Monthly Archives: February 2020

My last post outlined the work done by Cahalan’s The Great Pretender to expose Rosenhan’s deceits in On Being Sane in Insane Places. Highlighted was the question of misdiagnosis and the emergent conflict between Rosenhan and Spitzer.

Before continuing, it should be clear that, as regards pseudopatients being admitted to hospital, diagnosis was reliable; the system did work here, and doctors had no option (see the description of Rosenhan’s own committal, 184‑186).[1] Later, of course, the system didn’t work as well, exposed by Rosenhan’s subterfuge in telling hospitals to expect pseudopatients (120‑121). One can infer that diagnosis was likely a hit‑and‑miss affair; and even Spitzer, recognising a good career opportunity when he sees it, no misdiagnosis on his part there, is quoted as praising Rosenhan’s contribution to an important debate (193).[2] On 125 Cahalan has already commented on the impact Rosenhan’s revelations had, at a time (‘[i]t was no coincidence’) when the status of homosexuality as a mental disorder was finally being challenged, and this aspect of Rosenhan’s research readily lent itself to quantification, that is, the measurement of misdiagnosis.

He had reasons, then, to favour the experimental side of the study at the expense, perhaps, of those non-experimental aspects that might have proven more rewarding. In a short paper, he could get away with inflating the number of pseudopatients, for example (if that is what he did, The Great Pretender inconclusive here); however, when faced with the need to produce a book‑length study, mere number-crunching was never going to suffice.

Consequently, in The Great Pretender, the pages devoted to the experiences of pseudopatients in hospital – Rosenhan himself (86-105), Bill (142-147, 156‑161, and Harry (222-233) – are easily the most interesting, even though Cahalan, having concluded Rosenhan can only be untrustworthy, often misreads or ignores their significance.

The hole Rosenhan dug for himself

The Great Pretender is at its best when it offers a sometimes detailed description of Rosenhan’s own writing, fieldnotes from his time as the first pseudopatient, or draft versions of various chapters in the unfinished book‑length study his publisher would subsequently sue him for not completing. As published, the original paper is slight, presenting evidence that fits in with contemporary concerns, and perhaps its author failed to predict the impact it would have. Moreover, Rosenhan would not be the first researcher who undertook and published research to further his career (which is not to say others have necessarily served up bogus data). Cahalan alludes to the difficulties he might have had, and been aware of, in gaining a tenured post (268). For whatever reason, this expert on abnormality wanted to be a normal – that is, mainstream – experimental researcher. A wasted opportunity.

The road not taken – because it would have taken far too long?

So where did Rosenhan go wrong? The 1960s had opened with the publications of Goffman’s Asylums, Szasz’s The Myth of Mental Illness, Becker’s Outsiders, and Laing’s The Divided Self, all major texts referenced at the beginning of Rosenhan’s paper; he clearly wanted to signal straightaway the context for his own modest empirical study. However, given his selection, it’s striking that none of those texts would have claimed to be experimental. Goffman, for example, makes it quite clear at the beginning of Asylums that his approach was the precise opposite of Rosenhan’s: he was not committed as a patient and he prioritised ‘ethnographic detail regarding selected aspects of patient social life’, ignoring the ‘usual kinds of measurements and controls’; going after ‘statistical evidence for a few statements would preclude my gathering data on the tissue and fabric of patient life’.[3] Did Rosenhan read and think about that opening methodological statement?

Rosenhan does use the phrase ‘total institution’ in passing,[4] but appears less interested in basing his research on Goffman’s insights. Throughout, even when attempting to describe interactions, findings have to be presented reductively in statistical form – see, for example, the comparison of interactions as experienced by pseudopatients and the ‘young lady’ at Stanford;[5] or the passage, ‘[a]verage daily contact with psychiatrists, psychologists, residents, and physicians combined ranged from 3.9 to 25.1 minutes, with an overall mean of 6.8’.[6] Beyond the broadest of generalisations (‘not much time’), any attempt to ‘objectively’ measure the time nurses spent with patients is ridiculous: even were it possible to use a stop-watch and time the action reliably, such measures would still have no meaning (other than that some are so foolish as to think presentation of statistical data alone signals scientific rigour).

Goffman, by way of contrast, explains in detail how space is divided to reinforce the way patients interact with staff and with each other, the careful delineation of off-limits space from surveillance space, where patients might expect to have no or little freedom; or those spaces from free places, which are, in turn, not the same as group territories, which might, again, be distinguished from personal territories.[7] The meanings attached to space are inseparable from status and identity; as such, these meanings cannot be quantified in the way Rosenhan’s statistical data would have us believe. One simple illustration of personal territories will underscore their precarious nature:

… a few patients would carry their blankets around with them during the day and, in an act thought to be highly regressive, each would curl up on the floor with his blanket completely covering him; within this covered space each had some margin of control.[8]

This demonstration of the way ethnographic research might unfold has been chosen for two reasons. Firstly, it takes time. Secondly, this passage in Goffman’s book can be likened to the way, in Doing the Business, Hobbs divides the pub (a ‘rigidly stratified institution’) into four different sections, all as a prelude to the story of Keith, whose career as a failed entrepreneur begins and ends in the least prestigious part of the pub, section 4.[9] These examples, from Goffman and Hobbs, illustrate clearly what the sociologist means by the social construction of reality, or the subjective meanings actors attach to behaviour: whatever Rosenhan himself might have understood, and both his published and unpublished writing makes it clear he does get the sociology here, none of it is comprehensible to his critic Spitzer. So fixated is Spitzer on Rosenhan’s erroneous use of ‘in remission’, he ignores the more obviously bogus data.

What might have been – Rosenhan on the tissue and fabric of patient life

When looking at the parts of The Great Pretender that do offer tantalising glimpses into what an ethnographic version of Rosenhan’s published study might have looked like, it is possible to highlight passages that would provide substance.[10]

Rosenhan’s first impression comes with the, as reported, criticism of Szasz and the anti-psychiatry movement, when he notes: ‘They [patients] really are different from me’ (89). If his hypothesis, going in, is that hospitals cannot detect pseudopatients, he remains, nonetheless, committed to his belief in something, here called ‘sanity’, as distinct from something else, here called ‘insanity’. This observation would then help explain his reaction to Harris, who first offers an ‘unexpected intimacy’ (96) when treating Rosenhan as ‘a person, not a leper’ (97); and then rejects him.[11] On 99 Rosenhan, as quoted, ‘behaved like a patient’.

Think about it for a moment: if the confident, charismatic, middle-class professional Rosenhan can be so affected, a good example of what depersonalisation feels like, what might it be like for others, less confident, less assured in their social status?

From here it is easy to see how, over-compensating, Rosenhan would then ‘ham it up’ at the party described by Staub: ‘mesmerising the crowd with his dramatic tale’ (188). That he claimed to have worn a wig that did more than just change his appearance, which Cahalan finds ‘bewildering’ (189), makes perfect sense for someone who wants to live off the experience while, simultaneously, distancing himself from the shame he evidently felt when letting Harris put him in his place (and an understanding of what Goffman had written about space would have come in handy here, of course).

It also becomes easier to see the missed opportunity provided by Harry’s story. By this stage of the book, Rosenhan is a lost cause for Cahalan, who can only see how he has trashed poor Harry’s experience because it doesn’t fit in with what he wants to say. Yet, on 224, Harry is reminded that he is still a patient, and his response when the psychiatrist says, ‘It’s up to you’ – ‘Having that thrown back at me wasn’t entirely pleasant’ – is remarkably similar to Rosenhan’s feeling when Harris turns on him.

See also, in this vein, Bill’s account of the attendant, ‘as if seeing Bill as a human being for the first time’ (159). All three pseudopatients, then, provide compelling evidence of depersonalisation.

For Harry, there is also the moment a nurse gives him his own file: Cahalan can say this is ‘an unusual moment in any hospital, let alone a psychiatric one’ (224-225), while ignoring its similarity to Rosenhan’s description of the nurse unbuttoning her uniform to adjust her bra (95).[12] In each case, it matters little what the patient, a non-person, thinks.

Later, Harry, who has done all he can to ingratiate himself, is refused a pass: the unintentional irony of the older Harry’s recollection (‘… the most surreal experience. Here I am, I’m in a psychiatric institution and I can’t convince them that it’s safe to let me go’) is, of course, missed by Cahalan (225). Rosenhan’s take on Harry (‘HE LIKES IT’) is, supposedly, incredible: Cahalan seems to think Rosenhan is saying he cannot understand why anyone would not hate being in a hospital (224). But perhaps Rosenhan has seen the extent to which Harry, in embracing the role of patient, has failed to understand what is happening to him.

Conclusion – still defending Rosenhan

Rereading On Being Sane in Insane Places alongside The Great Pretender, I have been reminded of long-time reservations but one can only be impressed by how much Rosenhan did include about labelling, power relations, and depersonalisation. The experiences of Rosenhan himself, Bill, and Harry – that is what’s important, and the few examples given above will have to suffice as indicators of how much better a full-length sociological account might have been.

As for Cahalan – it’s clear, in so many places, that she cannot see beyond the label: stickiness, anyone?

Finally, this review has been an attempt to defend Rosenhan against the likely onslaught that will follow publication of The Great Pretender. In particular, to defend him as someone on the brink of producing a major research study when all people will want to talk about now are the lies. It’s only just, then, to end with a quotation from The Divided Self, Laing noting that ‘one frequently encounters “merely” before subjective, whereas it is almost inconceivable to speak of anyone being “merely” objective’.[13] Indeed.

[1] And don’t forget Rosenhan’s own regard, as published, for medical staff: ‘… our overwhelming impression of them was of people who really cared, who were committed and who were uncommonly intelligent. Where they failed, as they sometimes did painfully, it would be more accurate to attribute those failures to the environment in which they, too, found themselves than to personal callousness’ (On Being Sane, 399).

[2] One should also consider Rosenhan’s own career prospects at Stanford, as outlined, 117ff.

[3] Erving Goffman, 1991. Asylums: Essays on the Social Situations of Mental Patients and Other Inmates (first published 1961), Harmondsworth: Penguin Books, 7-8.

[4] On Being Sane, 399.

[5] Ibid, 393-394.

[6] Ibid, 396.

[7] Asylums, 203-220.

[8] Ibid, 219.

[9] Dick Hobbs, 1988. Doing the Business: Entrepreneurship, the Working Class, and Detectives in the East End of London, Oxford: Clarendon Press, 142-147.

[10] It’s important to note here that the gloss provided by Cahalan’s text is never an adequate substitute for Rosenhan’s own version, no matter how ‘faithful’ she has been to the ‘original’.

[11] In On Being Sane, Rosenhan writes: ‘The mentally ill are society’s lepers’ (390).

[12] A scene included in On Being Sane (395).

[13] RD Laing, 1965. The Divided Self (first published 1960), Harmondsworth: Penguin Books, 25.


I have always had reservations about David Rosenhan’s On Being Sane in Insane Places; its inferiority to the way a (qualitative) sociologist might have approached the same topic is evident.[1] And now Susannah Cahalan’s The Great Pretender has exposed Rosenhan as a fake: he made it all up (well – some of it).[2] However, reading her account of Rosenhan’s unpublished writing, fieldnotes in particular, has been more interesting. He should have had the courage to follow through his insights into power relations and produce a rich ethnographic account of life in a mental hospital. His inability to do so, and what that tells us about the doing of research, rather than any dishonesty on Rosenhan’s part is what we might more usefully take away from this sorry tale.

There’s at least one good review out there (if you ignore the headline)

I discovered this new book when reading a Guardian review, one that, predictably, failed even to summarise the case accurately.[3] Rosenhan describes the experience of pseudopatients, a small group of people admitted to hospital as schizophrenics when they claimed to be hearing voices. However, in its first paragraph, the Guardian review says they were institutionalised against their will, an odd mistake for even the most careless reader to make. A bit later we’re told Rosenhan’s study led to the first edition of the DSM, the classification system still used in the United States for diagnosing disorders. However, the first edition dates back to the 1950s; and so it’s left to a far better review in The Spectator to indicate the damage done by DSM-III. According to Andrew Scull, there is now a ‘rigidly biologically reductionist psychiatry, one that (falsely) claimed that the “diseases” the DSM identified and listed were akin to those that mainstream medicine diagnosed and treated’. Scull calls the case ‘a fraud whose real-world consequences still resonate today’, a conclusion the magazine then distorts for its misleading headline – ‘How a fraudulent experiment set psychiatry back decades’.[4]

If Rosenhan did set out to highlight misdiagnosis, such errors remain problematic today, and one might do better than blame the messenger whose message remains valid.[5]

Further, it’s worth pointing out here that Rosenhan’s modest little paper would have had a lesser impact had it not seemed to reinforce criticisms going back a decade or more.

Rosenhan’s nemesis?

On Being Sane in Insane Places was clearly of its time, the anti-psychiatry movement boldly dismissive of the pretensions of mainstream mental healthcare (summarised by Cahalan, 56-61). Nonetheless, Rosenhan’s study was rejected by the research community, ‘a chorus of hostile voices’ (175). By now Cahalan has anointed Robert Spitzer her hero and Rosenhan’s nemesis:[6] pouncing on Rosenhan’s claim that pseudopatients were released ‘in remission’, Spitzer points out that this term is seldom used – to do so would imply no symptoms of schizophrenia were being displayed.[7] Hence, the conclusion that, if pseudopatients were judged to be in remission, it must have been ‘a function of [their] behaviors and not of the setting (psychiatric hospital) in which the diagnoses were made’.[8] Rosenhan claims that, once the person has been labelled, all behaviours will be interpreted as evidence of their illness; Spitzer, pointing out that this cannot be so, questions the logic of Rosenhan’s argument – hence the sneering refrain, ‘logic in remission’.[9]

So far it is a conventional case of one professional researcher challenging the work of another, policing disciplinary boundaries, what Spitzer must have thought he was doing. However, he remains blissfully unaware that he has, inadvertently, done Rosenhan a favour. Cahalan makes the same mistake. When she starts to outline ‘sloppiness that seemed unprofessional and possibly unethical’ (173), she refers to pseudopatient Bill’s doubt that he had been released ‘in remission’ (see also Bill’s hospital record, reproduced on 160). On 231 another pseudopatient, Harry, also denies that he left hospital ‘in remission’, Cahalan again drawing attention to Rosenhan’s ‘outright fabrications’. Well, untruthful he might have been, but the absence of any ‘in remission’ statements here undermines, at the very least, Spitzer’s line of attack and suggests Rosenhan was right after all. If this is the case, why he chose to lie about it remains a matter for speculation (to be discussed later).

With the dismantling of Rosenhan’s study and reputation now underway, Cahalan appears to side with Spitzer when Rosenhan refuses to identify the hospitals pseudopatients had been admitted to (179). Spitzer can do little wrong.[10] However, one might pause here for thought. Ethical considerations would have to ensure that no participants were identifiable; it’s puzzling that anyone would expect Rosenhan to freely surrender such details, and this same reservation would cover some if not all ‘outright fabrications’. When Harry is quoted as saying ‘there are some basic factual inaccuracies that, I mean, don’t advance anything’ (231), one again wonders if, or how far, such ‘inaccuracies’ might legitimately be claimed as an attempt to maintain confidentiality. If a statement ‘doesn’t advance anything’, can it be said to be a significant distortion of research findings? At the very least, this is a question worth asking.

It’s certainly reasonable to ask for evidence that a researcher hasn’t lied – if legitimate suspicions have been raised. But is it also reasonable for the researcher to change details that don’t matter? With a strong emphasis on details that don’t matter – perhaps, if that means participants are protected. Cahalan says Rosenhan lied about the safeguards put in place to help pseudopatients if they were unable to get themselves released from hospital (145, 249). That would indeed be serious.[11] Changing details here and there might well be a lesser crime. I began by saying I have always had reservations about Rosenhan’s claim that his research could be called an experiment; and this, I suspect, the adherence to conventional methodological procedure, even when it means playing fast and loose with ‘the facts’, when he might have chosen a quite different route, is where Rosenhan got himself deeper into trouble.

To be concluded.

[1] David Rosenhan, 1973. Symposium: On Being Sane In Insane Places, Santa Clara Lawyer, 13/3, 379-399. Online:

[2] Susannah Cahalan, 2020. The Great Pretender, Edinburgh: Canongate Books. On the ‘replication crisis’ see 270-274.

[3] Stephen Poole, 2020. The Great Pretender by Susannah Cahalan review – psychiatry’s dubious past, The Guardian, 10 January. Online:

[4] Andrew Scull, 2020. How a fraudulent experiment set psychiatry back decades, The Spectator, 25 January. Online:

[5] Since the 1970s research has continued to appear highlighting the problems of misdiagnosis and the role played by different kinds of bias; the problem isn’t going away. See:

Maureen R. Ford and Thomas A. Widiger, 1989. Sex bias in the diagnosis of histrionic and antisocial personality disorders, Journal of Consulting and Clinical Psychology, 57/2, 301-305.

Rebecca Pinto, Mark Ashworth and Roger Jones, 2008. Schizophrenia in black Caribbeans living in the UK: an exploration of underlying causes of the high incidence rate, British Journal of General Practice, June, 429-434.

Jolynn L. Haney, 2016. Autism, females, and the DSM-5: Gender bias in autism diagnosis, Social Work in Mental Health, 14/4, 396-407.

Michael A. Gara, Shula Minsky, Steven M Silverstein, Theresa Miskimen, and Stephen M. Strakowski, 2019. A Naturalistic Study of Racial Disparities in Diagnoses at an Outpatient Behavioral Health Clinic, Psychiatric Services, 70/2,: 130-134.

On gender bias one might also note that the only pseudopatient (supposedly) diagnosed as manic depressive was a female painter, Laura.

[6] Robert L Spitzer, 1975. On Pseudoscience in Science, Logic in Remission, and Psychiatric Diagnosis: A Critique of Rosenhan’s ‘On Being Sane in Insane Places’, Journal of Abnormal Psychology, 84/5, 442-452.

[7] Ibid, 444.

[8] Ibid, 445.

[9] On 200 an erstwhile colleague is found quoting Spitzer, basking in reflected glory. One wonders how long Spitzer himself dined out on his own recycled witticism. As I point out below, blinded by his own methodological assumptions, he missed a more obvious opportunity to put the boot in. Hubris, indeed.

[10] See for example the passage at the bottom of 178: ‘the drollest piece of academic literature I’ve ever read’ etc. Really?

[11] Rosenhan’s own take: ‘I was not sensitive to these difficulties at the outset of the project, nor to the personal and situational emergencies that can arise, but later a writ of habeas corpus was prepared for each of the entering pseudopatients and an attorney was kept “on call” during every hospitalization. I am grateful to John Kaplan and Robert Bartels for legal advice and assistance in these matters’ (On Being Insane, 382, fn6). According to Cahalan, when she spoke to him, ‘Bartels [Kaplan’s assistant] was a bit hazy on details’ (145); and there is nothing from Kaplan. Sometimes Cahalan’s witnesses readily recall events in detail; sometimes they just don’t remember.