ON 6 February 1969, David Lurie told a psychiatrist at Haverford State Hospital in Pennsylvania that he had been hearing voices. 鈥淗ollow鈥�, 鈥渆mpty鈥� and 鈥渢hud鈥�, they said.
The voices were the only symptom experienced by the otherwise healthy 39-year-old copywriter. After an in-depth interview, in which Lurie was asked about his family life and two children, he was diagnosed with schizophrenia and hospitalised.
Yet all was not as it seemed. David Lurie didn鈥檛 exist. This was, in fact, an alias for psychologist David Rosenhan of Stanford University in California, who .
Advertisement
Published in 1973, his study contributed to an erosion of public faith in psychiatry, a mistrust memorably portrayed in the 1975 film One Flew Over the Cuckoo鈥檚 Nest starring Jack Nicholson. Rosenhan鈥檚 work held up for scrutiny the often harmful nature of psychiatric hospitals and galvanised a growing movement to shut the large ones and replace them with smaller, community-based mental health centres. In its wake, 鈥減sychiatrists looked like unreliable and antiquated quacks unfit to join in the research revolution鈥�, says psychiatrist Allen Frances, formerly at Duke University School of Medicine in North Carolina.
Rosenhan鈥檚 paper was 鈥渙ne of the most influential pieces of social science published in the 20th century鈥�, says sociologist and historian Andrew Scull at the University of California, San Diego.
But it wasn鈥檛 all it seemed. After spending six years investigating Rosenhan and his famous work, I believe he may have carried out a second deception, the effects of which are still being felt in psychiatry today.
Rosenhan wasn鈥檛 the first person to infiltrate a psychiatric hospital and expose its horrors. Journalists, writers and other psychiatrists had previously done the same 鈥� but none with such rigour, in such numbers and in such an attention-grabbing way. Published by the journal Science, Rosenhan鈥檚 paper 鈥淥n Being Sane in Insane Places鈥� exploded onto the scene 鈥溾��, as one article written three decades after its publication observed.
The experiment, as Rosenhan described it, went as follows: eight men and women with no history or symptoms of mental health problems, including himself, presented at 12 psychiatric hospitals across the US, claiming to hear voices that said 鈥渢hud, empty, hollow鈥�. The rest of their personal histories (save for a few biographical adjustments for privacy reasons) remained intact.
All of the participants were diagnosed with a mental health condition, seven with schizophrenia and one with what is now known as bipolar disorder. Once admitted into hospital, they dropped their hallucinations and behaved as they normally did, or as normally as being in hospital allowed. Yet in each case, clinicians viewed their behaviours through the prism of their presumed condition. When one took research notes on the activities of the ward, for example, nurses reported that the 鈥減atient engages in writing behavior鈥�, according to Rosenhan鈥檚 paper.
鈥淗ow many patients might be 鈥榮ane鈥� outside the psychiatric hospital but seem insane in it 鈥� not because craziness resides in them, as it were, but because they are responding to a bizarre setting?鈥� Rosenhan wrote in his paper.
During these hospitalisations, which lasted between seven and 52 days, some 2100 psychoactive drugs were administered to the participants. All were eventually released, although against medical advice and with their diagnoses labelled as 鈥渋n remission鈥�. 鈥淲e now know that we cannot distinguish insanity from sanity,鈥� Rosenhan concluded.
The experiment was published at the height of the anti-psychiatry movement, at a time when psychiatrist Thomas Szasz wrote that mental illness was a 鈥渕yth鈥� and psychiatrist R. D. Laing claimed that people with schizophrenia were the sane ones in an insane world. Sociologist Erving Goffman exposed how similar psychiatric hospitals were to prisons in his book Asylums, while pop culture spread fear and distrust. Meanwhile, the field faced an internal reckoning as it moved away from Sigmund Freud鈥檚 psychoanalysis.
By the late 1980s, . Despite virulent criticism from psychiatrists, who claimed when the study was first published that it was misleading and unscientific, it is .
鈥淗e concluded that it was impossible to distinguish insanity from sanity鈥�
Beyond the original nine-page paper in Science, which largely focuses on Rosenhan鈥檚 experience 鈥� the bad food, boredom, pacing, the neglect and even abuse of fellow patients he witnessed 鈥� scant public information exists about the study. The eight participants never revealed themselves, and Rosenhan wrote nothing further on the topic before his death in 2012. But in the course of my research, I gained access to a trove of documents he left behind 鈥� decades of correspondence, diary entries and even an unpublished book on the experiment 鈥� and interviewed his family and hundreds of friends and colleagues.
It quickly became clear that some things didn鈥檛 match up. In Rosenhan鈥檚 files, I found his Haverford Hospital medical records, which presented an entirely different portrait to that in his paper. One of the foundational principles of 鈥淥n Being Sane in Insane Places鈥� was that all of the pseudopatients presented with just one symptom: voices that said 鈥渢hud, empty, hollow鈥�. Rosenhan didn鈥檛 follow his own script.
The admitting psychiatrist Frank 鈥淟ewis鈥� Bartlett wrote that David Lurie was so disturbed by the voices that he had put copper over his ears 鈥� an almost cliched example of the 鈥渢infoil hat delusion鈥� commonly reported by people experiencing serious mental health problems. 鈥淗e has felt that he is 鈥榮ensitive to radio signals and hear[s] what people are thinking鈥�,鈥� Bartlett wrote. Hallucinations and disturbances in thought patterns, particularly believing you can hear or control other people鈥檚 thoughts, are considered key symptoms of schizophrenia, so would have contributed to his misdiagnosis.
Suspicious symptoms
Rosenhan also claimed that his wife 鈥渄id not know how disturbed and helpless and useless鈥� he was, that he had 鈥渢hought of suicide鈥� and believed that 鈥渆veryone would be better off if he was not around鈥�. Such ideation and threats of self-harm would provide strong grounds for immediate and necessary hospitalisation. 鈥淭o not hospitalise such a patient would be professionally unethical, and, in almost every circumstance, malpractice,鈥� says Michael Meade, chairman of psychiatry at Santa Clara Valley Health and Hospital System in San Jose, California. Intentionally or not, Rosenhan had served up a realistic characterisation of a man experiencing something more serious than the one-symptom portrait of illness his pseudopatients were supposed to describe.
The discovery of such inconsistencies prompted me to try to track down the other pseudopatients who participated in the study. I found the first with relative ease, thanks to a stray handwritten note on an outline for Rosenhan鈥檚 unpublished book, which mentioned the last name Underwood.
This led me to Bill Underwood, a retired software engineer living in Austin, Texas, who had volunteered to be a pseudopatient while taking Rosenhan鈥檚 psychopathology seminar during his graduate studies at Stanford. Underwood described the seven days he spent at what was then Agnews State Mental Hospital in Santa Clara, California, as harrowing 鈥� he was medicated, misdiagnosed and mistreated. He says he was hardly prepped for his undercover role, and that Rosenhan told him nothing about data collection or how to interact with the patients or staff. In fact, Underwood didn鈥檛 recall collecting much if any data. This made me wonder, how did Rosenhan get such specific figures for things like the total number of minutes psychiatrists spent on the ward, numbers so precise they came down to a decimal point?
Underwood led me to another of Rosenhan鈥檚 former students, , now a public health psychologist at the University of Minnesota. Although his misdiagnosis with schizophrenia and 19-day hospitalisation match the account in Rosenhan鈥檚 paper, little else does. Indeed, Lando wrote about his hospitalisation in an overlooked paper called 鈥溾�� in 1976. The takeaway? Lando had a positive experience. He described the ward he was on as a healing, comforting environment, where the staff truly cared for the patients and they recovered.
In further handwritten notes he made while in hospital, Lando described touching moments of compassion and connection. However, none of this made it into the Science paper. Rosenhan instead dropped Lando from his study, making him only a footnote: 鈥淒ata from a ninth pseudopatient are not incorporated in this report because, although his sanity went undetected, he falsified aspects of his personal history, including his marital status and parental relationships,鈥� Rosenhan wrote. Yet some details from Lando鈥檚 19-day hospitalisation, including his flirtation with a nurse, did make it into the paper.
What鈥檚 more, an earlier draft of the paper, which I found in Rosenhan鈥檚 files, featured nine pseudopatients and no footnote, leading me to believe that Lando hadn鈥檛 yet been removed. Yet the draft listed the same number of pills taken, the same average length of stay and the same number of minutes psychiatrists spent on the ward as the published version. Not one number changed after he was taken out.
鈥淩osenhan was interested in diagnosis, and that鈥檚 fine, but you鈥檝e got to respect and accept the data,鈥� Lando told me. 鈥淐learly, he had his idea, his hypothesis and he was going to confirm that hypothesis.鈥�
Adding it all up
What about the other participants? I spent years trying to find them and even hired a private investigator. Lando and Underwood only knew of each other, and even Rosenhan鈥檚 long-time research assistant could only recall their two names. I could find no proof that the other pseudopatients existed.
This isn鈥檛 the first time a classic study of mental health has come under fire. In October, researchers at King鈥檚 College London identified 26 papers written by world-renowned British psychologist Hans Eysenck, who died in 1997, as . Among other things, Eysenck had claimed that personality had a bigger impact on someone鈥檚 cancer risk than smoking.
Even the conclusions of Stanley Milgram鈥檚 famous 1960s experiments on obedience to authority, in which participants subjected unseen people to increasingly painful shocks when told to do so by experimenters, have recently been cast into doubt, following the identification of serious methodological issues.
If Rosenhan had merely exaggerated the findings of a small, anecdotal study, I wouldn鈥檛 be writing about him. But the effects of his paper can still be felt today. Within 10 years of its publication, the resident population in state and county psychiatric hospitals in the US had . In the UK, there has been a similar drop in places for people in need of in-patient mental health treatment: today there are , compared with nearly 70,000 in the 1980s. The UK now , while the US is a staggering .
鈥淒espite years of hunting, I found no proof that the other pseudopatients ever existed鈥�
Many psychiatric hospitals were neglectful, even harmful places. Rosenhan raised valid points about their disturbing limitations, the view of psychiatric illness as less worthy of our sympathy and support than physical illness, the lack of validity of psychiatric labels and the role of priming in diagnosis. Yet by excluding and exaggerating data, he presented a sensationalised and oversimplified take on these complex issues. We are worse off because of it.
Today, many people with untreated mental health problems are living on the streets or are in prison. When Rosenhan conducted his study, 5 per cent of prisoners in the US met the criteria for serious mental illness; today, it is . According to a Harvard Medical School report, .
People with mental illness are still often stigmatised and poorly treated. . By 2025, the US could be more than 15,000 psychiatrists short.
It is time to reassess the paths that took us here, and, in so doing, begin to set things right. Rosenhan may not have created these issues, but by only pretending to expose their true nature, he contributed to a culture that has caused harm to the people in greatest need.
