[Editor’s note: this article originally appeared in Raritan, a quarterly review, and is included in this issue of AOTC by kind permission of the publishers. A revised version is also included in the just-published Our most troubling madness: case studies in schizophrenia across cultures, edited by T.M. Luhrmann and Jocelyn Marrow.]

Katie was the age of my youngest sister. Perhaps because of that I was less emotionally guarded than I often am on the street. We sat giggling like schoolgirls in the open area of the women’s shelter. It must once have been an industrial warehouse, because the room was huge and airducts big enough to crawl through ran along the ceilings. By nightfall it would feel crowded and noisy. In the front room, thirty five women or more would be pulling out thin blue mattresses and arranging them on the floor. There would be another fifty women in the back room, the so-called “permanent” beds, permanent for six months. There the mattresses were out all the time, supported by boards on milk cartons full of shampoo, soap and other inexpensive things that nonetheless got stolen, but still the women who slept there had to find their clothes, change, chatter, argue, fight. Thin metal lockers lined the walls, and women would be banging the doors open and shut, struggling with locks, pulling out laundry and stuffing in their coats. A long line would snake out of the lavatory in the front room and around the mattresses, women waiting for the stalls, the sinks, the one shower scheduled throughout the evening in drill sergeant slots.

But in the afternoon, when I met Katie, the shelter was quiet, even charming. There were tables with doilies and potted plants. The plants, to my astonishment, were real, surviving on fluorescent light. The permanent beds were piled high with comforters and teddy bears. There was a little area in front with some sofas and a television, always on. Televisions were always on in this shelter. In the evening, in the back room, little portable televisions rented from the corner electronics store blared competing programs. That afternoon there was only an indistinct drone from the back. Katie and I sat by one of the tables with doilies.  She told me about the house she’d owned in Texas, how proud she was of it, and she took out pictures of her six kids. When I asked her how she’d come to the shelter, she said she’d been home in Texas and her husband came in and sat down and said, hon, I love you, took out a pistol, and blew his head off. He hadn’t been very nice to her in the decade or so they’d been together. He’d broken her nose once and when they went to the emergency room the doctors kept asking her questions but she didn’t say anything. They did methamphetamine together. They had done it, she said, for about five years straight. She thought that it wasn’t very good for the kids. The last baby—she’d had six – had died in childbirth. She’d drive the others to school in the morning, but on meth it was hard. You wouldn’t sleep for days on end and by the fourth or fifth day, driving wasn’t so easy. The kids sometimes got scared. Katie got scared too. She would lock herself in her daughter’s bedroom when her husband got into one of his rages. She’d be crying, her kids would be crying. Sometimes he’d be lovely for months, attentive and caring. Then he’d choke her, and leave handprints on her neck. He liked to choke, she said. It was confusing around the time he killed himself. She thought he was having lots of affairs, but then she wondered whether she was being paranoid. She found semen in his anus—how and why she didn’t say, but she was sure it was semen—so she thought some of the affairs were with men. She remembered that her husband had brought another women over to the house and that he was going to do something awful to her, something terrible, and then she saw that the woman had Katie’s mother’s face. And then Katie told me that her husband wasn’t really dead. He’d tried to hang himself three weeks before he shot himself. He’d stood on a bucket with his head in a noose and kicked the bucket away. Katie had found him and cut him down and saved him. She thought he was safe now, too, that he was just playing dead. She thought that someone was planning to hurt the kids—by that point, they were in foster care – and her husband was being clever and went away to protect them. Weird stuff was happening. Her face changed when she looked at it in the mirror. Sometimes her head became sunken in, deformed, like the head of an experimental monkey. She heard voices, she said. She would hear someone saying, I feel like chicken tonight, and then someone else would say, no, let’s have ham. The voices said that her mom cheated on her dad, that her mother was a whore. They said that they were going to kill her. I know, Katie said, that that’s not normal, and she shrugged.

When I got home in the evening all I wanted to do was to shower. When you walk out of the middle class to sit down in a homeless shelter, you want someone like Katie to be an innocent victim, sweet, helpless, deserving. You want there to be someone to blame and someone to save. That’s pathetically naïve, of course. The people who think like that are the newcomers. The effective staff are flinty and unsurprisable. But when I met Katie, I was naïve.

***

I am a social anthropologist, and the method of my field is ethnography, or participant observation, perhaps the most time-intensive research method in the social sciences. It was invented more or less by accident by an irascible Pole during the first world war. Bronislaw Malinowski had trained as a physicist when he came across Sir James George Frazer’s Golden Bough, an Edwardian hodgepodge of myth, folklore and wonderfully strange customs that enticed him into a new academic discipline still dominated by what we now call “armchair anthropology”—theories based on other people’s travelogues and colonial recollections. Malinowski went off to the London School of Economics as a student in 1910. When hostilities broke out in Europe four years later, he was on an island off the coast of Melanesia. Lonely and unable to return to England, he slowly entered native life. He learned the language, made friends, and—the legend goes—acquired a lover, and he stayed on the island for four years, a participant as well as an observer of local life. Few anthropologists take the immersion that far. At least, they rarely stay in a fieldsite for four years at a stretch. But Malinowski’s deep-in-the-bones knowledge set the tone of the discipline from that decade forward, and although much has changed, the basic principles of the method have remained intact: that the anthropologist should develop informal relationships in the local language over a long period of time (at least a year) and that the anthropologist should experience and participate in the lives lived by the local people. What the latter principle means, of course, has varied with the decade. The anthropologists Malinowksi taught may have learned the local language, but they also wore pith helmets and explorer’s gear. Not until the 1960s did anthropologists try to live as the locals did, like my own teacher in graduate school who went upriver on the Amazon. He wore a g-string and took ayahuasca, and his generation changed the scholarly discourse about why those natives thought the gods were real. The fact that (as an Englishman) he saw London double-decker buses and not the spirit of the jaguar when he first started taking the hallucinogen makes it no less an anthropological experience.

No anthropologist can really go native. No normal person goes to dinner and slips off to the washroom to write notes. And even when the eager anthropologist does go through with some initiation rite, how can even the best-intentioned outsider possibly infer from his or her own experience of dread and wonder what the moment feels like to those who don’t go home to write it up? I used to stew about these things when I began to work as an anthropologist on the street and returned home to my locked house and my private shower and my safe, stable marriage. I told myself that I didn’t sleep in the shelter because to stay there would take a bed from someone else, although the truth was that I was scared of the leering strangers and the thought of the shower made my toes curl.  There were places I didn’t go after dark and some places I never went at all. I gradually realized that Katie never went to those places either, that many women in the neighborhood moved in tight, careful circles. Katie literally walked in circles for hours, around and around the shelter on streets where she felt safe. But what really made her world feel opaque to an observer’s empathic gaze was psychosis.

Psychosis is the name we give to judgments and perceptions that seem so impaired to us as to be no longer within the bounds of normal reason. People who are psychotic speak furiously to the empty air, talk incoherently, or laugh when talking about something sad. Their faces and their words seem out of synch, disturbingly mismatched between what they seem to feel and what they say. They may fear that they are being followed by the government, tell you that they have radio transmitters implanted in their teeth, believe that they have been published in leading scientific journals. And of course, sometimes the people who say these things are right. The great stories that circulate in psychiatric residencies are about the crazy patient being right—the patient, for example, who claimed to be a famous astrophysicist and the resident who looked the man up in Science and found his paper. Psychosis is always a judgment. It is also a symptom of many different illnesses—schizophrenia, bipolar disorder, substance abuse, trauma—just as a sore throat is a symptom of many different diseases. Because of this, it is easy to cast doubt on our judgments altogether and to claim (in the tradition of Thomas Szasz and R.D. Laing) that we call people psychotic because they see the truth more nakedly than we do ourselves and we cannot bear their insight. Sometimes they do and we can’t. But many of those who are psychotic struggle in a phantasmagoria of blood and horror. They hear people scream at them and curse and jeer. Those voices seem real to them, more real than normal. At the same time they often don’t quite believe them. So the very evidence of their senses feels unreliable, the way you feel when your new glasses are fitted accidentally with the wrong prescription and wearing them, you see the world pitch sideways and you know it doesn’t and yet you cannot trust yourself to step off the curb. Psychosis is more terrifying, and those who experience it cannot return the glasses. The voices and images throb in their brains. Psychosis hurts, and although now I have seen it up close and personal, I know it is beyond my capacity to understand from the inside.

Nevertheless that was my job as an ethnographer. Malinowski thought that fieldwork was about accessing the inaccessible. He didn’t put it that way, of course. All the early anthropologists have a deadpan matter-of-factness in describing what was, in their era, the unimaginable.  When Evans-Pritchard recounted the way Sudanese Azande villagers organized their daily lives around the fear that their neighbors were actually flesh-eating witches, he casually explained that he himself had no difficulty organizing his life along similar principles and that after a while in the village he had done so routinely. For both those early anthropologists and more contemporary ones, fieldwork is a tool to bridge an overwhelming gap in the cultural divide. Yet the divide that is the traditional provenance of anthropology is the gap between middle class Euro-America and exotically different societies, such as witchcraft-fearing tribal Africa and caste-structured Hindu villages. I was trying to cross the gap with people who were in some sense part of my own culture but whose lives were shaped profoundly by an experience I could not learn to have. I didn’t have the illusion that I would really understand what it is like to live with psychosis. I was there because other techniques didn’t seem to be making sense of a basic puzzle: why do so many women like Katie fall into homelessness and seem to stay there by choice?

***

Back in the early 1960s, idealistic policy makers tried to do the right thing. In 1955, there were 339 psychiatric beds for every 100,000 Americans, and half of them held people diagnosed with schizophrenia, for months or years at a time, in institutions that became kin and community for those they held. When Kennedy proclaimed the Community Mental Health Act, in 1963—following the 1954 release of chlorpromazine, the first medication to treat psychotic symptoms successfully—the idea was to release patients from stagnation on the back wards of state psychiatric hospitals into the loving care of their community. We coined the word ‘deinstitutionalization’ to celebrate their liberation from the supposedly Goffmanesque settings that forced patients’ minds into institutional straitjackets. It was a time when people were more likely to believe that a psychiatric diagnosis was some kind of arrogant establishment mistake. Peter Schaffer had a Broadway hit with Equus, a play about a young boy who stabbed horses in the eye but—this was the play’s point–was really a misunderstood sensitive soul.

Forty years later, we have what we might as well call “reinstitutionalization,” although it is of a different character, given that these days there are only 22 psychiatric beds for every 100,000 Americans and about half the patients stay in them for five days or less. People who in earlier decades would have passed their lives in the back wards of cavernous state hospitals often spend their days in neighborhoods sociologists call  “service ghettoes.” I met Katie in a two or three block area in Chicago which probably has the densest concentration of persons with serious psychotic disorder in the entire state of Illinois. Within and around these blocks there are medical clinics, psychiatric clinics, housing services, social services, soup kitchens, drop in centers, and agencies funded variously by the city, the state, the federal government, different religious groups and a grab bag of charities. There are overnight shelters for single men, single women, women with infants, and families. There are halfway houses with trained staff who stay overnight and rambling single room occupancy buildings with many tenants and little oversight. There are so-called nursing homes and cheap hotels and the closest the city comes to flophouses. And still with all these services and thousands of subsidized beds, many people with psychotic disorder subsist on the margins of the neighborhood, sleeping in the park or in the shelter, eating at the soup kitchens, coming in periodically for medical care, and getting raped and beaten in the alleys.

By now we know that homelessness is commonplace in the lives of people with serious psychotic disorder. That is, not only is it the case that many of those who remain homeless for months and years are psychotic; it is also true that many of those who are psychotic end up becoming homeless. This is a shocking claim. But it is what the data tell us. In 1998 The American Journal of Psychiatry published a study that tracked patients after they first made contact at ten out of twelve Long Island, New York hospitals. In this study, one in six patients diagnosed with schizophrenia either had been homeless or would become homeless in the following two years. In 2005 the journal published another study that analyzed the records of all patients treated in the primary hospital cluster in San Diego. One in five clients diagnosed with schizophrenia was homeless at the time of contact. Both studies—by the nature of their measurement and method – undoubtedly underestimate the risk of periodic homelessness for those with schizophrenia or some other psychotic disorder. The San Diego study was a snapshot, a study of one moment in a patient’s life across a single year, and even then it excluded two thousand people with that diagnosis in locked psychiatric facilities or in jails. The New York study only looked at the first years of illness, usually the period before the patient’s exhausted family reaches the limits of its tolerance and throws the patient out. Meanwhile, these days we are as likely to jail than to hospitalize those who, because of their illness, disturb the public peace. At any one time in this country, there are four times as many people with serious mental illness behind bars than in hospitals.

How did that happen? How did it come to pass that in one of the richest nations in the world we care for our sickest, neediest citizens on the street? The most important reason for the failure of the community mental health movement was that the money to support it never materialized. Kennedy’s Act had guaranteed federal funding for the state for the first eight years, after which the states were meant to take on the costs of non-hospital care—subsidized housing, often with staff oversight. For the most part, the states underfunded the programs. Subsidized housing – often described variously as ‘halfway houses’ or ‘supported housing’—is far cheaper than inpatient hospitalization, but it costs real money, even in the bad part of town. From the beginning, in the 1960s and 1970s, there wasn’t enough housing. When the real estate boom of the 1980s led many landlords to convert their rental units into condos, the supply of cheap urban housing dwindled further still. These days the mental health system—at best a patchwork quilt of different institutions—is seriously strapped for cash. In Chicago, experts estimate that there are fewer than a tenth as many beds in supported housing than the city needs.

But there is another, more complicated reason why the community health care system fails, at least the community mental health system as we have created it in the aftermath of underfunding, in poor neighborhoods studded with services and supported housing. Many of the people who should be using the system often reject the help that it offers. Many people with schizophrenia end up on the street even when supported housing is available. In Chicago, the wait for non-disability related low-income housing is currently seven years. I know people who have been told that if they were willing to see a psychiatrist, they could get in two weeks housing as good as any they could get with a low income housing voucher. Yet many who are eligible repeatedly refuse offers of such housing, in many cases offered by decent, caring people. And they refuse many other offers: of medication or counseling or employment, not always consistently and not unambivalently but often, and for years at a time. People like Katie wander in nomadic squalor between the homeless shelter, supported housing, inpatient hospitalization and jail, a grim social cycle the anthropologist Kim Hopper calls “the institutional circuit.” Perhaps they do get housed—but then they become too disorganized to pay the rent, or they violate the curfew or end up in a fight. Eventually they land back on the street, evicted or by choice, living in the homeless shelter, ties broken with their families, hospitalized or jailed when their behavior gets out of hand, occasionally getting housed, then leaving or losing housing, and returning to the street again. The question is why.

***

In the beginning I settled in to what people who staff these neighborhood services call a drop in center. It was a large, cheerfully painted room in an old hotel that had been built around the First World War with hopes of glamour and now was home to an array of struggling social services. The psychiatric service on the third floor had changed its name three times in as many years, as one organization after another took it over in the hope that they would make ends meet. The drop-in center teetered on insolvency, edging along quarter to quarter on an unpredictable combination of donations and grants. It had been founded in the first flush of feminist enthusiasm as a haven for women down on their luck. In those early days it was open from morning until night. Now they could only afford the staff for four hours on weekday afternoons. Anyone was welcome, as long as they were female. Most of the women who came in were homeless, at least when they first arrived. You could get a hot meal most days, wash your clothes, take a shower, talk to staff about where to get help in the neighborhood. I’d go there in the early afternoon and sit down at one of the tables scattered across the room and strike up a conversation with the other people at the table. Sometimes a woman would glare at me, or brusquely tell me to mind my own business and to go away. But gradually people got used to me. At first, the women knew immediately that I wasn’t one of them. After a while, though, I seemed to pick up something of the aura of the place, and newcomers sometimes assumed that I was homeless too. I began to spend time in the shelters, meeting the women I saw at the drop-in center, watching television, passing time. I struck up relationships with staff at different agencies and sat over chicken soup with them in the local diner.

Most days I spent some time with Zaney. She was a white woman in her middle forties, well-spoken, clean, and neatly dressed. I was struck by that because about half of every month she slept on the El, Chicago’s elevated train. It is neither safe nor easy to fall asleep on the train, but it is warm. Zaney came to Chicago from Wisconsin in her late twenties when she began to be taunted by an angry but non-existent crowd. They shoved her on the street, they shouted “slut” and “whore” at her, and they banged on the walls when she tried to sleep. When she arrived in Chicago, the police picked her up and brought her to a hospital where she stayed for a few days, undoubtedly diagnosed with schizophrenia. She got a referral to a caseworker in a community mental health center, and she kept the appointment. The caseworker got her housing and, eventually, a monthly social security check, now around $600 a month. Both the housing and the check were available to her only because of her psychiatric diagnosis. Zaney stayed in the housing for about eight years, and then she lost it, either because she left or because she got evicted. Then she stayed in shelters for several years. She said that she was doing a routine chore at the shelter when someone rudely told her that she wasn’t doing it well, and of course, she says, she stuck up for herself and they threw her out. The shelter director remembers that Zaney left the shelter of her own accord.

Zaney wanted desperately not to be homeless. Two weeks a month she stayed at a fleabag hotel, for about $160 a week, but she couldn’t afford more than that. She came into the drop-in center everyday with the classified ads, looking for apartments and work. I knew she had been told repeatedly that she could get housing again if she were willing to see a psychiatrist. I’d heard the staff tell her. Anyway, as far as I could tell, everyone in the drop in center knew how you could get housed. Most women were homeless when they showed up at the drop in center, and they talked about housing volubly and frequently. They would tick off the ways to get housed on their fingers: you had to be “crazy,” “addicted” or you had to have a job. “I ain’t crazy and I don’t got a job,” one woman announced to me. “So I’m working on being addicted.” What she meant by this was that she was beginning to go to the meetings with case workers and to Alcoholics Anonymous-like group meetings which agencies usually required clients to attend for weeks before giving them an apartment of their own, usually a studio with a small refrigerator, a hotplate and a bathroom, one of many off a long corridor.

By far the largest amount of subsidized housing is reserved for people who could be diagnosed with serious psychotic illnesses, like schizophrenia or bipolar disorder. Even most of the housing associated by the women with addiction in fact depended on what psychiatrists call “dual diagnosis,” where someone is diagnosable not only with substance abuse but also serious psychiatric disorder. Depression, of course, and other psychiatric illnesses like post traumatic stress disorder can be crippling, and occasionally someone would obtain disability-related social security on the basis of those diagnoses—but it was rare. Typically, most psychiatric disability-related subsidies, including housing, depended upon a psychiatric diagnosis of psychotic disorder. And that was where Zaney balked. She was very clear that she is not “crazy,” as she puts it. I used to suggest to her that she just lie, that she “pretend” to hear voices, just to get a safer place to sleep. She always shook her head. “I’m not that kind of person,” she said.

If you listened to the clinicians and to other staff, you would conclude that someone like Zaney refuses to see a psychiatrist because her illness corrupts her ability to think, her capacity to have what psychiatrists call “insight.” That’s certainly the inference you would draw if you were the psychiatrist in a room with her for a fifteen minute appointment, taking in her unkempt hair and the discrete but telltale plastic bags she carries. There is truth to this. Some of what Zaney says about housing seems pretty irrational, like her complaint that one landlord evicted her because he didn’t like her birthdate. But most people aren’t psychotic all the time, in all dimensions of their lives. Much of the time Zaney was as coherent as I am.

In fact if she weren’t competent in some basic ways, she couldn’t survive. It takes moxie to make it on the street.  Zaney found herself homeless because she had no one who would give her shelter and no money with which to buy it. Someone—maybe a police officer, maybe staff in a hospital’s emergency room, maybe the person spooning food in the soupkitchen, the details have been lost in her fog—had handed her a list of shelters, maybe called ahead to secure a place, maybe even gave her a buspass to get there. She made her way on the bus alone, transferring from one line to another, dragging her stuff, numb with the newness. She arrived and stood on line. People are always standing in long, resigned lines on the street, waiting for doors to open. I used to find that moment when the shelter opened unbearable, squeezing in at the main door with a sudden press of women, women with bulky bags, women who hadn’t showered, women who were angry or boisterous or glazed with dull, dissociated stares.

Once you get a bed in a shelter, you need a nose for whom to trust, whom to avoid. When the doors open, everyone must be registered, new people are interviewed, people are rolling out mattresses and finding blankets, sheets and towels and lining up for showers. People are bumping into each other, looking for space, arranging their stuff. There’s often a blowup, someone enraged that there’s no space or no appointment, someone furious about some accidental insult. The staff shout her down or throw her out, sometimes with the help of the police. There’s a lot of warmth, too, sometimes. I liked hanging out in a corner, hugging the women I knew, joking with the staff.  They know this world well; they’ve often climbed out of it themselves. But the mental illness makes them uneasy. Once I was in the back office and the staff person, whose name was Jean, was on the phone about a woman who had been diagnosed with schizophrenia and should have been getting disability but wouldn’t sign her form because she’d changed her name when she got married and now she hated the guy. Jean was shouting into the phone, which was making this strange loud buzzing sound, and then a woman (of the many women who had knocked on the door at this point) put her head in and explained that she really needed help because people were trying to electrocute her with the fire alarm, and that “we” wanted to talk to you—and Jean said, “who is the we?” And the woman said, “Never mind that, the mafia is the one that started it,” and she went on to say that Jean needed to call the police and see how they rigged the place and she was going to have to get the fire department over and then the police because they were pumping electricity through the system against good people like her who were government people. Jean said, “okay, I’ll see what I can do,” but the woman continued as if she had not spoken. Eventually Jean pushed her out the door, because someone else came in, a woman who was not only psychotic, but deaf and mute, and upset because her locker didn’t shut properly. Then the woman who wouldn’t sign for her disability, who had a fibroid tumor which made her look pregnant, came in for a buspass. When Jean finally came back after getting her the buspass, I said, “This is really the way we treat serious mental illness in our society, isn’t it?” “Yes,” she said, and then she sighed, “And you know, I am not trained for that.”

Time behaves oddly in a shelter. There are stretched out patches of boredom, no way to get comfortable because everything’s a little damp and a little dirty and there’s no place to sit that’s yours. But time also has the compressed, intense quality of the unpredictable. The shock of arrival never fades. Getting used to a shelter is not like coming into a new school, sorting out the jocks and the nerds, and finding your niche. There are always new women, some newly released from prison or the psychiatric ward. Even the ones you know can explode, sometimes fueled by crack or psychosis, sometimes just from stress and noise. Even late in the evening, after the lights are out, it is not quiet. People turn and stretch in unfamiliar beds, lying next to people they do not know, tense about having cash or medication or even just their shampoo filched, clutching their most precious stuff under the pillow. People get up to pee during the night and trip over other people’s bags, reaching out to grab something to steady themselves. People talk and mutter out loud. One of the most startling features of shelter life is how god-awful difficult it is to sleep through the night. Most of the people you see on the street are exhausted.

Yet they have to be alert enough to find their way to soup kitchens and figure out where the social service agencies are and stand on line with people who are high or psychotic, whom, they worry, might harass them. They have to figure out where people deal and where they turn tricks and where to walk if they don’t want to do either. In the neighborhood around the shelters I would see men standing in small, predatory groups on the corners, sometimes calling out to women as they skirted paths around them. Most of the women spoke casually of men beating girlfriends and men raping women. Stories circulated about women who had been found dead in an alley. I never discovered whether the stories were true, but it was obvious that the women who told them were nervous. I was walking down the street with Katie once when she saw a guy she knew a few yards ahead. She saw that he had a beer, which I hadn’t noticed, and she grabbed my shoulder and pulled me across the street as if we were going somewhere else. By this point he was calling to her, saying hi, and she waved back cheerfully. Then she turned her face to me and lowered her voice, “I really don’t like to be with him when he’s drinking. Let’s go this way.”

To survive, someone like Zaney or Katie needs to learn all this fast. Neither had anyone to explain the cues or to protect her if she flubbed. Neither had any friends, or at least, regarded the people she talked with (like me) as friends. There are many reasons that few people develop friendships in a shelter. No one wants to be there. People in shelters say scathing, contemptuous things about each other and about people like themselves.  They sweep their arms out at the room and denounce the women sitting in it. “You can’t get away from the homeless here,” a woman spat at me one afternoon. “You just can’t get rid of them. You just trip over them when you walk out the door here.” Zaney and Katie need institutions because they offer free food and shelter. By contrast, peers are a threat. Once I asked Zaney to draw her social world, with her social relationships. She drew herself under a tree in the park. She drew the Buddhist temple where she sometimes went, and she drew the Department of Human Services. She didn’t draw anything else. That’s why, she said, it was important to be tough. “It’s like they say about men going to jail, even if they’re innocent, they gotta fight, and if they don’t stand up for themselves, the other guys will take advantage and get even rowdier and you can get hurt, so it’s better to try to stand up for yourself.” She called the other people on the street “cowboys.”

I could see that people on the street acted tough to each other. When I was with a woman in the park, and men from the neighborhood approached, she seemed to grow larger and belligerent, as if she were arching her back and stiffening her fur. “They don’t mess with me,” she would say when they left. In the face of uncertain danger from other human beings, people learn fast to signal to strangers that the strangers should let them be. They signal threat. They raise their voices, plant their feet and throw their shoulders back. Zaney seemed to have learned to threaten people who, willfully or not, intruded into what she defined as hers. She’d sit peaceably in the drop-in center, and something ordinary would happen – she was shoved in line, commented on as she walked by, elbowed aside in the bathroom. She’d flare, raising her voice, throwing her shoulders back and her chest forward, acting fierce to get the offender to back down.

After a while, after I’d been in the neighborhood long enough that people were sometimes confused about whether I was homeless, it dawned on me that refusing housing was the same kind of signal. I realized that Zaney, Katie and the other women shared a culture in which the refusal to accept housing was a meaningful social signal, rather than just another symptom of psychosis, and that its meaning was tied up with the toughness it took to survive, and with what it meant—given the toughness—to be “crazy.” On the street, people used the word “crazy” differently from the way I’d heard the term used by other psychiatric clients, even those struggling with schizophrenia and psychosis. Among people who were middle-class and ill, I’d heard the word “crazy” used with an ironic, grudging familiarity. “Yeh, that’s when I was crazy,” someone would say, telling a joke about how he thought he’d walk from San Diego to New York, and gave up when he got to the top of his first big hill. Politically active psychiatric clients—sometimes called “consumers” or, more angrily, “psychiatric survivors”—use the word “crazy” in a defiant way. They make political buttons with the word. They adopt it and reclaim it. A self-mocking edge runs throughout the psychiatric consumer literature. One of the main magazines, sadly no longer published, was called simply “Dendron,” as if the reader were an aberrant neuron.

But nobody called themselves “crazy” in this neighborhood. I’d rarely seen a word used with such contempt. That caught my attention, my anthropological ear. And when I began to ask women what people mean by it, they were remarkably consistent. They would point to a woman who was flagrantly ill and talking to herself—at a shelter or a drop-in center, there was almost always at least one such woman present—and they’d say something derisive. “She’s crazy, she don’t need no friends.”

That didn’t tell me what the word meant: it told me who best represented the word, what you might call its prototype. To figure out its meaning, I did what Malinowski would have done, which was to listen to many women use the word spontaneously and from that to infer its meaning. One strand of meaning was clearly being “weak.” Katie made that clear back in the beginning, talking about the way her husband’s death had caught her by surprise. “I didn’t think anything was wrong with his head because he was a strong man. I just thought he was this strong man, that that wouldn’t ever happen to him, you know, he would never be crazy, he would never be actually crazy because he was a strong minded person, strong minded man, strong, so it wouldn’t happen to him. But I was wrong, because it did.” Another strand was a kind of permanence, something that—as one woman said—“would never be fixed.” Being crazy “is something you absolutely cannot control. And a lot of them don’t even take medication. They have retardation and there’s nothing you can do about it. Alcoholism you can do something about. You can stop drinking. Smoking, you can stop smoking. You can do those things and thereby reverse your situation, but someone who appears mentally ill can’t do that.” In fact the women repeatedly spoke about mental illness as a kind of retardation. As one woman put it, “Half of these people show up here, you know what I’m saying, half of them got a little problem. They don’t think that well.”  And yet another strain was the idea that the street would drive you crazy, and that if you weren’t careful, you might be next. “She’s been on the street too long,” women would say about someone else, twirling their fingers or rolling their eyes to show that the person that they were talking about was “crazy.” Or: “some people can’t handle the pressure …They break and become mentally ill.”

The structure of this “cultural model,” the phrase anthropologists sometimes use for these more-or-less shared cognitive schemas, seemed to be that flagrant psychosis arises when a woman is not strong enough to cope with the difficulties of homelessness; that the condition was permanent; and that only those who give up the struggle to get out become flagrantly ill. And then I did something that Malinowski would not have done, perhaps because his model of mind was more pragmatic, perhaps because he never doubted that anyone could doubt his judgment, perhaps because he spent so long with the bareskinned natives that he assumed that no one would ever challenge his authority. I hired a student, a young ethnographer, to go into the neighborhood and ask women at the drop in center what “women up here” meant by the term “crazy.” Some refused to answer and some gave conventional psychiatric definitions, but fully three quarters of them made comments that made it clear that they thought that to be crazy was the horrific punishment for those who weren’t strong enough to make it off the street intact and alive, and it terrified them.

It didn’t matter that I didn’t understand from the inside what it was like to be psychotic. It only mattered that I could feel the toxic mixture of rage, despair and terror it produces, in which the fear that you would never get out could grip you at the throat. Sleeping in a shelter is about looking into the eyes of someone mad in the next bed and fearing that if you weren’t careful enough, if you didn’t watch out, you too could slide into that strange and eerie world forever and be caught. You can feel the simmering violence in a shelter, the distrust, the bracing against someone’s hostile outburst. You can feel the hypervigilance, the scanning for trouble, the incipient anger at people who steal and shout, the exhaustion, the dull rage and despair at life itself. You see people who have lost it and you fear that you too live on a knife-edge, clinging to reality. The threat of psychosis always hovers on the street. The women talked about it like something so primal you can smell it. “They down and out and you don’t want to be like that. You go in there [the shelter] and right away you feel the aroma.”

Sometimes, people who weren’t psychotic tried to act that way in self-protection. “Act crazy, they’ll leave you alone,” one woman shrugged. But for those who fear that they might be psychotic, those who smell the weakness in their own skin, the emotion is very different. Psychosis is a continuum. Not everyone who hears caustic, demeaning voices ends up as a gibbering idiot. But some do. And that endpoint is grim. You cannot know what it feels like to be a woman incapable of normal human communication, but what you can see is that she is despised.  The most flagrantly psychotic women—the ones who are visibly talking to people no one else can see, who gesture to the empty air—are the most disliked women in the shelter. They are the ones who don’t get the social cues, who talk volubly in the middle of the night, who don’t respond in a way that makes sense. They are also the most vulnerable to violence. Once I arrived in the drop-in center to discover that a group of women had been standing on the street corner, pointing and jeering at a woman who was visibly mad. Sociological data tells us that such people are more likely to be beaten up than to be a danger to others, but you don’t need those statistics to see that the street is more dangerous for the women who were most dramatically ill. It is blatant.

Zaney knows that she’s different. She knows that other people don’t hear the crowd that taunts at her from the streets outside the drop in center, even though she hears them as audibly as she hears me speak when we sit together over sudoku in the drop-in center. She knows that when she sees her son in the shelter, when he torments her by showing up by her bedside and slipping away suddenly, so fast she cannot see him go, that he may not have really been there, even though she saw him as clear and solid as her pillow. She is afraid that she might be going crazy. Yet she is also afraid that her son might be there and in trouble, and if she does not search for him he will be kidnapped and he will die, and she is afraid that if she does not pay attention to the voices that threaten her, they will lead her into an alley and kill her. This is the terrible dilemma of madness, that if you ignore the phenomena—if you tell yourself that the voices and the visions are twisted figments of your imagination – and you are wrong, the cost is so very high, because the voices promise your own destruction. Those are the grounds, after all, on which Pascal became a Christian, or at least so he recorded – that if he believed and he was wrong, he risked being a fool, but if he did not believe and he was wrong, he risked eternal damnation. He chose belief. We live, all of us, in the grey zone of interpretation, judging what in our world is truly real.

That dilemma is more precarious, disbelief perhaps more risky, on the street than in the world of the middle class because it is interlaced with the rage, humiliation and fear of everyday life. Most people with psychosis go through many years of doubt even at the most definitely made diagnosis. To accept the diagnosis, they must be able to see themselves as having symptoms – being able to identify their nagging worry that the house is bugged as meaning something quite different than what their mind insists. But when Zaney walks down the street, she does see people massing, watching her, often engaged in criminal activity. Women do get raped on her block. People do taunt her. I’ve seen them do it. The danger the voices warn her of is real. And she knows that she is not like the wretched women who are flagrantly psychotic in the drop-in center and on the street corner, who are completely out of touch with the everyday world and easy victims for theft and assault. She knows her situation is better than theirs. They are crazy. She is strong.

Meaning matters. The anthropological insight here is that when women on the street reject housing if it comes on the condition that they must see a psychiatrist and get a diagnosis, the rejection is not necessarily a symptom of psychiatric illness. The refusal is a social gesture, a way of indicating something to other people. When Zaney says that she doesn’t want to see a psychiatrist because she is not crazy, it is a defiant assertion that she can survive in a dangerous world.

That knowledge tells us that we should do things differently. And indeed, there are programs in which people don’t really mention psychiatric diagnoses, don’t really demand that clients are aware of their diagnosis, don’t move them along like packages on a conveyer belt. They just give housing to people who are obviously eligible, and that housing is permanent. There are no tests, there are no required visits with mental health providers, there are no rules about what someone can or cannot do inside the housing. It works. Pathways to Housing, an experimental program in New York of this kind, concludes that more people stay in housing longer when it’s offered without strings or apparent diagnosis. They also conclude that it’s no more expensive than the other more conventional programs. That, eventually, was how Katie got housed, after years of sleeping out and sleeping in shelters. The shelter gave her housing she was able to interpret as associated with her history of drugs and alcohol, not her schizophrenia. They didn’t demand a formal certification, a structured interview with a psychiatrist who would confirm to the world that she was crazy. They talked to her, told her that she was eligible because of her substance abuse—this was true; it was dual diagnosis housing—and she moved in gratefully.

But it is deeply non-American to go this route. Americans don’t like handouts (or so they claim). Americans like to think that everyone is equal and has an equal right to goods, and they have stringent tests for those they need to label as “disabled.” Americans say they believe in fairness, and to demonstrate our fairness they have accountability and, as a result, a monstrous bureaucracy in which people are evaluated and certified. And Americans say they encourage pride and self-sufficiency in their citizenry, so that someone like Zaney can believe that when she fends for herself on the street that she is doing something strong and good.

***

In recent years the National Institutes of Health has begun to fund anthropologists to study the lives of people with serious mental illness. In some ways this is a surprising move. The NIH is a scientists’ institution, founded to pursue detached, unbiased research. Overwhelmingly NIH scientists are experimenters. They measure and they count. Anthropologists like me sit uneasily in such company. We are our own best instruments. We get to know our subjects personally. We hang around with them, talking about nothing in particular, arguing about sports teams, playing charades. We laugh with them. From this informal, casual, undirected interaction we try to understand the ambitions and anxieties created by the social world in which they live. We try to understand why people offered choices that look good to us might nevertheless find it reasonable to reject them.

What ethnography can deliver, and statistics and experiments for the most part cannot, is meaning—the difference, as the anthropologist Clifford Geertz once said, between a blink and a wink, between random movement and action motivated by the complicated, often unarticulated assumptions and interpretations with which we make sense of our world. For Malinowski, the point of the ethnographic method is that people are rarely able to tell you directly all that guides their actions, any more than most people can explain the grammatical rules that govern their speech. Similarly, Malinowski argued, most people (certainly, in his view, “native” people) had an enormously difficult time telling an outsider the basic assumptions that guided their lives because they simply didn’t see them as assumptions. Those assumptions are what anthropologists after Malinowski came to call culture: the ideas people take for granted as the natural order of the world, ideas that seem so basic that they no longer look like choices but like the terra firma of a life. To get at culture, Malinowksi argued, informal interactions were far more helpful than abstract, explicit questions, because it is in casual conversation that people talk about the things that matter to them—not the things the naïve interviewer might think to be important. “By means of this natural discourse,” he wrote in his magisterial Argonauts of the Western Pacific, “you will learn to know him, and become familiar with his customs and his life, far better than when he is a paid and often bored informant.” It was the ethnographer’s job to listen to this everyday conversation, and infer from what people said and what they did an “anatomy” of the society, the implicit assumptions which guide its participants: “Nowhere in human mind or memory are those laws to be found definitely formulated.”

Malinowski was right about that. Culture is notoriously difficult to define, even for anthropologists, even when the concept is the dead center of their field. The term broadens or shrinks depending on the decade and the anthropological school. Yet at the heart of any concept of culture is the recognition that people draw inferences about each other’s intentions, motivations and probable action based on what they know about each other as members of various communities—as men, as women, as soccer fans, as children, as people whose habits of mind and body are formed through interaction with others. Those inferences rely on implicit models about the lives of those people, arising from some combination of actual experience and our own predispositions, which help us to identify what we should pay attention to and how we should respond. These models are inherently schematic (psychologists call them “schemas”), often not conscious, and adaptable to novel experience—though some are more rigid than others. To borrow a famous example from the Berkeley linguist Charles Filmore, the English verb “to write” invokes a writer, a surface on which the writing occurs, an instrument which leaves a trace, and language. But a great deal remains unspecified. A plane can write on the surface of the sky; a child can write with scribbles if she does so in imitation of an adult. And what comes to my mind when I think of a “writer”—maybe a lady novelist in a Cotswald cottage—may be quite different from the image that comes to yours.

These days anthropologists argue fiercely about how reliably even trained observers can identify real patterns in the way people think and talk and feel. The debates aren’t so different from the arguments about the validity of psychotherapy, and for much the same reason. Good anthropologists are like good psychodynamic clinicians. They listen carefully to the ambling flows of many conversations, they watch what people do and what choices they make, and then they infer meaning: the partially inaccessible cognitive models which people use to draw inferences and which guide their emotional experiences. Anthropologists pick out the assumptions they think that people make, the implicit structures, specific to particular social worlds, that people use to organize their thoughts, structures so basic that those people often cannot see them for themselves. They identify emotions which they think people feel and the way people deal with those feelings. This is what the psychodynamic clinician is also after: the landscape of meaning through which each of us picks our way. Like the judgment of a good clinician, the anthropologist’s judgment—her ability to hear under the surface of speech and action—involves art and intuition. Some of that can be taught, with examples and practice and years of reading. Some can’t, and just as there are bad therapists, there are anthropology students who go into the field and hear nothing. Sometimes, like the worst of psychoanalytic theorizing—like the worst in any scholarly or literary walk of life – anthropologists’ accounts are blustery egoistic blunderings. Yet just as good clinicians can help you see a problem in your life with pinpoint clarity, good anthropologists help puzzled outsiders to understand a group of people who feel alien and other, who make choices and decisions which outsiders cannot process.

And in this case, the anthropological insights are important, because culture affects the way we treat mental illness, and our treatment affects the course and outcome of the illness. With schizophrenia, our treatment outcomes are pretty poor. The kicker to this anthropological story is that the women are probably right. The street may well drive you crazy.

Back in the mid-twentieth century when psychoanalysis dominated American psychiatry and Prozac hadn’t yet been invented, the dominant American understanding of schizophrenia was that psychosis was the response to unbearable emotional conflict. In what is surely the worst moment of American psychiatry, mothers were often held responsible. The “schizophrenogenic mother” was intensely ambivalent, terrified of love yet insistent that she was loving. In the exemplary passage of the famous article on the “double bind,” a young man on an inpatient unit stretches out his arm to hug his mother when she visits. She flinches, and when he withdraws she says, don’t you love me? The willingness of relatives to pay for hospital care was thought to arise from the guilt they felt for their role in the patient’s suffering. A classic 1954 study of one of the best psychoanalytic hospitals comments, “In some cases it would be reasonably adequate to describe the ideal relative as a person who appeared, gave the history precisely, accurately and directly, and disappeared forever, except for paying his bills—by mail.”

When psychiatry shifted to a more biomedical stance in the late 1970s and began to emphasize the organic dimension to mental illness, psychiatrists were appalled by what they had done to parents already struggling with the grief of losing a child to madness. Clinicians began to describe schizophrenia as random bad luck, a kind of genetic lightning bolt. They knew of course that the illness was not entirely genetic. If one of two identical twins can be diagnosed with schizophrenia, there’s only a fifty percent chance that the other twin will also fall ill. It was known that if a first degree relative had schizophrenia the chance that another might was greatly increased, just as if you go outside during a storm it increases your chance of being struck by lightning. But we think of lightning as being unpredictable bad luck, and that was the way most psychiatrists I met seemed to want patients to think about schizophrenia. Most people with schizophrenia, after all, do not have first degree relatives with schizophrenia and do not know their own genetic vulnerability. In speaking with people diagnosed with schizophrenia and with their parents, then, clinicians—earnestly trying to ward off feelings of blame and guilt on the part of the parents—emphasized the accidental and unexpected, the bad luck that the disorder should strike your family, your son.

Certainly there is good evidence for biological causation in schizophrenia. What is striking is that now there is epidemiological evidence, mostly from Europe, that there are specific paths for social causation as well. It’s been known for a long time that schizophrenia is associated with poverty, but until recently, most people thought that this meant that people who developed schizophrenia became poor because they couldn’t hold their jobs. But a recent study – tracking down father’s job and mother’s address from the birth certificate of the person with schizophrenia – demonstrated that if you are born poor, your risk for schizophrenia increases. If you live in an urban area, your risk for schizophrenia increases. And if you have a dark skin, your risk for schizophrenia goes up as your neighborhood whitens, a remarkable, disturbing finding called the “ethnic density” effect.

Most strikingly, when dark-skinned people emigrate to the United Kingdom or to the Netherlands (the only places where the studies have been done) their risk of schizophrenia rises sharply, an effect which has now been shown in so many papers by so many researchers with such methodological care that it cannot be explained away by clinician’s racial bias. Those who arrive in England from the Caribbean have around seven or more times the incidence of schizophrenia and of other psychotic disorders than whites, even adjusting for social class and age.

Meanwhile one of the most interesting puzzles in culture and mental health today is the difference in the outcome of schizophrenia in developing and developed countries. In an old WHO study, researchers had found, two years after an initial diagnosis of schizophrenia, that patients looked better in Africa and India than they did in sites scattered throughout the west. The study was redone and done more carefully, and the results still held. No matter whether you look at symptoms, disability, clinical profile, or the ability to do productive work, significantly more people do well after a diagnosis of schizophrenia in the developing world (really, in India) than they do in the developed world. So in some sense the causal account of schizophrenia has at long last circled back to the old psychoanalytic explanation. Much is different. The mother is no longer the villain. Now we blame the weak organism, not complex ideas about unconscious motivation and defense. But the fundamental insight seems right: that individuals are caught in webs of human relationship which can strangle the biologically vulnerable. To read this new epidemiology is to confront the social dimension of our bodily experience in a manner as arresting as when Freud first suggested that illness was intrapsychic and interpersonal.

Many people look at this data and wonder what India is doing right. And there are many factors that are specific to India. Most people with schizophrenia stay with their families. There is, compared to America, little homelessness. When patients find jobs, those jobs are often agricultural, without the demand and stress of American entry-level work, packing French fries to order behind a counter.

But when I look at these data, I don’t think about India. I think about Katie’s and Zaney’s neighborhood and a world in which people struggling with psychosis live in the nomadic trawl of the institutional circuit. Their world arose from the accidental happenstance of many good intentions: the social safety net which means that families trust their sick children to the care of the state, the community mental health care movement that released people locked behind hospital walls into the brooding violence of the street. Yes, in long term hospitalization people’s lives were structured by the rules and culture of the place. Back wards aren’t pretty. Those I’ve seen are soul-deadening places. The corridors echo. They reek of industrial soap. And yet they are safer and more comfortable than homeless shelters and sleeping under the trees in an urban park. When we released people with schizophrenia from state hospitals into the care of the community, we patted ourselves on the back and dropped the ball.  We have no statistical data yet to suggest that living on the street actually increases the risk of schizophrenia or worsens its outcome. Yet it is sheer common sense that it is bad for sick and vulnerable people to be abandoned to the noise, chaos and aggression of the street.  Zaney may be lucky. But the last time I saw her I thought that she would be dead or hospitalized within the month.