back as I can remember. It is present almost 24/7 and is very high pitched. It sounds exactly like how cicadas sound when they come in droves back on Long Island in the summer. Sometime in the last year [I also became aware of] the music playing in my head. I kept hearing Bing Crosby, friends and orchestra singing "White Christmas" over and over. I thought it was coming from a radio playing in another room until I eliminated all possibilities of outside input. It went on for days, and I quickly discovered that I could not turn it off or vary the volume. But I could vary the lyrics, speed and harmonies with practice. Since that time I get the music almost daily, usually toward evenings and at times so loud that it interferes with my hearing conversations. The music is always melodies that I am familiar with such as hymns, favorites from years of piano playing and songs from early memories. They always have the lyrics …
To add to this cacophony, I now have started hearing a third level of sound at the same time that sounds like someone is listening to talk radio or TV in another room. I get a constant running of voices, male and female, complete with realistic pauses, inflections and increases and decreases in volume. I just can't understand their words.
Diane has had progressive hearing loss since childhood, and she is unusual in that she has hallucinations of both music and conversation.
There is a wide range in the quality of individual musical hallucinations — sometimes they are soft, sometimes disturbingly loud; sometimes simple, sometimes complex — but there are certain characteristics common to all of them. First and foremost, they are perceptual in quality and seem to emanate from an external source; in this way they are distinct from imagery (even "earworms," the often annoying, repetitious musical imagery that most of us are prone to from time to time). People with musical hallucinations will often search for an external cause — a radio, a neighbor's television, a band in the street — and only when they fail to find any such external source do they realize that the source must be in themselves. Thus they may liken it to a tape recorder or an iPod in the brain, something mechanical and autonomous, not a controllable, integral part of the self.
That there should be something like this in one's head arouses bewilderment and, not infrequently, fear — fear that one is going mad or that the phantom music may be a sign of a tumor, a stroke, or a dementia. Such fears often inhibit people from acknowledging that they have hallucinations; perhaps for this reason musical hallucinations have long been considered rare — but it is now realized that this is far from the case.
Musical hallucinations can intrude upon and even overwhelm perception; like tinnitus, they can be so loud as to make it impossible to hear someone speak (imagery never competes with perception in this way).
Musical hallucinations often appear suddenly, with no apparent trigger. Frequently, however, they follow a tinnitus or an external noise (like the drone of a plane engine or a lawn mower), the hearing of real music, or anything suggestive of a particular piece or style of music. Sometimes they are triggered by external associations, as with one patient of mine who, whenever she passed a French bakery, would hear the song "Alouette, gentille alouette."
Some people have musical hallucinations virtually nonstop, while others have them only intermittently. The hallucinated music is usually familiar (though not always liked; thus one of my patients hallucinated Nazi marching songs from his youth, which terrified him). It may be vocal or instrumental, classical or popular, but it is most often music heard in the patient's early years. Occasionally, patients may hear "meaningless phrases and patterns," as one of my correspondents, a gifted musician, put it.
Hallucinated music can be very detailed, so that every note in a piece, every instrument in an orchestra, is distinctly heard. Such detail and accuracy is often astonishing to the hallucinator, who may be scarcely able, normally, to hold a simple tune in his head, let alone an elaborate choral or instrumental composition. (Perhaps there is an analogy here to the extreme clarity and unusual detail which characterize many visual hallucinations.) Often a single theme, perhaps only a few bars, is hallucinated again and again, like a skipping record. One patient of mine heard part of "O Come, All Ye Faithful" nineteen and a half times in ten minutes (her husband timed this) and was tormented by never hearing the entire hymn. Hallucinatory music can wax slowly in intensity and then slowly wane, but it may also come on suddenly full blast in mid-bar and then stop with equal suddenness (like a switch turned on and off, patients often comment). Some patients may sing along with their musical hallucinations; others ignore them — it makes no difference. Musical hallucinations continue in their own way, irrespective of whether one attends to them or not. And they can continue, pursuing their own course, even if one is listening to or playing something else. Thus Gordon B., a violinist, sometimes hallucinated a piece of music while he was actually performing an entirely different piece at a concert.
Musical hallucinations tend to spread. A familiar tune, an old song, may start the process; this is likely to be joined, over a period of days or weeks, by another song, and then another, until a whole repertoire of hallucinatory music has been built up. And this repertoire itself tends to change — one tune will drop out, and another will replace it. One cannot voluntarily start or stop the hallucinations, though some people may be able, on occasion, to replace one piece of hallucinated music with another. Thus one man who said he had "an intracranial jukebox" found that he could switch at will from one "record" to another, provided there was some similarity of style or rhythm, though he could not turn on or turn off the "jukebox" as a whole.
Prolonged silence or auditory monotony may also cause auditory hallucinations; I have had patients report experiencing these while on meditation retreats or on a long sea voyage. Jessica K., a young woman with no hearing loss, wrote to me that her hallucinations come with auditory monotony:
In the presence of white noise such as running water or a central air conditioning system, I frequently hear music or voices. I hear it distinctly (and in the early days, often went searching for the radio that must have been left on in another room), but in the instance of music with lyrics or voices (which always sound like a talk radio program or something, not real conversation) I never hear it well enough to distinguish the words. I never hear these things unless they are "embedded," so to speak, in white noise, and only if there are not other competing sounds.
Musical hallucinations seem to be less common in children, but one boy I have seen, Michael, has had them since the age of five or six. His music is nonstop and overwhelming, and it often prevents him from focusing on anything else. Much more often, musical hallucinations are acquired at a later age — unlike hearing voices, which seems, in those who have it, to begin in early childhood and to last a lifetime.
Some people with persistent musical hallucinations find them tormenting, but most people accommodate and learn to live with the music forced on them, and a few even come to enjoy their internal music and may feel it as an enrichment of life.
Ivy L., a lively and articulate eighty-five-year-old, has had some visual hallucinations related to her macular degeneration, and some musical and auditory hallucinations stemming from her hearing impairment. Mrs. L. wrote to me:
In 2008 my doctor prescribed paroxetine for what she called depression and I called sadness. I had moved from St. Louis to Massachusetts after my husband died. A week after starting paroxetine, while watching the Olympics, I was surprised to hear languid music with the men's swim races. When I turned off the TV, the music continued and has been present virtually every waking minute since.
When the music began, a doctor gave me Zyprexa as a possible aid. That brought a visual hallucination of a murky, bubbling brown ceiling at night. A second prescription gave me hallucinations of lovely, transparent tropical plants growing in my bathroom. So I quit taking these prescriptions and the visual hallucinations ceased. The music continued.
I do not simply "recall" these songs. The music playing in the house is as loud and clear as any CD or concert. The volume increases in a large space such as a supermarket. The music has no singers or words. I have never heard "voices" but once heard my name called urgently, while I was dozing. There was a short time when I "heard" doorbells, phones, and alarm clocks ring although none were ringing. I no longer experience these. In addition to music, at times I hear katydids, sparrows, or the sound of a large truck idling at my right side.
During all these experiences, I am fully aware that they are not real. I continue to function, managing my accounts and finances, moving my residence, taking care of my household. I speak coherently while experiencing these aural and visual disturbances. My memory is quite accurate, except for the occasional misplaced paper.
I can "enter" a melody I think of or have one triggered by a phrase, but I cannot stop the aural hallucinations. So I cannot stop the "piano" in the coat closet, the "clarinet" in the living room ceiling, the endless "God Bless America"s, or waking up to "Good Night, Irene." But I manage.
PET and fMRI scanning have shown that musical hallucination, like actual musical perception, is associated with the activation of an extensive network involving many areas of the brain — auditory areas, motor cortex, visual areas, basal ganglia, cerebellum, hippocampi, and amygdala. (Music calls upon many more areas of the brain than any other activity — one reason why music therapy is useful for such a wide variety of conditions.) This musical network can be stimulated directly, on occasion, as by a focal epilepsy, a fever, or delirium, but what seems to occur in most cases of musical hallucinations is a release of activity in the musical network when normally operative inhibitions or constraints are weakened. The commonest cause of such a release is auditory deprivation or deafness. In this way, the musical hallucinations of the elderly deaf are analogous to the visual hallucinations of Charles Bonnet syndrome.
But although the musical hallucinations of deafness and the visual hallucinations of CBS may be akin physiologically, they have great differences phenomenologically, and these reflect the very different nature of our visual worlds and our musical worlds — differences evident in the ways we perceive, recollect, or imagine them. We are not given an already made, preassembled visual world; we have to construct our own visual world as best we can. This construction entails analysis and synthesis at many functional levels in the brain, starting with perception of lines and angles and orientation in the occipital cortex. At higher levels, in the inferotemporal cortex, the "elements" of visual perception are of a more complex sort, appropriate for the analysis and recognition of natural scenes, objects, animal and plant forms, letters, and faces. Complex visual hallucinations entail the putting together of such elements, an act of assemblage, and these assemblages are continually permuted, disassembled, and reassembled.
Musical hallucinations are quite different. With music, although there are separate functional systems for perceiving pitch, timbre, rhythm, etc., the musical networks of the brain work together, and pieces cannot be significantly altered in melodic contour or tempo or rhythm without losing their musical identity. We apprehend a piece of music as a whole. Whatever the initial processes of musical perception and memory may be, once a piece of music is known, it is retained not as an assemblage of individual elements but as a completed procedure or performance; music is performed by the mind/brain whenever it is recollected; and this is also so when it erupts spontaneously, whether as an earworm or as a hallucination.
Excerpted from Hallucinations by Oliver Sacks. Copyright © 2012 by Oliver Sacks. Excerpted by permission of Knopf, a division of Random House, Inc. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.