Eight healthy people said a single sentence at the hospital door: they heard a voice saying "empty", "hollow" and "thud". Nothing more. It was enough to have them all admitted, almost all with a diagnosis of schizophrenia. The 1973 Rosenhan experiment is famous as an indictment of psychiatry. For us something else about it is decisive: what triggered the diagnosis was not the person but a sentence about a voice they heard — exactly what mediums have always described.
"It Is Hollow, It Is Empty" – the 1973 Experiment
The American psychologist David Rosenhan (Stanford, also a trained lawyer — not a psychiatrist in the narrow sense) sent eight mentally healthy participants, himself included, into twelve psychiatric hospitals across five US states. Each of these "pseudopatients" presented with a single invented complaint: they heard a voice saying words like empty, hollow and thud. No one lied beyond that: names and occupations were sometimes changed, but every account of their life, mood and relationships was true.
Once admitted, the voice — unsurprisingly — stopped at once, and the pseudopatients behaved entirely normally. They said they felt fine. Still, none was recognized as healthy. Seven were diagnosed with schizophrenia, one with manic-depressive disorder. The stays lasted between 7 and 52 days, on average about 19. In the end no one was discharged as "healthy" but with the label "schizophrenia in remission" — the diagnosis remained, merely deemed temporarily quiet. Rosenhan published the results in 1973 under the title On Being Sane in Insane Places in the journal Science.
Who Saw Through the Hoax
Perhaps the most uncomfortable observation of the study: the staff noticed nothing, while the fellow patients often did. "You're not crazy — you're a journalist or a professor, you're checking up on the hospital," some pseudopatients were told. Once the label was attached, their normal behaviour was reinterpreted in its light: taking notes was logged as a symptom ("writing behaviour"), waiting outside the cafeteria as an "oral-acquisitive syndrome". The file coloured the perception of every single act.
This is the mechanism social psychology calls labelling — related to the conformity and authority pressure made visible by the Asch experiment. Once the group or the institution has decided what a person is, everyone sees only what is expected.
The Counter-Experiment
A teaching hospital that had heard of the results believed itself immune: such a thing could not happen there. Rosenhan announced that over the following weeks he would send one or more pseudopatients. Staff were to rate every new admission. Of 193 admitted patients, 41 were suspected of being impostors, 19 of them by at least one psychiatrist. In truth Rosenhan had sent no one. So both directions failed: the healthy were taken for ill — and as soon as fakes were expected, the ill were taken for healthy. The diagnosis said more about the observers' expectation than about the observed.
The Point: Not the Content, the Frame
Here lies what makes this experiment interesting for the question of mediums. The only "symptom" the pseudopatients presented with was a voice they heard. That very same basic experience — perceiving a voice others do not hear — is at the centre of what mediums, the clairaudient, prophets and mystics of every culture describe. In the clinic of 1973 it served as a ticket into schizophrenia. In a Spiritualist circle the same account would have been a sign of a gift.
So it is not the content of the experience that decides whether someone counts as ill or gifted, but the frame in which it is interpreted. Rosenhan does not prove that hearing voices is healthy — he proves how shaky and context-dependent the line is that we draw between "madness" and "gift".
Hearing Voices Is More Common and More Harmless Than We Think
That this line really does need redrawing is shown by the research of recent decades. In the 1980s the Dutch psychiatrists Marius Romme and Sandra Escher demonstrated that a substantial share of people who hear voices never enter psychiatric care: they live with their voices, make sense of them and function entirely normally in daily life. From this insight grew the international Hearing Voices Movement. Population studies estimate that several percent of people hear voices at some point without ever becoming psychotic. Hearing voices in itself is therefore not an illness — it is a widespread human experience that becomes distressing only under certain conditions.
This overturns the silent premise of 1973. Rosenhan's doctors leapt reflexively from "hears a voice" to "schizophrenic". That inference is empirically false.
Culture Decides How a Voice Sounds
Just how much the frame shapes the experience itself was shown by the Stanford anthropologist Tanya Luhrmann in a widely noted comparative study. She interviewed about 20 schizophrenia voice-hearers each in the USA, India and Ghana. The result split sharply along cultural lines: the Americans experienced their voices mostly as brutal, alien intruders telling them they were worthless and should die. Those in South India and Ghana far more often described familiar, advising, even kindly voices — relatives, gods, good powers. In Accra many called the experience predominantly positive, several heard God speaking audibly to them.
Luhrmann concludes that the content of voices responds to "cultural invitation" — her social kindling hypothesis. Where a society reads voices as illness they are experienced as hostile; where they count as contact with a spiritual world they become a relationship. The same neural basis, an entirely different experience — depending on the frame.
The Mediumistic Case: the Durham Study
That leaves the decisive question: does the voice-hearing of mediums actually differ in measurable ways from that of psychiatric patients? In 2021 the religion scholar Adam Powell and the psychologist Peter Moseley, within the Hearing the Voice research project at Durham University, presented a study on precisely this (When spirits speak, in Mental Health, Religion & Culture). They surveyed 65 Spiritualist mediums who report clairaudient contact and compared them with the general population.
The findings paint a clear picture that stands apart from the clinical hallucination:
- The mediums began hearing voices on average at 21.7 years, many already in childhood; 18% said they had known it "for as long as they could remember".
- 71% had not yet encountered Spiritualism as a movement before their first experience. So the experience came first, the interpretation later — it was not simply taught.
- The strongest factor was not a proneness to hallucinate but absorption: the capacity to become wholly immersed in inner experience. The voices were experienced mostly as positive, meaningful and controllable — not as an assault.
This is exactly the difference that matters: not "whether" someone hears voices but how. Mediumistic voice-hearing is ego-syntonic, controlled and meaningful — the opposite of the tormenting, externally driven voices Luhrmann's American patients described. The study sits seamlessly alongside the Brazilian research of Alexander Moreira-Almeida on 115 mediums, who likewise found that the frequency of anomalous experiences does not correlate with markers of mental disorder.
A Mirror Image to Rosenhan: Birgit Fischer
Just how exactly this pattern can be reversed is shown by a recent German-language case. The medium Birgit Fischer – today one of the best-known seers in the German-speaking world – recounts on the podcast {ungeskriptet} by Ben how in her early twenties she desperately wanted to be rid of her perceptions. She had a university degree, was "fully in the business world" and tried to suppress the voices – which only made them louder and stronger. Finally she went to doctors with the plea: "Please switch this off." She herself was convinced: "I see faces, I hear voices. I am schizophrenic."
The result was the exact opposite of 1973. After neurological and psychological testing a neurologist told her: "Frau Fischer, this is a gift." She was not schizophrenic – "we have now tested everything" – and she had that finding "in black and white", confirmed by more than one source. Where Rosenhan's pseudopatients received a diagnosis of schizophrenia from a single sentence about a voice, in Fischer's case the same experience – thoroughly examined – expressly did not lead to illness but to the recognition of a gift. (By her own account she then had the perception "shut down" with an antidepressant and slipped into a severe burnout – the suffering came not from the gift but from the attempt to suppress it.)
What Rosenhan Does Not Prove
Honesty is part of this. The Rosenhan experiment is itself contested today: in The Great Pretender (2019) the journalist Susannah Cahalan reconstructed that only some of the pseudopatients can be documented at all, that Rosenhan's own medical record diverged from his publication, and that a participant with a positive experience apparently dropped out of the data. The study should therefore be treated not as clean proof but as a historical spotlight.
And above all: the finding that mediums are mentally healthy and that their voice-hearing differs from psychosis says, in itself, nothing yet about where the content of those voices comes from. It only refutes — and that is a great deal — the convenient equation "hears a voice, therefore ill". How thin the line between gift and diagnosis can remain in an individual case is shown by the documented case of a medium in psychiatry.
The Open Question
This is exactly where it becomes interesting rather than tied off. If the voice-hearing of mediums is not pathological, occurs under control, often begins in childhood and is not first produced by Spiritualist doctrine — then the question shifts. It is no longer "Is this person disturbed?" but: Where does what they hear come from? Is the voice an inner product of a particularly absorption-prone brain — or a reception? Research cannot even ask this as long as it files every voice in advance as a symptom. It stands open the moment one treats, as this site does, the relationship between consciousness and brain as empirically unresolved rather than settled.
Rosenhan's eight pseudopatients set out to show how easily a healthy person is stamped a schizophrenic. Along the way they proved something else: a voice one hears is, in itself, no verdict — neither "ill" nor "gift". What it is is decided only by the frame we place it in. And the frame can be rethought.
Sources
- Rosenhan, D. L. (1973): On Being Sane in Insane Places. Science 179 (4070), 250–258 — eight pseudopatients, twelve hospitals, symptom "empty/hollow/thud", discharge as "schizophrenia in remission", counter-experiment with 41 of 193 admissions suspected.
- Cahalan, S. (2019): The Great Pretender. The Undercover Mission That Changed Our Understanding of Madness. Grand Central Publishing — critical reconstruction and doubts about the experiment's data.
- Romme, M., Escher, S. (1989): Hearing Voices. Schizophrenia Bulletin 15(2), 209–216 — basis of the Hearing Voices Movement; many voice-hearers without psychiatric care.
- Luhrmann, T. M., Padmavati, R., Tharoor, H., Osei, A. (2015): Differences in voice-hearing experiences of people with psychosis in the USA, India and Ghana. British Journal of Psychiatry 206(1), 41–44; and Hearing Voices in Different Cultures: A Social Kindling Hypothesis. Topics in Cognitive Science 7(4), 646–663.
- Powell, A. J., Moseley, P. (2021): When spirits speak: absorption, attribution, and identity among spiritualists who report "clairaudient" voice experiences. Mental Health, Religion & Culture — 65 mediums; first contact on average at 21.7 years, 18% "for as long as they could remember", 71% before encountering Spiritualism, central role of absorption.
- Moreira-Almeida, A., Lotufo Neto, F., Greyson, B. (2006): Dissociative and Psychotic Experiences in Brazilian Spiritist Mediums. Psychotherapy and Psychosomatics 76(1), 57–58 — see the study of 115 mediums.
- Birgit Fischer on the podcast "{ungeskriptet} by Ben": 2026: Ich weiß, was passieren wird (Hellseherin). YouTube — account of the neurological and psychological testing ("you are not schizophrenic") and the neurologist's statement "this is a gift" (from approx. 00:11:40); see also the Birgit Fischer portrait.