Mary sometimes imagined that God had a reason for keeping her in the hospital, but, she said, “I did not even want to go there, because the reasoning mind makes you feel very nervous and uncertain.” Her daily life became so narrow that she stopped noticing the weather. “They don’t remind you about the changing season—that spring is coming into summer or the winter is coming into spring,” she said. “You just pass the days as quickly as possible.”

Christine, who spoke to her mother every week, said that Mary never articulated her sorrow at the time. “I would have wanted to engage on the level of ‘I’m sorry you’re there, do you feel sad?’ And I’m sure she was having those emotions, but she wasn’t able to express them. It was always ‘I’m being attacked. I’m being held like a prisoner here.’ ” Christine felt that her mother was safer at the Bronx Psychiatric Center than she would be anywhere else, but she felt guilty for hoping that Mary would stay there for the rest of her life.

Mary was discharged in September, 2023, after a year. A week later, she collapsed in her bathroom and struggled to move. She was taken to a hospital in Brooklyn, where the doctors told her to stop taking antipsychotics, because they thought her condition may have been a side effect of the drugs. Then they discovered that she actually had lymphoma, a sometimes fatal form of cancer. She began seven cycles of a treatment that combined chemotherapy with rituximab, a medication that targets antibodies involved in the body’s immune response.

When Christine and Angie visited her at the hospital, Mary responded to their questions with one-word answers. Her face had a vacant expression. Christine and Angie thought she was dying. Mary did, too. She dreamed about being a child, playing with her sister and four brothers, in Kolkata, where she grew up. “I told myself, ‘This must be the end of it,’ ” Mary said.

Angie, who was now twenty-two and had recently graduated from Dartmouth, prepared in therapy for her mother’s death. She said, “I had multiple sessions where I was just crying about the fundamental things I wish she had given me, like ‘I wish she could have told me what was going on in her head,’ or ‘I wish she could have told me she was sorry for what she did.’ ”

By Christmas, two months after beginning chemotherapy, Mary was moving a little more freely, and she had begun to carry on conversations. Christine and Angie noticed that her personality seemed different: she was calm, outgoing, and polite, and she often expressed gratitude. Angie texted Christine that Mary seemed peaceful, as if maybe she was “getting some post-life or death clarity?”

Christine, who was twenty-nine and had settled in London, having moved there for a master’s degree in psychology, was struck by her mother’s ability to watch the news and absorb the information on its own terms. For years, the television had been a source of agitation; Mary had said that people were using her ideas and repeating her lines.

One day, when Christine was visiting the hospital, Mary asked for a phone. “I sort of teased her, ‘Now you’re asking for a phone?’ ” Christine said. “I wasn’t really making much of it, but then afterward I thought, Why has she asked for a phone? That’s quite unusual.” Mary already had a phone, but it was in storage because she’d told Christine that it contained spyware.

Angie gave her a flip phone and, to be safe, covered the camera with a piece of tape. “She seemed fine using it, which was odd,” Angie said.

In May, a month after Mary finished chemotherapy, Christine and Angie asked a psychiatrist at the hospital to examine her. Christine said, “The psychiatrist was, like, ‘Why have you called me here? I don’t understand. She has no symptoms.’ And we were, like, ‘Yeah, that’s the reason we’ve called you here.’ ”

Christine had the same feeling in her body that she’d had when her mother first became ill—the sense that something at Mary’s core had changed. She tried to get the doctors to grasp the scale of her mother’s recovery. By the summer, her cancer was in remission. She hadn’t taken antipsychotics for months, and yet “her psychotic symptoms are gone,” a doctor wrote. Christine told the doctors, “She had a twenty-year psychiatric history. Have you heard of this? Could any of her medications have caused this?” She spoke with a neurologist at the hospital, but he didn’t have an answer. Omid Heravi, one of Mary’s oncologists, didn’t understand what had happened, either. “Medicine is very specialized—we don’t get involved in other fields,” he said. He guessed only that one of the cancer drugs she’d been given had had collateral benefits. “In medicine, all side effects are not bad,” he offered.

When a person recovers from an illness, it is usually seen as the end of the story. But becoming sane also causes a kind of narrative collapse, a confrontation with a personal history that is no longer recognizable. Christine encouraged her mother’s friends and siblings, whom she’d been estranged from for years, to get back in touch. She wanted to restore her mother’s sense of connection, but, she said, “I also wanted them to be able to tell me—outside of my childhood memories—if this is the person she used to be.”

For years, Christine’s friends had known little about her mother beyond the details of her mental illness. “Suddenly, I was, like, ‘Hey, my mom’s better now. Would you like to call her up and talk to her?’ And that was a shocking concept for them,” she said. “I mean, there are a lot of people who wouldn’t necessarily be open to jumping on a call without planning, but my mom was now the type of person who is quite flexible and responsive and conversationally fluid.” Christine described her as a “psychological débutante.”

Angie, who was living in Queens and working at a firm where she analyzed data on sexual violence, felt skeptical that their mother’s transformation was real. Christine said, “I felt like, If Mom can disappear, then she can come back.” But Angie didn’t have memories of Mary from before the illness, and it felt to her like she was being asked to believe that her mother had become a new person. “I tend to choose security over the process of discovery,” Angie told me. “I didn’t have a curiosity that felt worth the emotional risks.” Angie had always felt that, on some level, her mother had “chosen her delusions” over her children. She didn’t want to experience that choice being made again.

Christine searched for medical papers that might explain her mother’s recovery and allow Angie to believe in it. She read about each of the medications that her mother had taken and concluded that the key drug may have been rituximab, the immunosuppressant. “I have a new working theory,” she texted Angie, in May, 2024. “Theoretically her chemo could have incidentally cured” her.

Christine found a handful of recent case studies that documented drastic psychiatric recoveries after people were treated with drugs that dampen immune activity. A 2017 study in Frontiers in Psychiatry described a woman with a twenty-five-year history of schizophrenia. She also had a skin disease, for which she was given drugs that reduced inflammation and suppressed her immune response. Her doctors noticed a pattern: when they treated her skin lesions, her psychosis went away. They hypothesized that the rash and the psychosis had been caused by a single autoimmune disorder, and were cured by the same drugs. Another paper in Frontiers in Psychiatry described a man with “treatment-resistant schizophrenia” who developed leukemia. After a bone-marrow transplant, which reconstituted his immune system, he startled his doctors by suddenly becoming sane. Eight years later, the authors wrote, “the patient is very well and there are no residual psychiatric symptoms.”

Christine also discovered a Washington Post article from 2023 about a woman named April, who had fallen into a catatonic state at the age of twenty-one and been diagnosed with schizophrenia. Sander Markx, a professor of psychiatry at Columbia, first encountered April at a psychiatric hospital on Long Island when he was a medical student; twenty years later, he was dismayed to find her at the same hospital, in the same condition. “She has not been outside for twenty years—out of sight,” he said, at a symposium at Weill Cornell’s medical school. He and his colleagues gave her an extensive workup and found that she had lupus, an autoimmune disorder that, in rare cases, can induce inflammation in the brain, causing symptoms that are indistinguishable from those of schizophrenia. After undergoing immunosuppressive therapy, including rituximab, April emerged from, essentially, a “twenty-five-year-long coma, and was able to tell us everything,” Markx said. “We don’t have a script for this. We don’t see patients coming back from this condition.”

April’s case helped give momentum to the founding, in 2023, of the Stavros Niarchos Foundation (S.N.F.) Center for Precision Psychiatry and Mental Health, at Columbia, which is working to uncover biologically distinct subtypes of illness that have been obscured by the broad categories in the DSM. Christine sent an e-mail to Markx, a co-director of the S.N.F. Center, with a brief time line of her mother’s life. “Her psychiatric symptoms disappeared and have yet to resurface months later,” she wrote. “But her current clinicians are stumped as to why it has happened.” When Markx didn’t respond, Christine, who was visiting New York, decided that she and Angie should go to Columbia to introduce themselves in person. Markx wasn’t in his office—he had just begun an ongoing medical leave—but they slipped a handwritten card in a pink envelope under his door and used inter-campus mail to send cards to the other directors of the center. They tried to think of this step, Angie said, as “the part in the documentary where the cameras go all shaky and you get the sense that someone is about to break a hole in the case.”

Emil Kraepelin, who developed psychiatry’s first modern diagnostic system, in the eighteen-nineties, defined the disease we now know as schizophrenia largely in terms of its hopelessness. The diagnosis allowed hospital administrators to separate patients with “periodic insanities” (like depression and bipolar disorder) from those who were believed to be incurable and belonged in asylums. Kraepelin hoped that schizophrenia would eventually reveal itself to be a disease like neurosyphilis, which was then responsible for a large portion of the cases of insanity in psychiatry wards. In 1913, scientists demonstrated that bacteria had infected the brains of these patients. “The diseases produced by syphilis are an object lesson,” Kraepelin wrote, four years later. “It is logical to assume that we shall succeed in uncovering the causes of many other types of insanity that can be prevented—perhaps even cured—though at present we have not the slightest clue.”

Psychiatry and neurology were originally one medical discipline, but gradually neurologists took responsibility for diseases like neurosyphilis and dementia, in which the pathology could be seen in an autopsied brain, and psychiatrists handled the illnesses that were left behind, their causes still a mystery. Schizophrenia, which affects roughly one per cent of the population, became the disorder through which psychiatry worked out its identity, in part because it seemed to embody the mystery and intractability of madness, presenting basic questions about what it means to have a self. “The history of modern psychiatry is, in fact, practically synonymous with the history of schizophrenia, the quintessential form of madness in our time,” the psychologist Louis Sass has written.

But psychiatrists struggled to pinpoint a single feature that unified the diagnosis. “The great question is what is this ‘something’ that underlies the symptoms,” Karl Jaspers wrote, in 1963. Three decades later, the psychiatrist Ian Brockington warned that the obsession with schizophrenia had stifled clinical curiosity. “Smaller, more homogeneous entities have been sucked in by the gravity of the big idea, and annihilated,” he wrote. For decades, scientists have been searching in vain for a biological marker that would confirm whether someone has schizophrenia. Last year, in a paper in Schizophrenia Research, seventeen international experts concluded that schizophrenia was defined by no single etiology, symptom, or biological mechanism. “It is prudent to wonder if the construct around which we are organizing this information is fundamentally flawed,” the authors wrote.