Hallucinations, OliverSacks. New York, NY: Knopf; 2012. $26.95

New York, NY: Knopf; 2012. $26.95Oliver Sacks, now famous for his highly accessible neurological case studies, used to be known for monographs on migraine and encephalitis lethargica. Interestingly, his migraine work focused not on the headache but on the fascinating sensory hallucinations that herald the condition in many migraineurs. Sacks argued that “It is in this sense that migraine is enthralling; for it shows us, in the form of a hallucinatory display, not only an elemental activity of the cerebral cortex, but an entire self-organizing system, a universal behavior, at work. It shows us not only the secrets of neuronal organization, but the creative heart of Nature itself.”1 In his new book, Hallucinations, he returns to migraine illusions, comparing them with hallucinations seen via pharmaceuticals (including Sacks’ private experiments in the sixties), but those varieties of hallucinogenic experience are just two in a larger cast of hallucinations induced by sensory and neural disorders. A disclosure—years ago, during experiments that induced “forbidden” hues, I incidentally created illusory colored textures, and Sacks helpfully pointed out that similar pattern formation occurs in other hallucinatory systems including migraine. At first I thought Sacks’ comment odd. Many hallucinatory systems form simple illusory patterns like spirals, fan shapes, or concentric circles, all of which correspond to oriented stripes of neural activity on visual cortex or they form illusory lattices that correspond to a superposition of cortical stripes. These distinctive shapes are known as Klüver forms, after the scientist who studied them first with mescaline; similar effects are inducible by flickering lights, transcranial electric currents and magnetic fields, epileptic seizures, and even by binocular pressure on the eyeballs. But migraine is usually different; the archetypical migraine hallucination is the fortification illusion, a moving flickering serrated arc. The details of the serrations carry valuable information about the functional neuroanatomy of the visual cortex. It was Sacks, a migraineur himself, who noticed that the fortification illusions and Klüver percepts could coexist in the same observer and during the same attack.1 In Sacks’ hands, migraine is less of a special case and more a member of the family of neural pattern-forming systems. Sacks interweaves his own hallucinatory experiences with those of his patients and others, forming an easy and coherent narrative that links many conditions. The sequential discussions of Charles Bonnet syndrome (CBS), sensory deprivation, and hemianopia are especially interesting. Donald Hebb, the Canadian pioneer of theoretical neuroscience, asked his students to study techniques used by Stalinists to force confessions, and they focused on sensory deprivation. Hebb’s students were aware that nature sometimes provides sensory deprivation experiments in subjects who are losing their sight, leading to hallucinations. Their use of opaque goggles as an analog to ocular cataracts was fortuitous because later experiments showed that subjects hallucinate more readily to diffuse light than total darkness. Sensory deprivation is one of those subjects that attracted wild attention before slipping back into obscurity, but its clinical equivalent, CBS, has recently undergone a serious revival of interest, and Sacks does a nice job of bringing the subject back to its roots by telling the story of Bonnet and his uncle, who shared the details of his hallucinations with Bonnet, a budding scholar, while the uncle slowly lost his sight. Some years later, Bonnet himself would experience the same condition. Sacks, who lost some vision to a melanoma, has written off his own CBS hallucinations.2 Like many vision loss patients, these three hallucinated a variety of simple geometric shapes and more complex forms, especially faces. What made the three of them unusual was that they were willing to talk about it. The more usual reticence is sad on two counts: many CBS sufferers are elderly and worry that the hallucinations are manifestations of dementia, whereas many clinicians do not warn patients with deteriorating vision of these harmless perceptual side effects, ensuring that some of their patients suffer in silence. Intriguingly, Sacks mentions that some physicians are now using quetiapine as an off-label treatment for CBS hallucinations. This joins previous reports of clinicians using carbamazepine to prevent occipital epileptic hallucinations and diphenylhydantoin to treat LSD flashbacks. What the later two treatments have in common is that they regulate sodium channels in nerve membranes, reducing cortical excitability. This is in keeping with evidence of excess cortical excitability in subjects with migraine auras; with reports of LSD flashbacks induced in susceptible people using caffeine, nicotine, and ocular pressure and flicker; and with a growing theoretical literature on pattern formation in neural networks when excitation is increased in a spatially uniform fashion.3 Sacks is sometimes gently criticized for letting the demands of writing for popular audiences overwhelm scholarly aspects of his work (Carl Sagan syndrome). At first glance, it might seem so here; Hallucinations is almost too enjoyably written to be considered a monograph but has a remarkable density and variety of information. I have been involved in hallucination research for a dozen years, and I learned something new in every chapter. Still, there is no effort to be encyclopedic, and so it is interesting to consider Sacks’ choices on what to include and what to elide. In Sacks’ early work, there was a strong emphasis on elementary hallucinations, and Sacks collaborated on some of the first neural simulations of elementary hallucinations. Little of this pioneering work is discussed in the new book; a major focus instead is on complex hallucinations involving faces, music, illegible text, unplayable musical scores, and autoscopy (Doppelgänger). The chapters on the hallucinations accompanying narcolepsy, sleep paralysis, and other sleep disorders are chilling. The exploration of the relationships between hallucinations, filling-in phenomena, and phantom limb disorders is fascinating and contains remarkable observations. Recently, there has been much work on the neural correlates of elementary hallucinations (see Billock and Tsou3 for a review), and a fuller understanding is starting to emerge. So perhaps it is useful that Sacks puts most of his emphasis on the less understood complex hallucinations. But if I had been as prescient on elementary hallucinations as Sacks, I don’t think I could have resisted taking a victory lap. Sacks is made of sterner stuff, and a good reward for his fortitude would be a serious revival of scholarly interest in these fascinating phenomena. For vision researchers, both basic and clinical, I would nominate Sacks’ discussions of unreadable text for further attention. Other researchers have pointed out how hallucinations—especially hallucinations in blind or peripheral fields—seem unusually high in contrast and detail and can either sharpen unexpectedly with accommodation and eye movements or be abolished by them. Sacks agrees that many hallucinations seem crisp in contrast and implausibly fine in detail, and yet hallucinated letters are often unreadable, and when attention is fixed on them, they seem “rune-like.” Elsewhere, Sacks distinguishes between realistic-looking hallucinatory faces and grotesque “neurological faces” formed by activation of primitive face-detecting mechanisms. Rune-like letters may be a text analog of “neurological faces,” but I’m reminded of letter distortions in amblyopia and the prosaic perception of high-contrast but unreadable text in peripheral vision. Several lines of research on deblurring and on the paradoxical effects of reducing contrast in blur perception suggest a high spatial frequency extrapolation in spatial vision. Rune-like shapes are a plausible outcome of extrapolating the spatial frequency content of text, absent a rule for extrapolating the corresponding phase information. With a little image processing, transcribed unreadable hallucinatory text may be decodable. Whether anyone—even the patients—would want to read those hallucinatory ramblings is a separate question. Vincent A. Billock Dayton, Ohio

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.