The paper appeared in the journal Science on 19 January 1973. It was eight pages long. Its title was “On Being Sane in Insane Places,” and its author was a Stanford University psychology professor named David Rosenhan. Before it was published, Rosenhan was a moderately well-regarded academic with a reasonable list of publications and no particular reputation outside his field. Within a year of publication, he was one of the most famous psychologists in America. Within a decade, the paper he had written was among the most-cited pieces of psychological research ever produced. And within a generation, the American Psychiatric Association had entirely rewritten its diagnostic manual, thousands of state psychiatric hospitals had closed their doors, and hundreds of thousands of severely mentally ill Americans had been released onto streets they were not equipped to survive on.
The paper did all of that.
The story it told was extraordinary, if it was true.
What Rosenhan claimed
According to Rosenhan’s original 1973 paper as published in Science, volume 179, pages 250 to 258, he and seven other volunteers, none of whom had any history of mental illness, presented themselves at twelve different psychiatric hospitals across five American states between 1969 and 1972. Each was instructed to report to the admitting staff that they had been hearing voices. The voices, they were to say, were unclear but seemed to be repeating three specific words: “empty,” “hollow,” and “thud.” They were to give false names and, if they were psychologists themselves, false professions. Beyond that single symptom and the false identifying details, they were to answer every subsequent question truthfully.
All eight were admitted. Seven were diagnosed with schizophrenia. The eighth received a diagnosis of manic-depressive psychosis. Once admitted, they stopped reporting any symptoms and began behaving normally, taking notes on the wards, chatting with staff, and requesting release. According to Rosenhan’s account, none of the medical staff at any of the twelve hospitals ever noticed that anything was unusual. Fellow patients, on the other hand, spotted the pseudopatients as frauds with startling regularity. In one hospital alone, thirty-five of a hundred and eighteen patients on the admissions ward publicly voiced their suspicions to the pseudopatients themselves. “You’re not crazy,” one patient reportedly told a pseudopatient. “You’re a journalist, or a professor.”
The pseudopatients spent between seven and fifty-two days each on the wards. The average stay was nineteen days. All eight were eventually discharged, and every one of them was released with a diagnosis of schizophrenia, or of manic-depressive illness, in remission. Not one was discharged with the note that they had been misdiagnosed. Not one was flagged as sane.
This video goes into more detail on the psychology behind the experiment:
The conclusion Rosenhan drew was blunt. Psychiatric diagnosis, he wrote, was so unreliable that trained professionals in a clinical setting could not distinguish a mentally ill person from a healthy one. The label, once applied, followed the patient through every subsequent interaction on the ward, and every ordinary behaviour was reinterpreted through the lens of the diagnosis. Writing in a notebook was documented as “engaging in writing behaviour.” Pacing the corridor was noted as agitation. The pseudopatients described themselves as feeling powerless, invisible, and depersonalised. Rosenhan concluded that the psychiatric hospital itself was not a place of healing but a place of amplified suffering, in which the sane could not be distinguished from the insane and in which the diagnosis functioned as a self-fulfilling prophecy.
The paper landed on the American psychiatric profession with the force of a public accusation. Newspapers covered it extensively. Television programmes debated it. Psychiatry, which for most of the twentieth century had operated as a comparatively confident branch of medicine, entered a period of institutional soul-searching that led, over the following decade, to a fundamental rewriting of the profession’s diagnostic standards. The third edition of the Diagnostic and Statistical Manual of Mental Disorders, published in 1980 and now known simply as DSM-III, was in significant part an attempt to answer the challenge Rosenhan’s paper had laid down. And in the political climate of the 1970s and 1980s, the paper was cited in support of the movement to close down large state psychiatric hospitals and shift the care of the severely mentally ill toward community-based programmes that most jurisdictions never adequately funded.
All of it, on the primary-source record, rested on Rosenhan’s account of what had happened in those twelve hospitals.
What Cahalan found
Forty-six years after the paper was published, an American investigative journalist named Susannah Cahalan set out to write a book about it. Cahalan had a personal stake in the story. Ten years earlier, at the age of twenty-four, she had been misdiagnosed as schizophrenic in a New York hospital when what she actually had was an autoimmune inflammation of her brain. She had, on the accumulated evidence, been on the exact wrong side of Rosenhan’s central question. She had assumed, when she began her research, that Rosenhan would be a hero of hers.
Her book, The Great Pretender: The Undercover Mission That Changed Our Understanding of Madness, published in November 2019 by Grand Central Publishing, did not turn out to be the book she had expected to write. What Cahalan found, when she gained access to David Rosenhan’s personal papers after his death and began comparing them with the published paper, was a body of documentary evidence that appeared to be irreconcilable with the story Rosenhan had told.
She started with what she thought would be the easiest part of the reporting, which was to identify and interview the other seven pseudopatients. She could only find two. Rosenhan himself, who had been the first pseudopatient, was one. The other was a graduate student named Bill Underwood. Cahalan located a third possible pseudopatient, a psychologist named Harry Lando, whose hospital admission fit the pattern Rosenhan had described. When she interviewed him, Lando confirmed he had participated in the study. He also told her that his hospital experience had been overwhelmingly positive, that the staff had been attentive and kind, and that he had come away with a much better impression of psychiatric care than he had gone in with. Lando’s participation was omitted entirely from the published paper. Rosenhan had excluded him.
The other five pseudopatients Cahalan simply could not find. There was no correspondence with them in Rosenhan’s papers. There were no names, no identifying details, no records of hospital admissions matching the descriptions in the study. When Cahalan tried to trace them through hospital records, professional networks, and Rosenhan’s surviving colleagues, the trail went cold in every direction. On the accumulated evidence of her investigation, she concluded that five of the eight pseudopatients Rosenhan had described in Science had probably never existed.
Then Cahalan obtained Rosenhan’s own psychiatric admission records from Haverford State Hospital in Pennsylvania, where he had been the first pseudopatient in early 1969. The records did not match his published account. Rosenhan had written in Science that he had reported only the three-word auditory hallucination and had otherwise answered every question truthfully. His actual admission notes, on the surviving medical record, showed that he had described elaborate suicidal thoughts, claimed the voices he was hearing were coming to him through radio waves, and given the admitting psychiatrist a fictional history of severe mental disturbance stretching back years. Any competent psychiatrist confronted with that presentation would have hospitalised the patient. On Cahalan’s assessment, the presenting symptoms Rosenhan had actually described bore almost no resemblance to the mild, single-symptom presentation he had reported to Science.
In 2023, the eminent medical historian Andrew Scull, working from records Cahalan had shared with him, published a formal academic assessment in the peer-reviewed journal History of Psychiatry, titled “Rosenhan revisited: successful scientific fraud”. Scull concluded, on the documentary evidence, that Rosenhan’s paper was not simply flawed or exaggerated. It was, in Scull’s phrase, a spectacularly successful case of scientific fraud. The study that had reshaped an entire profession, on the accumulated evidence of half a century of subsequent scrutiny, appears to have been substantially made up.
What is left, then, is one of the strangest questions in the history of twentieth-century science.
If Rosenhan invented most of the pseudopatients, exaggerated his own presenting symptoms, and excluded the participant whose experience contradicted his conclusions, then what should be done about the profession that reformed itself in response to his findings? The DSM-III rewrite was, by any measure, a significant improvement over its predecessors. The diagnostic reliability problem Rosenhan had claimed to reveal was, on the independent empirical evidence gathered in the years since, a real problem, even if his own data on it was fabricated. Psychiatric hospitals in the 1960s and early 1970s, on the accumulated historical record, really were the depersonalising and often harmful institutions Rosenhan had depicted. The deinstitutionalisation movement, whatever else it may have done, was responding to a genuine set of failings in the mid-century American mental-health system.
None of that changes the fact that the single most influential paper in twentieth-century psychiatric criticism was, on the currently available evidence, not really an empirical paper at all. It was a piece of moral argument dressed as an empirical study, whose central claims about the twelve hospitals appear to have been embellished, exaggerated, or in some cases simply invented by an author who understood exactly what kind of story his profession was ready to hear.
The uncomfortable observation that follows, and the one Cahalan herself has returned to repeatedly since the book was published, is that Rosenhan’s fabrications were successful because they were pushing on a door that psychiatry itself had already partially opened. The reforms his paper accelerated were, in the direction they moved, defensible. The mental-health system that emerged from them, on the accumulated evidence of the fifty years since, has been catastrophically inadequate for the severely mentally ill population it was meant to serve. And the specific pattern that made Rosenhan’s paper so persuasive in 1973, which is the pattern of a dramatic anecdotal claim that a whole profession is failing catastrophically, has repeated itself across many other domains of scientific and public life in the decades since. Nobody involved in that 1973 story has come out of it looking quite the way they thought they would when the paper first landed. Including, on the accumulated evidence, David Rosenhan himself.
Kiran Athar is not a psychologist or a psychiatrist. She writes about psychology, mental health, and the corners of scientific history where the two intersect, drawing on peer-reviewed research and primary-source scholarship.
