ON BEI NG SANE IN I NSANE PLACES*

D. L. Rosenhan**
If sanity and insanity exist, how shall we know them?
The question is neither capricious nor itself insane. However
much we may be personally convinced t hat we can tell the normal
from the abnormal, the evidence is simply not compelling. It is com-
monplace, for example, to read about murder trials wherein eminent
psychiatrists for the defense are contradicted by equally eminent psy-
chiatrists for the prosecution on the matter of the defendant' s sanity.
More generally, there are a great deal of conflicting dat a on the re-
liability, utility, and meaning of such terms as "sanity, " "insanity,"
"mental illness," and "schizophrenia" (1). Finally, as early as 1934,
Benedict suggested that normality and abnormality are not univer-
sal ( 2 ) . What is viewed as normal in one culture may be seen as quite
aberrant in another. Thus, notions of normality and abnormality
may not be quite as accurate as people believe they are.
To raise questions regarding normality and abnormality is in no
way to question the fact t hat some behaviors are deviant or odd.
Murder is deviant. So, too, are hallucinations. Nor does raising such
questions deny the existence of the personal angui sh that is often
associated with "mental illness." Anxiety and depression exist. Psy-
chological suffering exists. But normaht y and abnormality, sanity
and insanity, and the diagnoses t hat flow from them may be less
substantive t han many believe them to be.
At its heart, the question of whether the sane can be distin-
guished from the insane (and whether degrees of i nsani t y can be dis-
* Reprinted by permission of the author and Science magazine from Vol. 179, pp.
250-258, January 19, 1973. Copyright 1973 by the American Association for the Ad-
vancement of Science.
** The author is professor of psychology and law at Stanford University, Stanford,
California 94305. Portions of these data were presented to colloquiums of the psychol-
ogy departments at the Universityof California at Berkeley and at Santa Barbara;
University of Arizona, Tucson; and Harvard University, Cambridge, Massachusetts.
237
238
CLINICAL SOCIAL WORK JOURNAL
tinguished from each other) is a simple matter: do the salient charac-
teristics that lead to diagnoses reside in the patients themselves or in
the environments and contexts in which observers find them? From
Bleuler, through Kretchmer, through the formulators of the recently
revised Di agnos t i c and St at i s t i c al Ma n u a l of the American Psychi-
atric Association, the belief has been strong that patients present
symptoms, that those symptoms can be categorized, and, implicitly,
that the sane are distinguishable from the insane. More recently,
however, this belief has been questioned. Based in part on theoretical
and anthropological considerations, but also on philosophical, legal,
and therapeutic ones, the view has grown that psychological categor-
ization of mental illness is useless at best and downright harmful,
misleading, and pejorative at worst. Psychiatric diagnoses, in this
view, are in the minds of the observers and are not valid summaries
of characteristics displayed by the observed (3--5).
Gains can be made in deciding which of these is more nearly ac-
curate by getting normal people (that is, people who do not have, and
have never suffered, symptoms of serious psychiatric disorders) ad-
mitted to psychiatric hospitals and then determining whether they
were discovered to be sane and, if so, how. If the sanityof such pseu-
dopatients were always detected, there would be prima facie evidence
that a sane individual can be distinguished from the insane context
in which he is found. Normality (and presumably abnormality) is dis-
tinct enough that it can be recognized wherever it occurs, for it is
carried within the person. If, on the other hand, the sanity of the
pseudopatients were never discovered, serious difficulties would arise
for those who support traditional modes of psychiatric diagnosis.
Given t hat the hospital staff was not incompetent, t hat the pseu-
dopatient had been behaving as sanely as he had beenoutside of the
hospital, and that it had never been previously suggested that he be
belonged in a psychiatric hospital, such an unlikely outcome would
support the view that psychiatric diagnosis betrays little about the
patient but much about the environment in which an observer finds
him.
This article describes such an experiment. Eight sane people
gained secret admission to 12 different hospitals (6). Their diagnostic
experiences constitute the data of the first part of this article; the re-
mainder is devoted to a description of their experiences in psychiatric
institutions. Too few psychiatrists and psychologists, even those who
have worked in such hospitals, know what the experience is like.
They rarely talk about it with former patients, perhaps because they
distrust information coming from the previously insane. Those who
have worked in psychiatric hospitals are likely to have adapted so
thoroughly to the settings that they are insensitive to the impact of
D. L. R OS E NHAN 239
t hat experi ence. And whi l e t her e have been occas i onal report s of re-
sear cher s who submi t t ed t hemsel ves to ps ychi at r i c hospi t al i zat i on
(7), t hese r esear cher s have commonl y r e ma i ne d i n t he hospi t al s for
shor t peri ods of t i me, often wi t h t he knowl edge of t he hospi t al staff. It
is difficult to know t he ext ent to whi ch t hey wer e t r eat ed like pat i ent s
or like r es ear ch colleagues. Never t hel ess, t hei r r epor t s about t he in-
side of t he psychi at r i c hospi t al have been val uabl e. Thi s art i cl e ex-
t ends t hose efforts.
P s e u d o p a t i e n t s a n d The i r S e t t i n g s
The ei ght pseudopat i ent s were a var i ed group. One was a psy-
chol ogy gr a dua t e st udent i n hi s 20's. The r e ma i n i n g seven were older
and "est abl i shed. " Among t hem wer e t hr ee psychol ogi st s, a pediatri-
ci an, a psychi at r i st , a pai nt er, and a housewi fe. Thr ee pseudopat i ent s
were women, fi ve were men. All of t he m empl oyed psuedonyms, lest
t hei r al l eged di agnoses embar r as s t hem l at er. Those who were i n
ment al heal t h professi ons al l eged a not he r occupat i on i n order to
avoi d t he speci al at t ent i ons t hat mi ght be accor ded by st aff, as a mat -
ter of cour t esy or caut i on, to ai l i ng col l eagues (8). Wi t h t he exception
of mysel f (I was t he first pseudopat i ent a nd my pr esence was known
to t he hospi t al admi ni s t r at or a nd chi ef psychol ogi st and, so far as I
can tell, to t he m alone), t he pr esence of ps eudopat i ent s a nd t he nat ur e
of t he r esear ch pr ogr am was not known to t he hospi t al st affs (9).
The set t i ngs wer e si mi l ar l y var i ed. I n or der to gener al i ze t he find-
i ngs, admi s s i on i nt o a var i et y of hospi t al s was sought . The 12 hospi-
t al s i n t he s ampl e were l ocat ed i n fi ve di f f er ent st at es on t he East and
West coast s. Some were old and shabby, some wer e qui t e new. Some
were r esear ch- or i ent ed, ot hers not . Some ha d good st aff-pat i ent
rat i os, ot her s wer e qui t e under st af f ed. Onl y one was a st r i ct l y pri vat e
hospi t al . Al l of t he ot hers were suppor t ed by st at e or f eder al funds or,
i n one i ns t ance, by uni ver si t y funds.
Aft er cal l i ng t he hospi t al for a n appoi nt ment , t he pseudopat i ent
ar r i ved at t he admi ssi ons office compl ai ni ng t h a t he h a d been hear-
i ng voices. Asked what t he voices sai d, he r epl i ed t ha t t hey were often
uncl ear, but as f ar as he could tell t hey sai d " empt y, " "hol l ow, " and
"t hud. " The voi ces were unf ami l ar a nd wer e of t he s a me sex as t he
pseudopat i ent . The choice of t hese s ympt oms was occasi oned by t hei r
appar ent s i mi l ar i t y to exi st ent i al sympt oms. Such s ympt oms are al-
leged to ar i se f r om pai nf ul concer ns about t he per cei ved meani ngl ess-
ness of one' s life. I t is as i f t he ha l l uc i na t i ng per son wer e sayi ng, "My
life is empt y and hol l ow. " The choi ce of t hese s ympt oms was al so
det er mi ned by t he a b s e n c e of a si ngl e r epor t of exi st ent i al psychoses
i n t he l i t erat ure.
240 CLINICAL SOCIAL WORK JOURNAL
Beyond al l egi ng t he s ympt oms a nd f al si f yi ng name, vocat i on,
and empl oyment , no f ur t her al t er at i ons of person, hi st or y, or circum-
st ances wer e made. The s i gni f i cant event s of t he pseudopat i ent ' s life
hi st or y wer e present ed as t hey ha d act ual l y occurred. Rel at i onshi ps
wi t h par ent s and siblings, wi t h spouse and chi l dr en, wi t h people at
work and i n school, consi st ent wi t h t he af or ement i oned except i ons,
were descri bed as t hey were or h a d been. Fr us t r at i ons and upset s
were descri bed al ong wi t h j oys a nd sat i sf act i ons. These fact s are im-
por t ant to remember. I f anyt hi ng, t hey st r ongl y bi ased t he subse-
quent resul t s i n favor of det ect i ng s ani t y, si nce none of t hei r hi st ori es
or cur r ent behavi or s were seri ousl y pat hol ogi cal i n a n y way.
I mmedi at el y upon admi ssi on t o t he ps ychi at r i c war d, t he. pseu-
dopat i ent ceased s i mul at i ng any s ympt oms of abnor mal i t y. I n some
cases, t her e was a bri ef peri od of mi l d ner vousness a nd anxi et y, si nce
none of t he pseudopat i ent s r eal l y bel i eved t hat t hey woul d be ad-
mi t t ed so easily. Indeed, t hei r s har ed f ear was t ha t t hey woul d be
i mmedi at el y exposed as f r auds a nd gr eat l y embar r as s ed. Moreover,
ma ny of t hem ha d never vi si t ed a ps ychi at r i c war d; even t hose who
had, never t hel ess had some genui ne f ear s about wh a t mi ght happen
to t hem. Thei r ner vousness, t hen, was qui t e appr opr i at e to t he nov-
elty of t he hospi t al setting, and i t abat ed r api dl y.
Apar t from t hat short -l i ved ner vousness, t he ps eudopat i ent
behaved on t he war d as he " nor ma l l y" behaved. The pseudopat i ent
spoke to pat i ent s and st af f as he mi ght or di nar i l y. Because t her e is
uncommonl y little to do on a ps ychi at r i c war d, he at t empt ed to en-
gage ot hers i n conversat i on. When as ked by st af f how he was feeling,
he i ndi cat ed t hat he was fine, t ha t he no l onger exper i enced symp-
toms. He responded to i nst r uct i ons f r om at t endant s , to cal l s for med-
i cat i on (whi ch was not swal l owed), and to di ni ng- hal l i nst r uct i ons.
Beyond such activities as wer e avai l abl e to hi m on t he admi ssi ons
ward, he spent hi s t i me wr i t i ng down hi s obser vat i ons about t he
war d, its pat i ent s, and t he staff. I ni t i al l y t hese not es were wr i t t en
"secret l y, " but as it soon became cl ear t ha t no one muc h cared, t hey
were subsequent l y wr i t t en on s t a nda r d t abl et s of paper i n such publ i c
pl aces as t he dayroom. No secret was ma de of t hese act i vi t i es.
The pseudopat i ent , ver y muc h as a t r ue ps ychi at r i c pat i ent , en-
t ered a hospi t al wi t h no f or eknowl edge of when he woul d be dis-
charged. Ea c h was told t ha t he woul d have to get out by hi s own
devices, essent i al l y by convi nci ng t he st af f t ha t he wa s sane. The
psychol ogi cal st resses associ at ed wi t h hospi t al i zat i on wer e consider-
able, a nd all but one of t he pseudopat i ent s desi red to be di schar ged
al most i mmedi at el y aft er bei ng admi t t ed. They were, t herefore, mo-
t i vat ed not onl y to behave sanel y, but to be par agons of cooperat i on.
That t hei r behavi or was i n no wa y di sr upt i ve is conf i r med by nur s i ng
reports, whi ch have been obt ai ned on most of t he pat i ent s. These re-
D. L. ROSENHAN 241
port s uni f or ml y i ndi cat e t hat t he pat i ent s wer e "f r i endl y, " "coopera-
tive, " a nd "exhi bi t ed no abnor mal i ndi cat i ons . "
The No r ma l Ar e Not Det ect abl y S a n e
Despi t e t hei r public "show" of sani t y, t he pseudopat i ent s were
never det ect ed. Admi t t ed, except i n one case, wi t h a di agnosi s of
schi zophr eni a (10), each was di s char ged wi t h a di agnos i s of schizo-
phr eni a "i n r emi ssi on. " The l abel "i n r emi s s i on" shoul d i n no way be
di smi ssed as a formal i t y, for at no t i me dur i ng a ny hospi t al i zat i on
ha d a ny quest i on been r ai sed about a ny ps eudopat i ent ' s si mul at i on.
Nor ar e t her e any i ndi cat i ons i n t he hospi t al r ecor ds t h a t t he pseu-
dopat i ent ' s st at us was suspect. Rat her , t he evi dence is st r ong t hat ,
once l abel ed schi zophreni c, t he ps eudopat i ent was st uck wi t h t hat
label. I f t he pseudopat i ent was to be di schar ged, he mus t nat ur al l y be
"i n r emi ssi on"; but he was not sane, nor, i n t he i nst i t ut i on' s view,
ha d he ever been sane.
The uni f or m fai l ure to r ecogni ze s ani t y c a nnot be at t r i but ed to
t he qual i t y of t he hospi t al s, for, al t hough t her e wer e consi der abl e var-
i at i ons a mong t hem, several ar e consi der ed excel l ent . Nor can it be
al l eged t h a t t her e was si mpl y not enough t i me to obser ve t he pseu-
dopat i ent s. Lengt h of hospi t al i zat i on r a nge d f r om 7 to 52 days wi t h an
aver age o f l 9 days. The pseudopat i ent s wer e not , i n fact , car ef ul l y ob-
served, but t hi s fai l ure cl ear l y s peaks mor e to t r adi t i ons wi t hi n
psychi at r i c hospi t al s t han to l ack of oppor t uni t y.
Fi nal l y, it cannot be sai d t ha t t he f ai l ur e to r ecogni ze t he pseu-
dopat i ent s' s ani t y was due to t he f act t ha t t hey wer e not behavi ng
sanel y. Whi l e t her e was cl ear l y some t ensi on pr es ent i n all of t hem,
t hei r dai l y vi si t ors could det ect no seri ous behavi or al consequences
- - nor , i ndeed, coul d ot her pat i ent s. I t was qui t e common for t he pa-
t i ent s to "det ect " t he pseudopat i ent s' sani t y. Dur i ng t he fi rst t hr ee
hospi t al i zat i ons, when accur at e count s wer e kept, 35 of a t ot al of 118
pat i ent s on t he admi ssi ons war d voi ced t hei r suspi ci ons, some vigor-
ously. "You' r e not crazy. You' re a j our nal i st , or a pr of essor [referri ng
to t he cont i nual not e-t aki ng]. You' re checki ng up on t he hospi t al . "
Whi l e mos t of t he pat i ent s were r eas s ur ed by t he pseudopat i ent ' s in-
si st ence t h a t he ha d been si ck before he came i n but was fi ne now,
some cont i nued to believe t ha t t he pseudopat i ent was s ane t hr ough-
out hi s hospi t al i zat i on (11). The f act t h a t t he pat i ent s oft en recog-
ni zed nor mal i t y when st aff di d not r ai ses i mpor t a nt quest i ons.
Fai l ur e to det ect sani t y dur i ng t he course of hospi t al i zat i on ma y
be due to t he f act t hat phys i ci ans oper at e wi t h a s t r ong bi as t owar d
wha t st at i st i ci ans call t he t ype 2 er r or (5). Thi s is to s ay t hat physi -
ci ans ar e mor e i ncl i ned to call a he a l t hy per son si ck (a fal se positive,
t ype 2) t h a n a sick person heal t hy (a fal se negat i ve, t ype 1). The rea-
242 CLINICAL SOCIAL WORK JOURNAL
sons for t hi s are not har d to fi nd: i t is cl ear l y mor e da nge r ous to mis-
di agnos e i l l ness t h a n heal t h. Bet t er t o err on t he si de of caut i on, to
s us pect i l l ness even a mong t he he a l t hy.
But wh a t hol ds for medi ci ne does not hol d equal l y well for psy-
chi at r y. Medi cal i l l nesses, whi l e unf or t una t e , are not c ommonl y pe-
j or at i ve. Ps ychi at r i c di agnos es , on t he cont r ar y, car r y wi t h t he m per-
sonal , l egal , a nd soci al s t i gma s (12). I t was t her ef or e i mp o r t a n t to see
whe t he r t he t endency t owar d d i a g n o s i n g t he s ane i ns a ne coul d be
r ever sed. The fol l owi ng e xpe r i me nt wa s a r r a nge d at a r es ear ch a nd
t e a c hi ng hos pi t al whose s t af f h a d he a r d t hes e f i ndi ngs but doubt ed
t ha t such a n error coul d occur i n t he i r hospi t al . The s t a f f was in-
f or med t ha t at some t i me dur i ng t he f ol l owi ng 3 mont hs , one or mor e
ps eudopat i ent s woul d a t t e mpt to be a dmi t t e d i nt o t he ps ychi at r i c
hospi t al . Ea c h s t af f me mbe r was a s ke d t o r at e each pa t i e nt who pre-
s ent ed hi ms e l f at admi s s i ons or on t he war d accor di ng to t he likeli-
hood t ha t t he pat i ent was a ps e udopa t i e nt . A 10-point scal e was used,
wi t h a 1 a nd 2 r ef l ect i ng hi gh conf i dence t ha t t h e pa t i e nt was a
ps eudopat i ent .
J u d g me n t s were obt ai ned on 193 pa t i e nt s who wer e a dmi t t e d for
ps ychi at r i c t r eat ment . All s t a f f who h a d ha d s us t a i ne d cont act wi t h
or pr i ma r y r esponsi bi l i t y for t he p a t i e n t - - a t t e n d a n t s , nur s es , psychi -
at r i st s, phys i ci ans , a nd ps yc hol ogi s t s - - we r e as ked to ma ke j udg-
ment s . For t y- one pat i ent s wer e al l eged, wi t h hi gh conf i dence, to be
ps eudopat i ent s by at l east one me mb e r of t he st aff. Twent y- t hr ee
were consi der ed suspect by at l eas t one ps ychi at r i s t . Ni ne t e e n were
suspect ed by one ps ychi at r i s t and one ot her s t af f member . Act ual l y,
no ge nui ne ps eudopat i ent (at l eas t f r om my group) pr es ent ed hi ms e l f
dur i ng t hi s period.
The exper i ment is i nst r uct i ve. I t i ndi cat es t ha t t he t e nde nc y to
des i gnat e s a ne peopl e as i ns a ne c a n be r ever sed whe n t he s t akes (in
t hi s case, pr est i ge and di agnos t i c acumen) are hi gh. But wh a t can be
sai d of t he 19 peopl e who wer e s us pect ed of bei ng " s a ne " by one psy-
chi at r i s t a nd anot her s t af f me mbe r ? Were t hese peopl e t r ul y " s ane, "
or was i t r a t he r t he case t ha t i n t he cour se of avoi di ng t he t ype 2 error
t he s t a f f t ended t o ma ke mor e er r or s of t he f i r st s or t - - c a l l i ng t he
cr azy " s a ne " ? Ther e is no wa y of knowi ng. But one t h i n g is cer t ai n:
a ny di agnos t i c process t ha t l ends i t s el f so r eadi l y t o ma s s i ve errors of
t hi s sor t c a nnot be a ver y rel i abl e one.
The Stickiness of Psychodiagnostic Labels
Beyond t he t endency to cal l t he he a l t hy s i c k- - a t e nde nc y t ha t
account s bet t er for di agnos t i c be ha vi or on a dmi s s i on t h a n i t does for
s uch be ha vi or af t er a l e ngt hy per i od of expos ur e- - t he d a t a s peak to
I). L. R OS E NHAN 243
t he massi ve rol e of l abel i ng i n ps ychi at r i c as s es s ment . Havi ng once
been l abel ed schi zophreni c, t her e is not hi ng t he pseudopar i ent can do
to overcome t he t ag. The t ag pr of oundl y colors ot her s' percept i ons of
hi m and hi s behavi or .
Fr om one vi ewpoi nt , t hese dat a ar e ha r dl y sur pr i si ng, for it has
l ong been known t hat el ement s are gi ven me a ni ng by t he cont ext i n
whi ch t hey occur. Gest al t psychol ogy ma de t hi s poi nt vigorously,
and Asch (13) demonst r at ed t ha t t her e ar e " cent r al " per sonal i t y
t rai t s (such as " wa r m" versus "col d") whi ch ar e so power f ul t hat t hey
mar kedl y col or t he me a ni ng of ot her i nf or mat i on i n f or mi ng an im-
pressi on of a gi ven per sonal i t y (14). " I ns a ne , " "schi zophr eni c, "
"mani c- depr es s i ve, "and "cr azy" ar e pr obabl y a mo n g t he most power-
ful of such cent r al t rai t s. Once a per son is des i gnat ed abnor mal , all of
hi s ot her behavi or s and char act er i st i cs ar e colored by t ha t label. In-
deed, t ha t l abel is so powerful t ha t ma n y of t he pseudopat i ent s' nor-
mal behavi or s were overlooked ent i r el y or pr of oundl y mi.~interpreted.
Some exampl es ma y cl ari fy t hi s issue.
Ear l i er I i ndi cat ed t hat t her e wer e no changes i n t he pseudopa-
t i ent ' s per s onal hi st or y and cur r ent s t at us beyond t hose of name,
empl oyment , and, wher e necessar y, vocat i on. Ot her wi se, a veri di cal
descri pt i on of per sonal hi st or y and ci r cums t ances wa s offered. Those
ci r cumst ances were not psychot i c. How wer e t hey ma d e cons onant
wi t h t he di agnos i s of psychosi s? Or wer e t hose di agnos es modi fi ed i n
such a wa y as to br i ng t hem i nt o accor d wi t h t he ci r cums t ances of t he
pseudopat i ent ' s life, as descri bed by hi m?
As f ar as I can det er mi ne, di agnos es wer e i n no wa y affect ed by
t he r el at i ve heal t h of t he ci r cums t ances of a pseudopat i ent ' s life.
Rat her , t he r ever se occurred: t he per cept i on of hi s ci r cumst ances was
shaped ent i r el y by t he di agnosi s. A cl ear exampl e of such t r ans l at i on
is found i n t he case of a pseudopat i ent who ha d h a d a close relation-
shi p wi t h hi s mot her but was r a t he r r emot e f r om hi s f at her dur i ng hi s
earl y chi l dhood. Dur i ng adol escence a nd beyond, however , hi s f at her
became a cl ose friend, whi l e hi s r el at i ons hi p wi t h hi s mot her cooled.
His pr esent r el at i onshi p wi t h hi s wi fe was char act er i s t i cal l y close
and war m. Apar t from occasi onal a ngr y exchanges , fri ct i on was min-
i mal . The chi l dr en ha d r ar el y been spanked. Sur el y t her e is not hi ng
especi al l y pat hol ogi cal about such a hi st or y. I ndeed, ma n y readers
may see a si mi l ar pat t er n i n t hei r own experi ences, wi t h no mar kedl y
del et eri ous consequences. Observe, however , how s uch a hi st or y was
t r ans l at ed i n t he psychopat hol ogi cal cont ext , t hi s f r om t he case sum-
mar y pr epar ed af t er t he pat i ent was di s char ged.
This white 39-year-old mal e. . , manifests a long histo~- of con-
siderable ambivalence in close relationships, which begins in early
244 CLINICAL SOCIAL WORK JOURNAL
childhood. A warm relationship with his mot her cools du~_ng his
adolescence. A di st ant relationship to his fat her is described as
becoming very intense. Affective stability is absent. His at t empt s
to control emotionality with his wife and children are punct uat ed
by angry outbursts and, in t he case of t he children, spanki ngs. And
while he says t hat he has several good friends, one senses
considerable ambivalence embedded in those relationships also . . . .
The f act s of t he case wer e uni nt e nt i ona l l y di s t or t ed by t he s t a f f
to achi eve cons i s t ency wi t h a popul a r t he or y of t he d y n a mi c s of a
s chi zophr eni c r eact i on ( 1 5 ) . No t h i n g of a n a mbi va l e ne na t ur e h a d
been descr i bed i n r el at i ons wi t h pa r e nt s , spouse, or f r i ends. To t he
ext ent t h a t ambi vl ance coul d be i nf er r ed, i t was pr oba bl y not gr eat er
t h a n is f ound i n all h u ma n r el at i ons hi ps . I t is t r ue t he pseudopa-
t i ent ' s r el at i ons hi ps wi t h hi s pa r e nt s c ha nge d over t i me , but i n t he
or di nar y cont ext t ha t woul d h a r d l y be r e ma r ka bl e - - i nde e d, it mi ght
ver y wel l be expect ed. Cl earl y, t he me a n i n g ascr i bed to hi s verbal i za-
t i ons ( t hat is, ambi val ence, af f ect i ve i nst abi l i t y) was de t e r mi ne d by
t he di agnos i s : schi zophr eni a. An ent i r el y di f f er ent me a n i n g woul d
ha ve been ascr i bed i f it were k n o wn t h a t t he ma n was " nor ma l . "
Al l ps eudopat i ent s t ook ext ens i ve not es publ i cl y. Un d e r or di na r y
ci r cums t ances , such behavi or woul d ha ve r ai sed ques t i ons i n t he
mi nds of observers, as, i n fact , i t di d a mo n g pat i ent s . I ndeed, i t
seemed so cer t ai n t ha t t he not es woul d el i ci t s us pi ci on t h a t el abor at e
pr ecaut i ons were t a ke n to r emove t h e m f r om t he wa r d each day. But
t he pr ecaut i ons pr oved needl ess. The cl osest a ny s t a f f me mb e r came
to que s t i oni ng t hes e not es occur r ed whe n one ps e udopa t i e nt as ked
hi s phys i c i a n wh a t ki nd of me di c a t i on he was r ecei vi ng a n d be ga n to
wr i t e down t he response. "You ne e dn' t wr i t e i t , " he wa s t ol d gent l y.
" I f you ha ve t r oubl e r emember i ng, j us t as k me agai n. "
I f no ques t i ons were as ked of t he ps eudopat i ent s , h o w was t hei r
wr i t i ng i nt er pr et ed? Nur s i ng r ecor ds for t hr ee pa t i e nt s i ndi c a t e t h a t
t he wr i t i ng was seen as a n as pect of t hei r pat hol ogi cal behavi or .
" Pa t i e nt e nga ge s i n wr i t i ng behavi or " was t he dai l y n u r s i n g com-
me n t on one of t he ps eudopat i ent s who was never que s t i one d about
hi s wr i t i ng. Gi ven t ha t t he pa t i e nt is i n t he hospi t al , he mu s t be psy-
chol ogi cal l y di st ur bed. And gi ven t h a t he is di st ur bed, c ont i nuous
wr i t i ng mu s t be a behavi or al ma ni f e s t a t i on of t ha t di s t ur bance, per-
h a p s a s ubs et of t he compul si ve behavi or s t h a t are s ome t i me s corre-
l at ed wi t h schi zophr eni a.
One t aci t char act er i st i c of ps yc hi a t r i c di agnos i s i s t h a t i t l ocat es
t he sources of aber r at i on wi t hi n t he i ndi vi dua l a nd onl y r ar el y wi t hi n
t he compl ex of st i mul i t ha t s ur r ounds hi m. Cons equent l y, behavi or s
t h a t ar e s t i mul at ed by t he e n v i r o n me n t ar e c ommonl y mi s a t t r i but e d
D. L. ROSENHAN 245
to the patient' s disorder. For example, one kindly nurse found a pseu-
dopatient pacing the long hospital corridors. "Nervous, Mr. X?" she
asked. "No, bored," he said.
The notes kept by pseudopatients are full of patient behaviors
that were misinterpreted by well-intentioned staff. Often enough, a
patient would go "berserk" because he had, wittingly or unwittingly,
been mistreated by, say, an attendant. A nurse coming upon the
scene would rarely inquire even cursorily into the environmental
stimuli of the patient' s behavior. Rather, she assumed t hat his upset
derived from his pathology, not from his present interactions with
other staff members. Occasionally, the staff might assume that the
patient' s family (especially when they had recently visited) or other
patients had stimulated the outburst. But never were the staff found
to assume t hat one of themselves or the structure of the hospital had
anything to do with a patient' s behavior. One psychiatrist pointed to
a group of patients who were sitting outside the cafeteria entrance
half an hour before lunchtime. To a group of young residents he in-
dicated t hat such behavior was characteristic of the oral-acquisitive
nature of the syndrome. It seemed not to occur to him t hat there were
very few things to anticipate in a psychiatric hospital besides eating.
A psychiatric label has a life and an influence of its own. Once
the impression has been formed t hat the patient is schizophrenic, the
expectation is that he will continue to be schizophrenic. When a suffi-
cient amount of time has passed, during which the patient has done
nothing bizarre, he is considered to be in remission and available for
discharge. But the label endures beyond discharge, with the uncon-
firmed expectation that he will behave as a schizophrenic again.
Such labels, conferred by mental health professionals, are as influen-
tial on the patient as they are on his relatives and friends, and it
should not surprise anyone that the diagnosis acts on all of them as a
self-fulfilling prophecy. Eventually, the patient himself accepts the
diagnosis, with all of its surplus meani ngs and expectations, and
behaves accordingly ( 5 ) .
The inferences to be made from these matters are quite simple.
Much as Zigler and Phillips have demonstrated t hat there is enor-
mous overlap in the symptoms presented by patients who have been
variously diagnosed ( 1 6 ) , so there is enormous overlap in the behav-
iors of the sane and the insane. The sane are not "sane" all of the
time. We lose our tempers "for no good reason." We are occasionally
depressed or anxious, again for no good reason. And we may find it
difficult to get along with one or another person--again for no reason
that we can specify. Similarly, the i nsane are not always insane. In-
deed, it was the impression of the pseudopatients while living with
them t hat they were sane for long periods of t i me--t hat the bizarre
246 CIANICAL SOCIAL WORK JOURNAL
behavi or s upon whi ch t hei r di agnoses wer e al l egedl y pr edi cat ed con-
st i t ut ed onl y a smal l fract i on of t hei r t ot al behavi or. I f it makes no
sense to l abel oursel ves per manent l y depressed on t he basi s of an
occasi onal depressi on, t hen it t akes bet t er evi dence t h a n is pr esent l y
avai l abl e to l abel all pat i ent s i ns a ne or schi zophr eni c on t he basi s of
bi zarre behavi or s or cogni t i ons. I t seems mor e useful, as Mi schel (17)
has poi nt ed out, to l i mi t our di scussi ons t o behaviors, t he st i mul i t hat
provoke t hem, and t hei r correlates.
I t is not known why powerful i mpr essi ons of per s onal i t y t rai t s,
such as "cr azy" or "i nsane, " ari se. Concei vabl y, whe n t he ori gi ns of
and st i mul i t hat gi ve ri se to a behavi or ar e r emot e or unknown, or
when t he behavi or st ri kes us as i mmut abl e, t r ai t l abel s r egar di ng t he
behaver arise. When, on t he ot her hand, t he ori gi ns and st i mul i are
known and avai l abl e, di scourse is l i mi t ed to t he behavi or itself. Thus,
I ma y hal l uci nat e because I am sl eepi ng, or I ma y hal l uci nat e be-
cause I have i ngest ed a pecul i ar drug. These are t er med sl eep-i nduced
hal l uci nat i ons, or dr eams, and dr ug- i nduced hal l uci nat i ons , respec-
tively. But when t he st i mul i to my hal l uci nat i ons ar e unknown, t hat
is called crazi ness, or s chi zophr eni a- - as i f t hat i nf er ence were some-
how as i l l umi nat i ng as t he others.
The Experience of Psychiatric Hospi t al i zat i on
The t er m " ment al i l l ness" is of r ecent origin. I t was coi ned by
people who wer e huma ne i n t hei r i ncl i nat i ons and who want ed very
much to r ai se t he st at i on of ( and t he publ i c' s s ympat hi es t owar d) t he
psychol ogi cal l y di st urbed from t ha t of wi t ches and "cr azi es" to one
t hat was aki n to t he physi cal l y ill. And t hey were at l east par t i al l y
successful, for t he t r eat ment of t he me nt a l l y ill has i mpr oved con-
si derabl y over t he years. But whi l e t r eat ment has i mpr oved, it is
doubt ful t ha t people r eal l y r egar d t he ment al l y ill i n t he s ame way
t hat t hey vi ew t he physi cal l y ill. A br oken leg is s ome t hi ng one recov-
ers from, but ment al i l l ness al l egedl y endur es forever (18). A broken
leg does not t hr eat en t he observer, but a cr azy schi zophr eni c? Ther e
is by now a host of evi dence t hat at t i t udes t owar d t he me nt a l l y ill ar e
char act er i zed by fear, host i l i t y, al oofness, suspi ci on, a nd dr ead (19).
The ment al l y ill are society' s lepers.
That such at t i t udes i nfect t he gener al popul at i on is per haps not
surpri si ng, onl y upset t i ng. But t ha t t hey affect t he pr of essi onal s- - at -
t endant s, nur ses, physi ci ans, psychol ogi st s, and soci al wor ker s - -
who t r eat a nd deal wi t h t he ment al l y ill is mor e di sconcer t i ng, bot h
because such at t i t udes are sel f-evi dent l y perni ci ous a nd because t hey
are unwi t t i ng. Most ment al heal t h professi onal s woul d i nsi st t hat
t hey are s ympat het i c t owar d t he ment al l y ill, t hat t hey ar e nei t her
D. L. ROSENHAN 247
avoidant nor hostile. But it is more likely t hat an exquisite ambiva-
lence characterizes their relations with psychiatric patients, such
that their avowed impulses are only part of their entire attitude. Neg-
ative attitudes are there too and can easily be detected. Such attitudes
should not surprise us. They are the natural offspring of the labels
patients wear and theplaces in which they are found.
Consider the structure of the typical psychiatric hospital. Staff
and patients are strictly segregated. Staff have their own living
space, including their dining facilities, bathrooms, and assembly
places. The glassed quarters t hat contain the professional staff,
which the pseudopatients came to call "the cage," sit out on every
dayroom. The staff emerge primarily for caretaking purposes--to
give medication, to conduct a therapy or group meeting, to instruct or
reprimand a patient. Otherwise, staff keep to themselves, almost as if
the disorder that afflicts their charges is somehow catching.
So much is patient-staff segregation the rule that, for four public
hospitals in which an attempt was made to measure the degree to
which staff and patients mingle, it was necessary to use "time out of
the staff cage" as the operational measure. While it was not t he case
that all time spent out of the cage was spent mingling with patients
(attendants, for example, would occasionally emerge to watch televi-
sion in the dayroom), it was the only way in which one could gather
reliable data on time for measuring.
The average amount of time spent by attendants outside of the
cage was 11.3 percent (range, 3 to 52 percent). This figure does not
represent only time spent mingling with patients but also includes
time spent on such chores as folding laundry, supervising patients
while they shave, directing ward cleanup, and sending patients to
off-ward activities. It was the relatively rare at t endant who spent
time talking with patients or playing games with them. It proved
impossible to obtain a "percent mingling time" for nurses, since the
amount of time they spent out of the cage was too brief. Rather, we
counted instances of emergence from the cage. On the average, day-
time nurses emerged from the cage 11.5 times per shift, including the
instances when they left the ward entirely (range, 4 to 39 times). Late
afternoon and night nurses were even less available, emerging on the
average 9.4 times per shift (range, 4 to 41 times). Data on early morn-
ing nurses, who arrived usually after midnight and departed at 8
a.m., are not available because patients were asleep during most of
this period.
Physicians, especially psychiatrists, were even less available.
They were rarely seen on the wards. Quite commonly, they would be
seen only when they arrived and departed, with the remaining time
being spent in their offices or in the cage. On the average, physicians
248 CLINICAL SOCIAL WORK JOURNAL
emerged on t he war d 6.7 t i mes per day (range, 1 to 17 times). It proved
difficult to make an accur at e est i mat e i n t hi s r egar d, si nce phys i ci ans
often mai nt ai ned hour s t hat al l owed t hem to come a nd go at di fferent
times.
The hi er ar chi cal or gani zat i on of t he ps ychi at r i c hospi t al has
been comment ed on before ( 2 0 ) , but t he l at ent me a n i n g of t ha t ki nd of
or gani zat i on is wor t h not i ng agai n. Those wi t h t he most power have
l east to do wi t h pat i ent s, and t hose wi t h t he l east power are most
i nvol ved wi t h t hem. Recall, however, t h a t t he acqui si t i on of role-ap-
propri at e behavi or s occurs ma i nl y t hr ough t he obser vat i on of others,
wi t h t he mos t powerful havi ng t he most i nfl uence. Consequent l y, it is
under s t andabl e t hat at t endant s not onl y spend mor e t i me wi t h pa-
t i ent s t ha n do any ot her member s of t he s t a f f mt ha t is requi red by
t hei r st at i on i n t he hi e r a r c hy- - but also, i nsof ar as t hey l ear n from
t hei r superi ors' behavi or, spend as little t i me wi t h pat i ent s as t hey
can. At t endant s are seen mai nl y i n t he cage, whi ch is wher e t he mod-
els, t he act i on, and t he power are.
I t ur n now to a di fferent set of st udi es, t hese deal i ng wi t h st af f
response to pat i ent -i ni t i at ed cont act . I t ha s l ong been known t hat t he
amount of t i me a person spends wi t h you can be a n i ndex of your
si gni fi cance to hi m. I f he i ni t i at es and ma i nt a i ns eye cont act , t her e is
r eason to bel i eve t hat he is consi der i ng your r equest s a nd needs. I f he
pauses to chat or act ual l y stops and t al ks, t her e is added r eason to
i nfer t hat he is i ndi vi duat i ng you. I n four hospi t al s, t he pseudopa-
t i ent appr oached t he st aff member wi t h a r equest whi ch t ook t he fol-
l owi ng form: " Par don me, Mr. [or Dr. or Mrs.] X, coul d you tell me
when I wi l l be eligible for gr ounds pri vi l eges?" (or " . . . when I will be
present ed at t he st af f meet i ng?" or " . . . . when I a m l i kel y to be dis-
charged?"). While t he cont ent of t he quest i on var i ed accor di ng to t he
appr opr i at eness of t he t ar get a nd t he pseudopat i ent ' s ( appar ent ) cur-
r ent needs t he form was al ways a court eous and r el evant request for
i nf or mat i on. Car e was t aken never to appr oach a par t i cul ar member
of t he st af f mor e t ha n once a day, l est t he st af f member become suspi-
cious or i rri t at ed. I n exami ni ng t hese dat a, r emember t ha t t he behav-
ior of t he pseudopat i ent s was nei t her bi zar r e nor di srupt i ve. One
could i ndeed engage i n good conver sat i on wi t h t hem.
The dat a for t hese exper i ment s ar e s hown i n Tabl e 1, separ at el y
for phys i ci ans (column 1) and for nur ses and a t t e nda nt s (col umn 2).
Mi nor di fferences bet ween t hese four i nst i t ut i ons wer e over whel med
by t he degree to whi ch st af f avoi ded cont i nui ng cont act s t hat pa-
t i ent s ha d i ni t i at ed. By far, t hei r most common r esponse consi st ed of
ei t her a br i ef r esponse to t he quest i on, offered whi l e t hey wer e "on t he
move" and wi t h head avert ed, or no r esponse at all.
The encount er frequent l y t ook t he fol l owi ng bi zar r e form: (pseu-
D. L. ROSENHAN 249
TABLE 1. Self-initiated Contact by Pseudopatients with Psychi at ri st s and Nurses
and Attendants, Compared to Contact with Other Groups
Univer-
sity cam-
Psychiatric pus (non-
hospitals medical)
Uni versi t y medical center
Physicians
Contact
( 2 )
Nurses
(1) and
Psychi- attend- (3)
atrists ant s Faculty
(4)
"Look- (5) (6)
ing "Look- No addi-
for a i ng tional
psychi- for an in- com-
atrist" ternist" ment
R e s p o n s e s
Moves on, head averted (%) 71 88 0 0 0 0
Makes eye contact (%) 23 10 0 11 0 0
Pauses and chats (%) 2 2 0 11 0 10
Stops and talks (%) 4 0.5 100 78 100 90
Mean number of questions
answered (out of 6) * * 6 3.8 4.8 4.5
Respondents (No.) 13 47 14 18 15 10
Attempts (No.) 185 1283 14 18 15 10
* Not applicable
dopat i ent ) " Pa r don me, Dr. X. Coul d you tell me wh e n I a m eligible for
gr ounds pri vi l eges?" (physi ci an) "Good mor ni ng, Dave. How are you
t oday?" (Moves off wi t hout wai t i ng for a response. )
I t is i nst r uct i ve to compar e t hese da t a wi t h dat a recent-
ly obt ai ned at St anf or d Uni ver si t y. I t ha s been al l eged t hat
l ar ge a nd emi nent uni versi t i es are char act er i zed by f acul t y who are
so busy t h a t t hey have no t i me for st udent s. For t hi s compar i son, a
young l ady appr oached i ndi vi dual f acul t y member s who seemed to be
wal ki ng purposeful l y to some meet i ng or t e a c hi ng e nga ge me nt and
asked t hem t he fol l owi ng six quest i ons.
1) " Pa r don me, could you di rect me to En c i n a Hal l ?" (at t he med-
ical school: " . . . to t he Cl i ni cal Resear ch Cent er?").
2) "Do you know wher e Fi sh Anne x i s?" (t here is no Fi s h Annex
at St anford).
3) "Do you t each here?"
4) "How does one appl y for admi s s i on to t he col l ege?" (at t he
medi cal school: " . . . to t he medi cal school?").
5) "Is i t di ffi cul t to get i n?"
6) "I s t her e f i nanci al ai d?"
Wi t hout exception, as can be seen i n Tabl e 1 ( col umn 3), all of t he
quest i ons wer e answer ed. No mat t er how r us hed t he y were, all re-
spondent s not onl y mai nt ai ned eye cont act , but st opped to talk. In-
250 CLINICAL SOCIAL WORK JOURNAL
deed, many of the respondents went out of their way to direct or take
the questioner to the office she was seeking, to try to locate "Fish
Annex," or to discuss with her the possibilities of being admitted to
the university.
Similar data, also shown in Table 1 (columns 4, 5, and 6), were
obtained in the hospital. Here too, the young lady came prepared with
six questions. After the first question, however, she remarked to18 of
her respondents (column 4), "I' m looking for a psychiatrist," and to
15 others (column 5), "I' m looking for an internist." Ten other respon-
dents received no inserted comment (column 6). The general degree of
cooperative responses is considerably higher for these university
groups t han it was for pseudopatients in psychiatric hospitals. Even
so, differences are apparent within the medical school setting. Once
having indicated that she was looking for a psychiatrist, the degree
of cooperation elicited was less t han when she sought an internist.
Powerlessness and Depersonalization
Eye contact and verbal contact reflect concern and individua-
tion; their absence, avoidance and depersonalization. The data I have
presented do not do justice to the rich daily encounters that grew up
around matters of depersonalization and avoidance. I have records of
patients who were beaten by staff for the sin of havi ng initiated ver-
bal contact. During my own experience, for example, one patient was
beaten in the presence of other patients for having approached an
attendant and told him, "I like you." Occasionally, punishment
meted out to patients for misdemeanors seemed so excessivethat it
could not be justified by the most radical interpretations of psychi-
atric canon. Nevertheless, they appeared to go unquestioned.Tempers
were often short. A patient who had not heard a call for medication
would be roundly excoriated, and the morning attendants would
often wake patients with, "Come on, you m---f--s, out of bed!"
Neither anecdotal nor "hard" data can convey the overwhelming
sense of powerlessness which invades the individual as he is con-
tinually exposed to the depersonalization of the psychiatric hospital.
It hardly matters which psychiatric hospital--the excellent public
ones and the very plush private hospital were better t han the rural
and shabby ones in this regard, but, again, the features that psychi-
atric hospitals had in common overwhelmed by far their apparent
differences.
Powerlessness was evident everywhere. The patient is deprived
of many of his legal rights by dint of his psychiatric commitment
(21). He is shorn of credibility by virtue of his psychiatric label. His
freedom of movement is restricted. He cannot initiate contact with
D. L. R OS E NHAN 251
t he staff, but ma y onl y r espond to such over t ur es as t hey make. Per-
sonal pr i vacy is mi ni mal . Pa t i e nt quar t er s and possessi ons can be
ent er ed a nd exami ned by any st af f member , for what ever r eason. His
per sonal hi s t or y and angui sh is avai l abl e to a ny st af f member (often
i ncl udi ng t he "gr ey l ady" a nd " c a n d y st r i per " vol unt eer) who
chooses to r ead hi s folder, r egar dl es s of t hei r t her apeut i c r el at i onshi p
to hi m. Hi s per sonal hygi ene and wa s t e evacuat i on ar e oft en mon-
itored. The wat er closets ma y have no doors.
At t i mes, deper sonal i zat i on r e a c he d such pr opor t i ons t hat pseu-
dopat i ent s h a d t he sense t hat t hey wer e i nvi si bl e, or at l east unwor t hy
of account . Upon bei ng admi t t ed, I a nd ot her ps eudopat i ent s took t he
i ni t i al phys i cal exami nat i ons i n a semi publ i c room, wher e st af f mem-
bers went about t hei r own busi ness as i f we wer e not t here.
On t he wa r d, at t endant s del i ver ed ver bal and occas i onal l y seri ous
physi cal abus e to pat i ent s in t he pr es ence of ot her obser vi ng pat i ent s,
some of whom (t he pseudopat i ent s) wer e wr i t i ng i t al l down.
Abusi ve behavi or , on t he ot her ha nd, t e r mi na t e d qui t e abr upt l y when
ot her st af f member s were known to be comi ng. St af f are credi bl e wit-
nesses. Pa t i e nt s ar e not.
A nur s e unbut t oned her uni f or m to adj ust her br assi er e i n t he
pr esence of a n ent i re war d of vi ewi ng men. One di d not have t he
sense t h a t she was bei ng seduct i ve. Rat her , she di dn' t not i ce us. A
group of st af f persons mi ght poi nt to a pat i ent i n t he dayr oom and
di scuss h i m ani mat edl y, as i f he wer e not t here.
One i l l umi nat i ng i nst ance of deper sonal i zat i on and i nvi si bi l i t y
occurred wi t h r egar d to medi cat i ons. Al l told, t he pseudopat i ent s
were admi ni s t er ed near l y 2100 pills, i ncl udi ng El avi l , St el azi ne, Com-
pazi ne, a nd Thor azi ne, to n a me but a few. ( That such a var i et y of
medi cat i ons shoul d have been admi ni s t er ed to pat i ent s pr es ent i ng
i dent i cal s ympt oms is i t sel f wor t hy of note. ) Onl y t wo wer e swal-
lowed. The r est were ei t her pocket ed or deposi t ed i n t he toilet. The
pseudopat i ent s were not al one i n t hi s. Al t hough I have no precise
records on how ma n y pat i ent s r ej ect ed t hei r medi cat i ons, t he pseu-
dopat i ent s f r equent l y found t he medi cat i ons of ot her pat i ent s i n t he
toilet before t hey deposited t hei r own. As l ong as t hey wer e coopera-
tive, t hei r behavi or and t he ps eudopat i ent s ' own i n t hi s mat t er , as i n
ot her i mpor t a nt mat t er s, went unnot i ced t hr oughout .
React i ons to such deper sonal i zat i on a mong pseudopat i ent s were
i nt ense. Al t hough t hey had come t o t he hospi t al as ~participant ob-
ser ver s a nd wer e ful l y awar e t ha t t he y di d not "bel ong, " t he y nevert he-
less f ound t hemsel ves caught up i n a nd f i ght i ng t he pr ocess of de-
per sonal i zat i on. Some exampl es: a gr a dua t e st udent i n psychol ogy
asked hi s wi fe to br i ng hi s t ext books to t he hospi t al so he could
"cat ch up on hi s homewor k" - - t hi s despi t e t he el abor at e pr ecaut i ons
252 CLINICAL SOCIAL WORK JOURNAL
t aken to conceal hi s pr of es s i onal as s oci at i on. The s a me s t udent , who
ha d t r ai ned for qui t e s ome t i me to get i nt o t he hos pi t al , a nd who ha d
l ooked f or war d to t he experi ence, " r e me mbe r e d" s ome dr a g races t ha t
he h a d wa nt e d to see on t he we e ke nd and i ns i s t ed t h a t he be dis-
char ged by t h a t t i me. An o t h e r ps e udopa t i e nt a t t e mpt e d a r omance
wi t h a nur s e. Subs equent l y, he i nf or me d t he s t af f t h a t he was appl y-
i ng for a dmi s s i on to gr a dua t e school i n ps ychol ogy a nd was very like-
ly to be admi t t ed, si nce a gr a dua t e pr of essor was one of hi s r egul ar
hos pi t al vi si t or s. The s a me per s on be ga n t o e nga ge i n ps yc hot he r a py
wi t h ot her pa t i e nt s - - a l l of t hi s as a wa y of be c omi ng a per son i n an
i mper s onal envi r onment .
The Sources of Depersonalization
Wha t ar e t he or i gi ns of deper s onal i zat i on? I h a v e al r eady men-
t i oned t wo. Fi r s t are at t i t udes hel d by all of us t owa r d t he ment al l y
i l l - - i ncl udi ng t hose who t r eat t h e m- - a t t i t u d e s char act er i zed by fear,
di st r ust , a n d hor r i bl e expect at i ons on t he one h a n d , a nd benevol ent
i nt ent i ons on t he ot her. Our a mbi va l e nc e l eads, i n t hi s i ns t a nc e as i n
ot hers, t o avoi dance.
Second, a nd not ent i r el y separ at e, t he hi e r a r c hi c a l st r uct ur e of
t he ps ychi at r i c hos pi t al f aci l i t at es deper s onal i zat i on. Thos e who are
at t he t op h a v e l east to do wi t h pat i ent s , a nd t hei r be ha vi or i nspi r es
t he r est of t he st aff. Aver age da i l yc ont a c t wi t h ps yc hi a t r i s t s , psychol -
ogi st s, r es i dent s , a nd phys i c i a ns c ombi ne d r a n g e d f r om 3.9 to 25.1
mi nut es, wi t h a n overal l me a n of 6.8 (six ps e udopa t i e nt s over a t ot al
of 129 da ys of hospi t al i zat i on) . I ncl uded i n t hi s a ve r a ge are t i me
s pent i n t he a dmi s s i ons i nt er vi ew, wa r d me e t i ngs i n t he pr esence of a
seni or s t a f f member , gr oup a nd i ndi vi dua l p s y c h o t h e r a p y cont act s,
case pr e s e nt a t i on conf er ences, a nd di s c ha r ge me e t i ngs . Cl earl y, pa-
t i ent s do n o t s pend mu c h t i me i n i nt e r pe r s ona l c ont a c t wi t h doct or al
st aff. And doct or al s t af f ser ve as model s f or nur s es a n d a t t e nda nt s .
Ther e ar e pr obabl y ot her sources. Ps yc hi a t r i c i ns t al l at i ons are
pr esent l y i n ser i ous f i nanci al st r ai t s. St a f f s hor t a ge s ar e per vasi ve,
s t af f t i me at a pr emi um. Some t hi ng h a s to give, a n d t h a t s ome t hi ng
is pa t i e nt cont act . Yet , whi l e f i na nc i a l s t r es s es ar e r eal i t i es, t oo muc h
can be ma d e of t hem. I ha ve t he i mpr e s s i on t h a t t he ps ychol ogi cal
forces t h a t r es ul t i n deper s onal i zat i on ar e mu c h st r onger , t h a n t he
fi scal ones a n d t h a t t he a ddi t i on of mor e s t a f f woul d not cor r espond-
i ngl y i mpr ove pa t i e nt Care i n t hi s r egar d. The i nci dence of s t af f
me e t i ngs a n d t he e nor mous a mo u n t o f r ecor d- keepi ng on pat i ent s , for
exampl e, h a v e not been as s ubs t a nt i a l l y r educed as h a s pa t i e nt con-
t act . Pr i or i t i es exi st , even dur i ng h a r d t i mes. Pa t i e n t c ont a c t is not a
s i gni f i cant pr i or i t y i n t he t r adi t i onal ps ychi at r i c hos pi t al , a nd fi scal
D. L. ROSENHAN 253
pressures do not account for this. Avoi dance and depersonalization
may.
Heavy reliance upon psychotropic medi cat i on tacitly contributes
to depersonalization by convi nci ng st af f t hat t r eat ment is indeed
being conduct ed and t hat furt her pat i ent cont act may not be neces-
sary. Even here, however, caution needs to be exercised in under-
st andi ng t he role of psychotropic drugs. If pat i ent s were powerful
rat her t han powerless, if t hey were vi ewed as i nt erest i ng individuals
rat her t han diagnostic entities, if t hey were socially si gni fi cant rath-
er t han social lepers, if their angui sh t rul y and wholly compelled our
sympat hi es and concerns, would we not seek cont act with them, de-
spite the availability of medications? Per haps for t he pleasure of it
all?
The Consequences of Labeling and Depersonalization
Whenever the ratio of what is known to what needs to be known
approaches zero, we tend to i nvent "knowl edge" and assume t hat we
under st and more t han we actually do. We seem unabl e to acknowl-
edge t hat we simply don' t know. The needs for di agnosi s and remedi-
ation of behavi oral and emotional problems are enormous. But rath-
er t han acknowl edge t hat we are j ust embar ki ng on underst andi ng,
we cont i nue to label patients "schi zophreni c, " "manic-depressive, "
and "i nsane, " as if in those words we had capt ured the essence of
underst andi ng. The facts of the mat t er are t hat we have known for a
long time t hat diagnoses are often not useful or reliable, but we have
nevert hel ess continued to use t hem. We now know t hat we cannot
di st i ngui sh i nsani t y from sanity. It is depressing to consider how
t hat i nformat i on will be used.
Not merel y depressing, but fri ght eni ng. How many people, one
wonders, are sane but not recognized as such in our psychi at ri c in-
stitutions? How many have been needl essl y stripped of their privi-
leges of citizenship, from the r i ght to vot e and drive to t hat of handl i ng
t hei r own accounts? How many have feigned i nsani t y in order
to avoid t he criminal consequences of t hei r behavior, and, conversely,
how ma ny would rat her st and trial t han live i nt er mi nabl y in a psy-
chiatric hospi t al --but are wrongl y t hought to be ment al l y ill? How
many have been stigmatized by well-intentioned, but nevertheless
erroneous, diagnoses? On the l ast point, recall agai n t hat a "type 2
error" in psychi at ri c diagnosis does not have t he same consequences
it does in medical diagnosis. A di agnosi s of cancer t hat has been
found to be in error is cause for celebration. But psychi at ri c diagnoses
are rarel y found to be in error. The label sticks, a mar k of i nadequacy
forever.
254 CLINICAL SOCIAL WORK JOURNAL
Finally, how many patients mi ght be "sane" outside the psychi-
atric hospital but seem i nsane in i t unot because craziness resides in
them, as it were, but because t hey are respondi ng to a bizarre setting,
one t hat may be unique to i nst i t ut i ons which harbor net her people?
Goffman (4) calls the process of socialization to such institutions
"mor t i f i cat i on"uan apt met aphor t hat includes t he processes of de-
personalization t hat have been described here. And while it is impos-
sible to know whet her the pseudopatients' responses to these pro-
cesses are characteristic of all i nmat es- - t hey were, aft er all, not real
pat i ent s--i t is difficult to believe t hat these processes of socialization
to a psychi at ri c hospital provide useful attitudes or habi t s of response
for living in the "real world."
SUMMARY AND CONCLUSIONS
It is clear t hat we cannot di st i ngui sh the sane from the i nsane in
psychi at ri c hospitals. The hospital itself imposes a special environ-
ment in whi ch the meani ngs of behavi or can easily be misunder-
stood. The consequences to pat i ent s hospitalized in such an environ-
ment - - t he powerlessness, depersonalization, segregation, mortifica-
tion, and self-labeling--seem undoubt edl y countertherapeutic.
I do not, even now, underst and this problem well enough to per-
ceive solutions. But two mat t ers seem to have some promise. The first
concerns the proliferation of communi t y ment al heal t h facilities, of
crisis i nt ervent i on centers, of the human potential movement , and of
behavior therapies that, for all of t hei r own problems, t end to avoid
psychi at ri c labels, to focus on specific problems and behaviors, and
to ret ai n the individual in a relatively nonpej orat i ve environment.
Clearly, to the extent t hat we refrai n from sendi ng the distressed to
i nsane places, our impressions of t hem are less likely to be distorted.
(The risk of distorted perceptions, it seems to me, is al ways present,
since we are much more sensitive to an individual' s behaviors and
verbalizations t han we are to the subtle contextual stimuli t hat often
promote them. At issue here is a mat t er of magni t ude. And, as I have
shown, the magni t ude of distortion is exceedingly hi gh in the ex-
treme context t hat is a psychiatric hospital.)
The second mat t er t hat mi ght prove promi si ng speaks to the need
to increase the sensitivity of ment al heal t h workers and researchers
to the C a t c h 22 position of psychi at ri c patients. Si mpl y readi ng ma-
terials in this area will be of help to some such workers and research-
ers. For others, directly experiencing the i mpact of psychi at ri c hospi-
talization will be of enormous use. Clearly, furt her research into the
social psychol ogy of such total institutions will bot h facilitate treat-
ment and deepen underst andi ng.
D. L. ROSENHAN 255
I and t he other pseudopat i ent s i n t he ps ychi at ri c s et t i ng had dis-
ti nctl y negat i ve reactions. We do not pret end to descri be t he subjec-
ti ve experi ences of true pati ents. Thei rs ma y be di fferent from ours,
parti cul arl y wi t h the pas s age of t i me and the neces s ary process of
adapt at i on to one' s envi ronment . But we can and do speak to the
rel ati vel y more objecti ve i ndi ces of t reat ment wi t hi n t he hospi t al . It
coul d be a mi st ake, and a very unf ort unat e one, to consi der t hat what
happened to us derived from mal i ce or st upi di t y on t he part of the
staff. Qui te the contrary, our ove r whe l mi ng i mpres s i on of t hem was
of peopl e who real l y cared, who were commi t t ed and who were un-
commonl y i ntel l i gent. Where t hey fai l ed, as t hey s ome t i me s did pain-
fully, it woul d be more accurate to attri bute t hos e f ai l ures to the
envi ronment i n whi ch they, too, f ound t he ms e l ve s t han to personal
cal l ousness. Thei r percepti ons and behavi or were control l ed by the
si t uat i on, rather t han bei ng mot i vat ed by a mal i ci ous di sposi t i on. In
a more beni gn envi ronment , one t hat wa s l ess at t ached to gl obal
di agnosi s, thei r behavi ors and j udgme nt s mi ght have been more be-
ni gn and effective.
REFERENCES AND NOTES
1. P. Ash, J. Abnorm. Soc. Psychot. 44, 272 (1949); A. T. Beck, Amer. J. Psychiat.
119, 210 (1962); A. T. Boisen, Psychiatry 2, 233 (1938); N. Kr ei t man, J. Ment. ScL
107, 876 {1961); N. Kr ei t man, P. Sai ns bur y, J. Morri sey, J. Towers, J. Scri vener,
ibid., p. 887; H. O. Schmi t t and C. P. Fonda, J. Abnorm. Soc. Psychol. 52, 262
(1956); W. Seeman, J. Nerv. Ment. Dis. 118, 541 (1953). For an a na l ys i s of t hese
ar t i f act s a nd summar i es of t he di sput es, see J. Zubi n, Annu. Rev. Psychol. 18, 373
(1967); L. Phi l l i ps and J. G. Dr aguns, ibid. 22, 447 (1971).
2. R. Benedi ct , J. Gen. Psychol. 10, 59 (1934).
3. See i n t hi s r egar d H. Becker, Outsiders: Studies in the Sociology of Deviance (Free
Press, New York, 1963); B. M. Br agi ns ky, D. D. Br agi ns ky, K. Ri ng, Methods of
Madness: The Mental Hospital as a Last Resort (Holt, Ri nehar t & Wi nst on, New
York, 1969); G. M. Crocet t i and P. V. Lemkau, Amer. Sociol. Rev. 30, 577 (1965); E.
Goffman, Behavior in Public Places (Free Pr ess, New York, 1965); R. D. Lai ng, The
Divided Self: A Study of Sanity and Madness ( Quadr angel , Chi cago, 1960); D. L.
Phi l l i ps, Amer. Sociol. Rev. 28, 963 (1963); T. R. Sar bi n, Psychol. Today 6, 18
(1972); E. Schur , Amer. J. Sociol. 75, 309 (1969); T. Szasz, Law, Liberty and Psychi-
atry ( Macmi l l an, New York, 1963); The Myt h of Mental Illness: Foundations of a
Theory of Mental Illness (Hoeber Har per , New York, 1963). For a cri t i que of some
of t hese vi ews, see W. R. Gove, Amer. Sociol. Rev. 35, 873 (1970).
4. E. Goffi nan, Asylums (Doubleday, Gar den Ci t y, N.Y., 1961).
5. T. J . Scheff, Being Mentally Ill: A Sociological Theory (Al di ne, Chi cago, 1966).
6. Dat a from a ni nt h pseudopat i ent ar e not i ncor por at ed i n t hi s r epor t because, al-
t hough hi s s a ni t y went undet ect ed, he f al si f i ed aspect s of hi s per s onal hi st ory,
i ncl udi ng hi s ma r i t a l st at us and pa r e nt a l r el at i ons hi ps . Hi s exper i ment al behav-
i ors t her ef or e were not i dent i cal to t hose of t he ot her pseudopat i ent s.
7. A. Bar r y, Bellevue Is a State of Mind ( Har cour t Brace J ovanovi ch, New York,
1971); I. Bel knap, Human Problems of a State Mental Hospital ( McGr aw Hill, New
York, 1956); W. Caudi l l , F. C. Redlich, H. R. Gi l more, E. B. Brody, Amer. J. Or-
thopsychiat. 22, 314 (1952); A. R. Gol dman, R. H. Bohr, T. A. St ei nber g, Prof. Psy-
chol. 2, 427 (1970); unaut hored, Roche Report 1 (No. 13), 8 (1971).
8. Beyond t he per sonal di ffi cul t i es t ha t t he ps eudopat i ent i s l i kel y to experi ence in
256 CLINICAL SOCIAL WORK JOURNAL
the hospital, there are legal and social ones that, combined, require considerable
attention before entry. For example, once admitted to a psychiatric institution, it is
difficult, if not impossible, to be discharged on short notice, state law to the con-
trary notwithstanding. I was not sensitive to these difficulties at the outset of the
project, nor to the personal and situational emergencies t hat 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.
9. However distasteful such concealment is, it was a necessary first step to examin-
ing these questions. Without concealment, there would have been no way to know
how valid these experiences were; nor was there any way of knowing whether
whatever detections occurred were a tribute to the diagnostic acumen of the staff or
to the hospital' s rumor network. Obviously, since my concerns are general ones
that cut across individual hospitals and staff, I have respected their anonymity
and have eliminated clues that might lead to t hei r identification.
10. Interestingly, of the 12 admissions, 11 were diagnosed as schizophrenic and one,
with the identical symptornatology, as manic-depressive psychosis. This diagnosis
has a more favorable prognosis, and it was given by the only private hospital in
our sample. On the relations between social class and psychiatric diagnosis, see A.
deB. Hollingshead and F. C. Redlich, Social Class and Mental Illness: A Commu-
nity Study (Wiley, New York, 1958).
11. It is possible, of course, that patients have quite broad latitudes in diagnosis and
therefore are inclined to call many people sane, even those whose behavior is pat-
ently aberrant. However, although we have no hard dat a on this matter, it was our
distinct impression that this was not the case. In many instances, patients not
only singled us out for attention, but came to imitate our behaviors and styles.
12. J. Cummi ng and E. Cumming, Community Ment. Health 1, 135 (1965); A. Fari na
and K. Ring, J. Abnorm. Psychol. 70, 47 (1965); H. E. Freeman and 0. G.Sim-
mons, The MentalPatient Comes Home (Wiley, New York, 1963); W. J. Johannsen,
Ment. Hygiene 53,218 (1969); A. S. Linsky, Soc. Psychiat. 6, 166 (1970).
13. S. E. Asch, J. Abnorm. Soc. Psychol. 41, 258 (1946); Social Psychology (Prentice-
Hall, New York, 1952).
14. See also I. N. Mensh and J. Wishner, J. Personality 16, 188 (1947); J. Wishner,
Psychol. Reu. 67, 96 (1960); J. S. Bruner and R. Tagiuri, in Handbook of Social
Psychology, G. Lindzey, Ed. (Addison-Wesley, Cambridge, Mass., 1954), vol. 2, pp.
634-654; J. S. Bruner, D. Shapiro, R. Tagiuri, in Person Perception and Interperson-
al Behavior, R. Tagiuri and L. Petrullo, Eds. (Stanford Univ. Press, Stanford,
Calif., 1968), pp. 277-288.
15. For an example of a similar self-fulfilling prophecy, in this instance dealing with
the "cent ral " trait of intelligence, see R. Rosenthal and L. Jacobson, Pygmalion in
the Classroom (Holt, Rinehart & Winston, New York, 1968).
16. E. Zigler and L. Phillips, J. Abnorm. Soc. Psychol. 63, 69 (1961). See also R. K.
Freudenberg and J. P. Robertson, A.M.A. Arch. Neurol. Psychiatr. 76, 14 (1956).
17. W. Mischel, Personality and Assessment (Wiley, New York, 1968).
18. The most recent and unfortunate instance of this tenet is t hat of Senator Thomas
Eagleton.
19. T. R. Sarbin and J. C. Mancuso, J. Clin. Consult. Psychol. 35, 159 (1970); T. R.
Sorbin, ibid 31, 447 (1967); J. C. Nunnal l y, Jr., Popular Conceptions of Mental
Health (Holt, Rinehart & Winston, New York, 1961).
20. A. H. St ant on and M. S. Schwartz, The Mental Hospital: A Study of Institutional
Participation in Psychiatric Illness and Treatment (Basic, New York, 1954).
21. D.B. Wexler and S. E. Scoville, Ariz. Law Rev. 13, 1 (1971).
22. I thank W. Mischel, E. Orne, and M. S. Rosenhan for comments on an earlier draft
of this manuscript.
Department of Psychology
Stanford University
Stanford, California 94305
Clinical Social Work Journal
Vol. 2, No. 4,1974