in Insane Places - University of...

9
The Geographical Ditribution of Animas (Wiley, New York, 1957); B. Rensch, Evolu- tion Above the Specie Level (Methuen, London, 1959); V. Grant, The Origin of Adaptations (Columbia Univ. Press, New York, 1963). 5. S. Wright, Genetics 16, 97 (1931). 6.- , Ibid. 28, 114 (1943); Ibid. 31. 39 (1946); Evolution and the Genetics ot Popu- lations, vol. 2, The Theory of Gene Fre- quencies (Univ. of Chicago Press, Chicago, 1969); F. 3. Rohlf and 0. D. Schnell, Amer. Natur. 105, 295 (1971). 7. 3. B. S. Haldane, J. Genet. 48, 277 (1948). 8. R. A. Fisher, Biometrics 6, 353 (1950); M. Kimura, Annu. Rep. Nat. Inst. Genet. Mishima-City, Japan 9, 84 (1958). 9. M. Kimura and G. H. Weiss, Genetics 49, 561 (1964); M. Kimura and T. Maruyama, Genet. Res. 18, 125 (1971). 10. P. R. Ehrlich and P. H. Raven, Science 165, 1228 (1969). 11. For example, J. Maynard-Smith, Amer. Natur. 100, 637 (1966). 12. J. M. Thoday, Nature 181, 1124 (1958); and T. B. Boam, Heredity 13, 204 (1959); E. Millicent and J. M. Thoday, Ibid. 16, 219 (1961); M. Thoday and J. B. Gibson, Amer. Natur. 105, 86 (1971). .13. F. A. Streams and D. Pimentel, Ibid. 9S, 201 (1961); Th. Dobzhansky and B. Spassky, Proc. Roy. Soc. London Ser. B. 168, 27 (1967); , J. Sved, Ibid. 173, 191 (1969); Th. Dobzhansky, H. Levene, B. Spassky, ibid. 18W, 21 (1972). 14. M. Slatkin, thesis, Harvard University (1971). 15. S. K. Jain and A. D. Bradshaw, Heredity 21, 407 (1966). 16. Parapatric divergence is divergence between adjacent but genetically continuous popula- tions. See H. M. Smith, Syst. Zool. 14, 57 (1965); ibid. 18, 254 (1969); M. 3. D. White, R. E. Blacidth, R. M. Blaclidth, J. Cheney, Aust. J. Zool. 15, 263 (1967); M. J. D. White, Science 159, 1065 (1968); K. H. L. Key, Syst. Zool. 17, 14 (1968). 17. J. S. Huxley, Nature 142, 219 (1938); BlJdr. Dierk. Leiden 27, 491 (1939). 18. F. B. Sumner, BIbliogr. Genet. 9, 1 (1932). 19. F. Salomonsen, Dan. Blol. Medd. 22, 1 (1955). 20. E. B. Ford, Blol. Rev. Cambridge Phil. Soc. 20, 73 (1945). 21. Examples of morph-ratio clines include: H. B. D. Kettlewell and R. J. Berry, Heredity 16, 403 (1961); Ibid. 24, 1 (1969); H. B. D. Kettlewell, R. J. Berry. C. J. Cadbury, G. C. Phillips, Ibid., p. 15; H. N. Southern, J. Zool. London Ser. A 138, 455 (1966); A. J. Cain and 3. D. Currey. Phil. Trans. Roy. Soc. London Ser. D. 246, 1 (1962); A. P. Plant and L P. Brower, Evolu- tion t2, 699 (1968); 0. Halkka and E. Mikkola, Ierditas 54, 140 (1965); B. C. Clarke, in Evolution and Environment, E. T. Drake, Ed. (Yale Univ. Press, New Haven, 1968), p. 351; B. C. Clarke and J. J. Murray, in Ecological Genetics and Evolution, R. Greed, Ed. (Blackwells, Oxford, 1971), p. 51; J. A. Bishop and P. S. Harper, Heredity 25, 449 (1969); J. A. Bishop, J. Anim. Ecol. 41, 209 (1972); G. Hewitt and F. M. Brown, Heredity 25, 365 (1970); G. Hewitt and C. Ruscoe, 1. Anim. Ecol. 48, 753 (1971); H. Wolda, Ibid. 38, 623 (1969); F. B. LIving- stone, Amer. J. Phys. Anthropol. 31, 1 (1969). 22. C. P. Hasklns, E. F. Haskins, J. J. A. McLaughlan, R. E. Hewitt, in Vertebrate Speclatlon, W. F. Blair, Ed. (Univ. of Texas Press, Austin, 1961), p. 320. 23. A. 3. Bateman, Heredity 1, 234, 303 (1947); Ibid. 4, 353 (1950); R. N. Colwell, Amer. I. Bot. 38, 511 (1951); M. R. Roberts and H. Lewis, Evolution 9, 445 (1955); C. P. Haskins, personal commnunication; K. P. Lamb, E. Hassan, D. P. Sooter, Ecology 52, 178 (1971). For localized distribution and problem of establishment see also: W. F. Blair, Ann. N.Y. Acad. Sc. 44, 179 (1943); Evolution 4, 253 (1950); Lf R. Dice, Amer. Natut. 74, 289 (1940); P. Labine, Evolution 20, 580 (1966); H. Lewis, Ibid. 7, 1 (1953); W. Z. Lidicker, personal communication; J. T. Marshall, Jr., Condor 50, 193, 233 (1948); R. K. Sealander, Amer. Zool. 10, 53 (1970); P. Voiplo, Ann. Zool. Fenn. 15, 1 (1952); P. K. Anderson, Science 145, 177 (1964). 24. N. W. rumofeeff-Ressovsky, In The New Systematics, 3. S. Huxley, Ed. (Oxford Univ. Press, Oxford, 1940), p. 73. 25. The null point is the position at which selection changes over from favoring one type to favoring another. 26. 3. A. Endler, in preparation. 27. L. M. Cook, Coefficients of Natural Selection (Hutchinson Univ. Library, Biological Sci- ences No. 153, London, 1971); P. B. Lving- stone, Amer. J. Phys. Anthropol. 31, 1 (1969). 28. W. C. Allee, A. E. Emerson, 0. Park, T. Park, K. P. Schmidt, Prncdpks of Animal Ecology (Saunders, Philadelphia, 1949); H. C. Andrewartha and L. C. Birch, The Distrib- tion and Abundance of Animal (Univ. of Chicago Press, Chicago, 1954); G. L. Clarke, Elements of Ecology (Wiley, New York, 1954); R. Geiger, The Climate Near the Ground (transation, Harvard Univ. Press, Cambridge, 1966). 29. Results for autosomal and sex-lnked systenu do not differ for the models to be discussed, except that, for a given amount of selection, the sex-linked system is l sensitive to the effect of gene flow. 'This Is because the effective selection on males In sex-inked lod makes the net selection sronger, com- pared to autosomal loci, for the population as a whole. See C. C. Li, Populion Genedes (Univ. of Chicago Press, Chicago, 1955) for a good dion of sex-linkage and selection. 30. The equilibrium configurations are not sig- nificandy altered if the emigrants from the end demes -do not retur, unless the number of demes (d) very small (3. A. Endler, unpublished data). 31. See, for example, the models of B. C. Clarke [Amer. Natr. 1W1, 389 (1966)1 and those in (14). 32. This model incorporates Clarke's model of frequency-dependence; see B. C. Clarke, Evolution 18, 364 (1964). 33. R. A. Fisher and F. Yates, Statsteal Tables for Biologkia, Agricultural, ad Medical Re- search (Oliver & Boyd, Edinburgh, 1948); R. R. Sokcal and F. J. Rohlf, Biometry (Freeman, San Francisco, 1969). 34. See, for example, C. G. Johnson, Migration ad Dispersal of Inects by FUght (Methuen, London, 1969); 3. Antonovics, Amer. Sd. 99, 593 (1971). 35. E. C. Pielou, An Introduction to Mathematical Ecology (Wiey-Interscience, New York, 1969). 36. W. F. Blair, Contrtb. Lab. Vertebrate Btol. Univ. Mich. No. 36, 1 (1947). 37. P. A. Parsons, Genetica 33, 184 (1963). 38. G. Hewitt and B. John. Chromosome 21, 140 (1967); Evolution 24, 169 (1970); G. Hewitt, personal communication; H. Wolda, J. Anim. Ecol. 38, 305, 623 (1969). 39. L. R. Dice, Contrib. Lab. Vertebrate Genet. Univ. Mich. No. 8 (1939), p. 1; ibid. No. 15 (1941), p. 1. 40. 1. C. J. Galbraith, Bull. Brit. Mus. Natur. Hist. Zoolk 4, 133 (1956). 41. I am grateful to the National Science Founda- tion for a graduate fellowship In support of this study. I thank Prof. Alan Robertson and the Institute of Animal Genetics, Uni- versity of Edinburgh, for the Drosophila, and for klindly providing me with fresh medium throughout the study. Criticism of the manu- script by Professors John Bonner and Jane Potter, Dr. Philip Ashmole, Peter Tuft, Dr. David Noakes, Dr. John Godfrey, Dr. Caryl P. Haskins, and M. C. Bathgate was very welcome. In particular, I thank my supervisor, Professor Bryan C. Clarke, for help and criti- cism throughout this study. Any errors or omissions are entirely my own. I thank the Edinburgh Regional Computing Center and the Edinburgh University Zoology Department for generous computer tiWm allowanoes. I will supply the specially wrtten IMP language program upon request. On Being Sane in Insane 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 that we can tell the normal from the abnormal, the evidence is simply not compelling. It is commonplace, for example, to read about murder trials wherein eminent psychiatrists for the defense are con- 250 tradicted by equally eminent psychia- trists for the prosecution on the matter of the defendant's sanity. More gen- erally, there are a great deal of conflict- ing data on the reliability, utility, and meaning of such terms as "sanity," "in- sanity," "mental illness," and "schizo- phrenia" (1). Finally, as early as 1934, Benedict suggested that normality and abnormality are not universal (2). What is viewed as normal in one cul- ture may be seen as quite aberrant in another. Thus, notions of normality and abnormality may not be quite-as accu- rate as people believe they are. To raise questions regarding normal- ity and abnormality is in no way to question the fact that some behaviors are deviant or odd. Murder is deviant. So, too, are hallucinatio,ns. Nor does raising such questions deny the exis- tence of the personal anguish that is often associated with "mental illness." Anxiety and depression exist. Psycho- logical sufferng' exists. But normality and abnormality, sanity and insanity, and the diagnoses that flow from them Th author is professor of psychoWgy and law at Stanford University, Stanford, Caifornis 9430S. Portions of thee data were presented to collo- quiums of the psychology departents at the University of Cafomra at Berkeley and at Santa Barbar; University of Arizona, Tucson; and Harvard Univwrsity, Cambridge, Massachusetts. SCIENCE6 VOL. 179

Transcript of in Insane Places - University of...

The Geographical Ditribution of Animas(Wiley, New York, 1957); B. Rensch, Evolu-tion Above the Specie Level (Methuen,London, 1959); V. Grant, The Origin ofAdaptations (Columbia Univ. Press, NewYork, 1963).

5. S. Wright, Genetics 16, 97 (1931).6.- , Ibid. 28, 114 (1943); Ibid. 31. 39

(1946); Evolution and the Genetics ot Popu-lations, vol. 2, The Theory of Gene Fre-quencies (Univ. of Chicago Press, Chicago,1969); F. 3. Rohlf and 0. D. Schnell, Amer.Natur. 105, 295 (1971).

7. 3. B. S. Haldane, J. Genet. 48, 277 (1948).8. R. A. Fisher, Biometrics 6, 353 (1950); M.

Kimura, Annu. Rep. Nat. Inst. Genet.Mishima-City, Japan 9, 84 (1958).

9. M. Kimura and G. H. Weiss, Genetics 49,561 (1964); M. Kimura and T. Maruyama,Genet. Res. 18, 125 (1971).

10. P. R. Ehrlich and P. H. Raven, Science 165,1228 (1969).

11. For example, J. Maynard-Smith, Amer. Natur.100, 637 (1966).

12. J. M. Thoday, Nature 181, 1124 (1958);and T. B. Boam, Heredity 13, 204 (1959); E.Millicent and J. M. Thoday, Ibid. 16, 219(1961); M. Thoday and J. B. Gibson, Amer.Natur. 105, 86 (1971).

.13. F. A. Streams and D. Pimentel, Ibid. 9S, 201(1961); Th. Dobzhansky and B. Spassky, Proc.Roy. Soc. London Ser. B. 168, 27 (1967);

, J. Sved, Ibid. 173, 191 (1969); Th.

Dobzhansky, H. Levene, B. Spassky, ibid. 18W,21 (1972).

14. M. Slatkin, thesis, Harvard University (1971).15. S. K. Jain and A. D. Bradshaw, Heredity

21, 407 (1966).16. Parapatric divergence is divergence between

adjacent but genetically continuous popula-tions. See H. M. Smith, Syst. Zool. 14, 57(1965); ibid. 18, 254 (1969); M. 3. D. White,R. E. Blacidth, R. M. Blaclidth, J. Cheney,Aust. J. Zool. 15, 263 (1967); M. J. D. White,Science 159, 1065 (1968); K. H. L. Key,Syst. Zool. 17, 14 (1968).

17. J. S. Huxley, Nature 142, 219 (1938); BlJdr.Dierk. Leiden 27, 491 (1939).

18. F. B. Sumner, BIbliogr. Genet. 9, 1 (1932).19. F. Salomonsen, Dan. Blol. Medd. 22, 1

(1955).20. E. B. Ford, Blol. Rev. Cambridge Phil. Soc.

20, 73 (1945).21. Examples of morph-ratio clines include:

H. B. D. Kettlewell and R. J. Berry, Heredity16, 403 (1961); Ibid. 24, 1 (1969); H. B. D.Kettlewell, R. J. Berry. C. J. Cadbury,G. C. Phillips, Ibid., p. 15; H. N. Southern,J. Zool. London Ser. A 138, 455 (1966);A. J. Cain and 3. D. Currey. Phil. Trans.Roy. Soc. London Ser. D. 246, 1 (1962);

A. P. Plant and L P. Brower, Evolu-tion t2, 699 (1968); 0. Halkka and E.Mikkola, Ierditas 54, 140 (1965); B. C.Clarke, in Evolution and Environment, E. T.Drake, Ed. (Yale Univ. Press, New Haven,1968), p. 351; B. C. Clarke and J. J. Murray,in Ecological Genetics and Evolution, R.Greed, Ed. (Blackwells, Oxford, 1971), p.51; J. A. Bishop and P. S. Harper, Heredity25, 449 (1969); J. A. Bishop, J. Anim. Ecol.41, 209 (1972); G. Hewitt and F. M. Brown,Heredity 25, 365 (1970); G. Hewitt and C.Ruscoe, 1. Anim. Ecol. 48, 753 (1971);H. Wolda, Ibid. 38, 623 (1969); F. B. LIving-stone, Amer. J. Phys. Anthropol. 31, 1 (1969).

22. C. P. Hasklns, E. F. Haskins, J. J. A.McLaughlan, R. E. Hewitt, in VertebrateSpeclatlon, W. F. Blair, Ed. (Univ. of TexasPress, Austin, 1961), p. 320.

23. A. 3. Bateman, Heredity 1, 234, 303 (1947);Ibid. 4, 353 (1950); R. N. Colwell, Amer. I.

Bot. 38, 511 (1951); M. R. Roberts and H.Lewis, Evolution 9, 445 (1955); C. P. Haskins,personal commnunication; K. P. Lamb, E.Hassan, D. P. Sooter, Ecology 52, 178 (1971).For localized distribution and problem ofestablishment see also: W. F. Blair, Ann.N.Y. Acad. Sc. 44, 179 (1943); Evolution 4,253 (1950); LfR. Dice, Amer. Natut. 74, 289(1940); P. Labine, Evolution 20, 580 (1966);H. Lewis, Ibid. 7, 1 (1953); W. Z. Lidicker,personal communication; J. T. Marshall, Jr.,Condor 50, 193, 233 (1948); R. K. Sealander,Amer. Zool. 10, 53 (1970); P. Voiplo, Ann.Zool. Fenn. 15, 1 (1952); P. K. Anderson,Science 145, 177 (1964).

24. N. W. rumofeeff-Ressovsky, In The NewSystematics, 3. S. Huxley, Ed. (Oxford Univ.Press, Oxford, 1940), p. 73.

25. The null point is the position at whichselection changes over from favoring onetype to favoring another.

26. 3. A. Endler, in preparation.27. L. M. Cook, Coefficients of Natural Selection

(Hutchinson Univ. Library, Biological Sci-ences No. 153, London, 1971); P. B. Lving-stone, Amer. J. Phys. Anthropol. 31, 1 (1969).

28. W. C. Allee, A. E. Emerson, 0. Park, T.Park, K. P. Schmidt, Prncdpks of AnimalEcology (Saunders, Philadelphia, 1949); H. C.Andrewartha and L. C. Birch, The Distrib-tion and Abundance of Animal (Univ. ofChicago Press, Chicago, 1954); G. L. Clarke,Elements of Ecology (Wiley, New York,1954); R. Geiger, The Climate Near theGround (transation, Harvard Univ. Press,Cambridge, 1966).

29. Results for autosomal and sex-lnked systenudo not differ for the models to be discussed,except that, for a given amount of selection,the sex-linked system is l sensitive to

the effect of gene flow. 'This Is because theeffective selection on males In sex-inked

lod makes the net selection sronger, com-pared to autosomal loci, for the populationas a whole. See C. C. Li, Populion Genedes(Univ. of Chicago Press, Chicago, 1955) fora good dion of sex-linkage and selection.

30. The equilibrium configurations are not sig-nificandy altered if the emigrants from theend demes -do not retur, unless the numberof demes (d) very small (3. A. Endler,unpublished data).

31. See, for example, the models of B. C.Clarke [Amer. Natr. 1W1, 389 (1966)1 andthose in (14).

32. This model incorporates Clarke's model offrequency-dependence; see B. C. Clarke,Evolution 18, 364 (1964).

33. R. A. Fisher and F. Yates, Statsteal Tablesfor Biologkia, Agricultural, ad Medical Re-search (Oliver & Boyd, Edinburgh, 1948);R. R. Sokcal and F. J. Rohlf, Biometry(Freeman, San Francisco, 1969).

34. See, for example, C. G. Johnson, Migrationad Dispersal of Inects by FUght (Methuen,London, 1969); 3. Antonovics, Amer. Sd. 99,

593 (1971).35. E. C. Pielou, An Introduction to Mathematical

Ecology (Wiey-Interscience, New York, 1969).36. W. F. Blair, Contrtb. Lab. Vertebrate Btol.

Univ. Mich. No. 36, 1 (1947).37. P. A. Parsons, Genetica 33, 184 (1963).38. G. Hewitt and B. John. Chromosome 21,

140 (1967); Evolution 24, 169 (1970); G.Hewitt, personal communication; H. Wolda,J. Anim. Ecol. 38, 305, 623 (1969).

39. L. R. Dice, Contrib. Lab. Vertebrate Genet.Univ. Mich. No. 8 (1939), p. 1; ibid. No. 15(1941), p. 1.

40. 1. C. J. Galbraith, Bull. Brit. Mus. Natur.Hist. Zoolk 4, 133 (1956).

41. I am grateful to the National Science Founda-tion for a graduate fellowship In supportof this study. I thank Prof. Alan Robertsonand the Institute of Animal Genetics, Uni-versity of Edinburgh, for the Drosophila, andfor klindly providing me with fresh mediumthroughout the study. Criticism of the manu-script by Professors John Bonner and JanePotter, Dr. Philip Ashmole, Peter Tuft, Dr.David Noakes, Dr. John Godfrey, Dr. CarylP. Haskins, and M. C. Bathgate was verywelcome. In particular, I thank my supervisor,Professor Bryan C. Clarke, for help and criti-cism throughout this study. Any errors oromissions are entirely my own. I thank theEdinburgh Regional Computing Center andthe Edinburgh University Zoology Departmentfor generous computer tiWm allowanoes. I willsupply the specially wrtten IMP language

program upon request.

On Being Sane in Insane Places

D. L. Rosenhan

If sanity and insanity exist, how shallwe know them?The question is neither capricious nor

itself insane. However much we maybe personally convinced that we can

tell the normal from the abnormal, theevidence is simply not compelling. It iscommonplace, for example, to readabout murder trials wherein eminentpsychiatrists for the defense are con-

250

tradicted by equally eminent psychia-trists for the prosecution on the matterof the defendant's sanity. More gen-

erally, there are a great deal of conflict-ing data on the reliability, utility, andmeaning of such terms as "sanity," "in-sanity," "mental illness," and "schizo-phrenia" (1). Finally, as early as 1934,Benedict suggested that normality andabnormality are not universal (2).

What is viewed as normal in one cul-ture may be seen as quite aberrant inanother. Thus, notions of normality andabnormality may not be quite-as accu-rate as people believe they are.To raise questions regarding normal-

ity and abnormality is in no way toquestion the fact that some behaviorsare deviant or odd. Murder is deviant.So, too, are hallucinatio,ns. Nor doesraising such questions deny the exis-tence of the personal anguish that isoften associated with "mental illness."Anxiety and depression exist. Psycho-logical sufferng' exists. But normalityand abnormality, sanity and insanity,and the diagnoses that flow from them

Th author is professor of psychoWgy and lawat Stanford University, Stanford, Caifornis 9430S.Portions of thee data were presented to collo-quiums of the psychology departents at theUniversity of Cafomra at Berkeley and at SantaBarbar; University of Arizona, Tucson; andHarvard Univwrsity, Cambridge, Massachusetts.

SCIENCE6 VOL. 179

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Mar

ch 1

3, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

may be less substantive than many be-,lieve them to be.At its heart, the question of whether

the sane can be distinguished from theinsane (and whether degrees of insanitycan be distinguished from each other)is a simple matter: do the salient char-acteristics that lead to diagnoses residein the patients themselves or in the en-vironments and contexts in which ob-servers find them? From Bleuler,through Kretchmer, through the formu-lators of the recently revised Diagnosticand Statistical Manual of the AmericanPsychiatric Association, the belief hasbeen strong that patients present symp-toms, that those symptoms can be cate-gorized, and, implicitly, that the saneare distinguishable from the insane.More recently, however, this belief hasbeen questioned. Based in part on theo-retical and anthropological considera-tions, but also on philosophical, legal,and therapeutic ones, the view hasgrown that psychological categorizationof mental illness is useless at best anddownright harmful, misleading, andpejorative at worst. Psychiatric diag-noses, in this view, are in the minds ofthe observers and are not valid sum-maries of characteristics displayed bythe observed (3-5).

Gains can be made in deciding whichof these is more nearly accurate bygetting normal people (that is, peoplewho do not have, and have never suf-fered, symptoms of serious psychiatricdisorders) admitted to psychiatric hos-pitals and then determining whetherthey were discovered to be sane and, ifso, how. If the sanity of such pseudo-patients were always detected, therewAlNd be prima facie evidence that a

sane individual can be distinguishedfrom the insane context in which he isfound. Normality (and presumably ab-normality) is distinct enough that itcan be recognized wherever it occurs,for it is carried within the person. If,on the other hand, the sanity of thepseudopatients were never discovered,serious difficulties would arise for thosewho support traditional modes of psy-chiatric diagnosis. Given that the hospi-tal staff was not incompetent, that thepseudopatient had been behaving assanely as he had been outside of thehospital, and that it had never beenpreviously suggested that he belongedin a psychiatric hospital, such an un-likely outcome would support the view

that psychiatric diagnosis betrays littleabout the patient but much about theenvironment in which an observer findshim.19 JANUARY 1973

This article describes such an experi-ment. Eight sane people gained secretadmission to 12 different hospitals (6).Their diagnostic experiences constitutethe data of the first part of this article;the remainder is devoted to a descrip-tion of their experiences in psychiatricinstitutions. Too few psychiatrists andpsychologists, even those who haveworked in such hospitals, know whatthe experience is like. They rarely talkabout it with former patients, perhapsbecause they distrust information com-ing from the previously insane. Thosewho have worked in psychiatric hospi-tals are likely to have adapted so thor-oughly to the settings that they areinsensitive to the impact of that expe-rience. And while there have been oc-casional reports of researchers whosubmitted themselves to psychiatric hos-pitalization (7), these researchers havecommonly remained in the hospitals forshort periods of time, often with theknowledge of the hospital staff. It isdifficult to know the extent to whichthey were treated like patients or likeresearch colleagues. Nevertheless, theirreports about the inside of the psychi-atric hospital have been valuable. Thisarticle extends those efforts.

Pseudopatients and Their Settings

The eight pseudopatients were avaried group. One was a psychologygraduate student in his 20's. The re-maining seven were older and "estab-lished." Among them were three psy-chologists, a pediatrician, a psychiatrist,a painter, and a housewife. Threepseudopatients were women, five weremen. All of them employed pseudo-nyms, lest their alleged diagnoses em-barrass them later. Those who were inmental health professions alleged an-other occupation in order to avoid thespecial attentions that might be ac-corded by staff, as a matter of courtesyor caution, to ailing colleagues (8).With the exception of myself (I was thefirst pseudopatient and my presence wasknown to the hospital administrator andchief psychologist and, so far as I cantell, to them alone), the presence ofpseudopatients and the nature of the re-search program was not known to thehospital staffs (9).The settings were similarly varied. In

order to generalize the findings, admis-sion into a variety of hospitals wassought. The 12 hospitals in the samplewere located in five different states onthe East and West coasts. Some were

old and shabby, some were quite new.Some were research-oriented, othersnot. Some had good staff-patient ratios,others were' quite understaffed. Onlyone was a strictly private hospital. Allof the others were supported by stateor federal funds or, in one instance, byuniversity funds.

After calling the hospital for an ap-pointment, the pseudopatient arrived atthe admissions office complaining thathe had been hearing voices. Asked whatthe voices said, he replied that theywere often unclear, but as far as hecould tell they said "empty," "hollow,"and "thud." The voices were unfamiliarand were of the same sex as the pseudo-patient. The choice of these symptomswas occasioned by their apparent sim-ilarity to existential symptoms. Suchsymptoms are alleged to arise frompainful concerns about the perceivedmeaninglessness of one's life. It is asif the hallucinating person were saying,"My life is empty and hollow." Thechoice of these symptoms was also de-termined by the absence of a singlereport of existential psychoses in theliterature.Beyond alleging the symptoms and

falsifying name, vocation, and employ-ment, no further alterations of person,history, or circumstances were made.The significant events of the pseudo-patient's life history were presented asthey had actually occurred. Relation-ships with parents and siblings, withspouse and children, with people atwork and in school, consistent with theaforementioned exceptions, were de-scribed as they were or had been. Frus-trations and upsets were describedalong with joys and satisfactions. Thesefacts are important to remember. Ifanything, they strongly biased the sub-sequent results in favor of detectingsanity, since none of their histories orcurrent behaviors were seriously patho-logical in any way.

Immediately upon admission to thepsychiatric ward, the pseudopatientceased simulating any symptoms of ab-normality. In some cases, there was abrief period of mild nervousness andanxiety, since none of the pseudopa-tients really believed that they would beadmitted so easily. Indeed, their sharedfear was that they would be immedi-ately exposed as frauds and greatlyembarrassed. Moreover, many of themhad never visited a psychiatric ward;even those who had, nevertheless hadsome genuine fears about what mighthappen to them. Their nervousness,then, was quite appropriate to the nov-

251

elty of the hospital setting, and -it abatedrapidly.

Apart from that short-lived nervous-ness, the pseudopatient behaved on theward as he "normally" behaved. Thepseudopatient spoke to patients andstaff as he might ordinarily. Becausethere is uncommonly little to do on apsychiatric ward, he attempted to en-gage others in conversation. Whenasked by staff how he was feeling, heindicated that he was fine, that he nolonger experienced symptoms. He re-sponded to instructions from attendants,to calls for medication (which was notswallowed), and to dining-hall instruc-tions. Beyond such activities as wereavailable to him on the admissionsward, he spent his time writing downhis observations about the ward, itspatients, and the staff. Initially thesenotes were written "secretly," but as itsoon became clear that no one muchcared, they were subsequently writtenon standard tablets of paper in suchpublic places as the dayroom. No secretwas made of these activities.The pseudopatient, very much as a

true psychiatric patient, entered a hos-pital with no foreknowledge of whenhe would be discharged. Each was toldthat he would have to get out by hisown devices, essentially by convincingthe staff that he was sane. The psycho-logical stresses associated with hospital-ization were considerable, and all butone of the pseudopatients desired to bedischarged almost immediately afterbeing admitted. They were, therefore,motivated not only to behave sanely,but to be paragons of cooperation. That,their behavior was in no way disruptiveis confirmed by nursing reports, whichhave been obtained on most of thepatients. These reports uniformly indi-cate that the patients were "friendly,"("cooperative," and "exhibited no ab-normal indications."

The Normal Are Not Detectably Sane

Despite their public "show" of sanity,the pseudopatients were never detected.Admitted, except in one case, with a

diagnosis of schizophrenia (10), eachwas discharged with a diagnosis ofschizophrenia "in remission." The label"in remission" should in no way bedismissed as a formality, for at no timeduring any hospitalization had anyquestion been raised about any pseudo-patient's simulation. Nor are there anyindications in the hospital records thatthe pseudopatient's status was suspect.Rather, the evidence is strong that, once

252

labeled schizophrenic, the pseudopatientwas stuck with that label. If the pseudo-patient was to be discharged, he mustnaturally be "in remission"; but he wasnot sane, nor, in the institution's view,had he ever been sane.The uniform failure to recognize san-

ity cannot be attributed to the qualityof the hospitals, for, although therewere considerable variations amongthem, several are considered excellent.Nor can it be alleged that there wassimply not enough time to observe thepseudopatients. Length of hospitaliza-tion ranged from 7 to 52 days, with anaverage of 19 days. The pseudopatientswere not, in fact, carefully observed,but this failure clearly speaks more totraditions within psychiatric hospitalsthan to lack of opportunity.

Finally, it cannot be said that thefailure to recognize the pseudopatients'sanity was due to the fact that theywere not behaving sanely. While therewas clearly some tension present in allof them, their daily visitors could detectno serious behavioral consequences-nor, indeed, could other patients. It wasquite common for the patients to "de-tuct" the pseudopatients' sanity. Duringthe first three hospitalizations, whenaccurate counts were kept, 35 of a totalof 118 patients on the admissions wardvoiced their suspicions, some vigorously."You're not crazy. You're a journalist,or a professor [referring to the con-tinual note-taking]. You're checking upon the hospital." While most of thepatients were reassured by the pseudo-patient's insistence that he had beensick before he came in but was finenow, some continued to believe thatthe pseudopatient was sane throughouthis hospitalization (11). The fact thatthe patients often recognized normalitywhen staff did not raises importantquestions.

Failure to detect sanity during thecourse of hospitalization may be dueto the fact that physicians operate witha strong bias toward what statisticianscall the type 2 error (5). This is tosay that physicians are more inclinedto call a healthy person sick (a falsepositive, type 2) than a sick personhealthy (a false negative, type 1). Thereasons for this are not hard to find:it is clearly more dangerous to mis-diagnose illness than health. Better toerr on the side of caution, to suspectillness even among the healthy.

But what holds for medicine doesnot hold equally well for psychiatry.Medical illnesses, while unfortunate, arenot commonly pejorative. Psychiatricdiagnoses, on the contrary, carry with

them personal, legal, and social stigmas(12). It was therefore important to see%ihether the tendency toward diagnosingthe sane insane could be reversed. Thefollowing experiment was arranged ata research and teaching hospital whosestaff had heard these findings butdoubted that such an error could occurin their hospital. The staff was informedthat at some time during the following3 months, one or more pseudopatientswould attempt to be admitted into thepsychiatric hospital. Each staff memberwas asked to rate each patient who pre-sented himself at admissions or on theward according to the likelihood thatthe patient was a pseudopatient. A 10-point scale was used, with a 1 and 2reflecting high confidence that the pa-tient was a pseudopatient.

Judgments were obtained on 193 pa-tients who were admitted for psychi-atric treatment. All staff who had hadsustained contact with or primary re-sponsibility for the patient-attendants,nurses, psychiatrists, physicians, andpsychologists-were asked to makejudgments. Forty-one patients were al-leged, with high confidence, to bepseudopatients by at least one memberof the staff. Twenty-three were consid-ered suspect by at least one psychiatrist.Nineteen were suspected by one psychi-atrist and one other staff member.Actually, no genuine pseudopatient (atleast from my group) presented himselfduring this period.The experiment is instructive. It indi-

cates that the tendency to designatesane people as insane can be reversedwhen the stakes (in this case, prestigeand diagnostic acumen) are high. Butwhat can be said of the 19 people whowere suspected of being "sane" by onepsychiatrist and another staff member?Were these people truly "sane," or wasit rather the case that in the course ofavoiding the type 2 error the stafftended to make more errors of the firstsort-calling the crazy "sane"? There isno way of knowing. But one thing iscertain: any diagnostic process thatlends itself so readily to massive errorsof this sort cannot be a very reliableone.

The Stickiness of

Psychodiagnostic Labels

Beyond the tendency to call thehealthy sick-a tendency that accountsbetter for diagnostic behavior on admis-sion than it does for such behavior aftera lengthy period of exposure-the dataspeak to the massive role of labeling in

SCINCE, VOL. 179

psychiatric assessment. Having oncebeen labeled schizophrenic, there isnothing the pseudopatient can do toovercome the tag. The tag profoundlycolors others' perceptions of him andhis behavior.From one viewpoint, these data are

hardly surprising, for it has long beenknown that elements are given meaningby the context in which they occur.Gestalt psychology made this pointvigorously, and Asch (13) demon-strated that there are "central" person-ality traits (such as "warm" versus"cold") which are so powerful that theymarkedly color the meaning of otherinformation in forming an impressionof a given personality (14). "Insane,""schizophrenic," "manic-depressive,"and "crazy" are probably among themost powerful of such central traits.Once a person is designated abnormal,all of his other behaviors and character-istics are colored by that label. Indeed,that label is so powerful that many ofthe pseudopatients' normal behaviorswere overlooked entirely or profoundlymisinterpreted. Some examples mayclarify this issue.

Earlier I indicated that there wereno changes in the pseudopatient's per-sonal history and current status beyondthose of name, employment, and, wherenecessary, vocation. Otherwise, a veridi-cal description of personal history andcircumstances was offered. Those cir-cumstances were not psychotic. Howwere they made consonant with thediagnosis of psychosis? Or were thosediagnoses modified in such a way as tobring them into accord with the cir-cumstances of the pseudopatient's life,as described by him?As far as I can determine, diagnoses

were in no way affected by the relativehealth of the circumstances of a pseudo-patient's life. Rather, the reverse oc-curred: the perception of his cir-cumstances was shaped entirely by thediagnosis. A clear example of suchtranslation is found in the case of apseudopatient who had had a close re-lationship with his mother but wasrather remote from his father duringhis early childhool. During adolescenceand beyond, however, his father be-came a close friend, while his relation-ship with his mother cooled. His presentrelationship with his wife was charac-teristically close and warm. Apart fromOccasional angry exchanges, frictionwas minimal. The children had rarelybeen spanked. Surely there is nothingespecially pathological about such a lhistory. Indeed, many readers may seea similar pattern in their own experi-19 JANUARY 1973

ences, with no markedly. deleteriousconsequences. Observe, however, howsuch a history was translated in theipsychopathological context, this fromthe case summary prepared after thepatient was discharged.

This white 39-year-old male . . . mani-fests a long history of considerable ambiv-alence in close relationships, which beginsin early childhood. A warm relationshipwith his mother cools during his adoles-cence. A distant relationship to his fatheris described as becoming very intense.Affective stability is absent. His attemptsto control emotionality with his wife andchildren are punctuated by angry out-bursts and, in the case of the children,spankings. And while he says that he hasseveral good friends, one senses consider-able ambivalence embedded in those rela-tionships also....The facts of the case were uninten-

tionally distorted by the staff to achieveconsistency with a popular theory ofthe dynamics of a schizophrenic reac-tion (15). Nothing of an ambivalentnature had been described in relationswith parents, spouse, or friends. To theextent that ambivalence could be in-ferred, it was probably not greater thanis found in all human relationships. Itis true the pseudopatient's relationshipswith his parents changed over time, butin the ordinary context that wouldhardly be remarkable-indeed, it mightvery well be expected. Clearly, themeaning ascribed to his verbalizations(that is, ambivalence, affective instabil-ity) was determined by the diagnosis:schizophrenia. An entirely differentmeaning would have been ascribed ifit were known that the man was"normal."

All pseudopatients took extensivenotes publicly. Under ordinary circum-stances, such behavior would haveraised questions in the minds of ob-servers, as, in fact, it did among pa-tients. Indeed, it seemed so certain thatthe notes would elicit suspicion thatelaborate precautions were taken to re-move them from the ward each day.But the precautions proved needless.The closest any staff member came toquestioning these notes occurred whenone pseudopatient asked his physicianwhat kind of medication he was receiv-ing and began to write down the re-sponse. "You needn't write it," he wastold gently. "If you have trouble re-membering, just ask me again."

If no questions were asked of thepseudopatients, how was their writinginterpreted? Nursing records for threepatients indicate that the writing wasseen as an aspect of their pathologicalbehavior. "'Patient engages in writingbehavior" was the daily nursing com-

ment on one of the pseudopatients whowas never questioned about his writing.Given that the patient is in the hospital,he must be psychologically disturbed.And given that he is disturbed, continu-ous writing must be a behavioral mani-festation of that disturbance, perhaps asubset of the compulsive behaviors thatare sometimes correlated with schizo-phrenia.One tacit characteristic of psychiatric

diagnosis is that it locates the sourcesof aberration within the individual andonly rarely within the complex of stim-uli that surrounds him. Consequently,behaviors that are stimulated by theenvironment are commonly misattrib-uted to the patient's disorder. For ex-ample, one kindly nurse found apseudopatient pacing the long hospitalcorridors. "Nervous, Mr. X?" she asked."No, bored," he said.The notes kept by pseudopatients are

full of patient behaviors that were mis-interpreted by well-intentioned staff.Often enough, a patient would go "ber-serk" because he had, wittingly or un-wittingly, been mistreated by, say, anattendant. A nurse coming upon thescene would rarely inquire even cursor-ily into the environmental stimuli ofthe patient's behavior. Rather, she as-sumed that his upset derived from hispathology, not from his present inter-actions with other staff members. Oc-casionally, the staff might assume thatthe patient's family (especially whenthey had recently visited) or other pa-tients had stimulated the outburst. Butnever were the staff found to assumethat one of themselves or the structureof the hospital had anything to do witha patient's behavior. One psychiatristpointed to a group of patients who weresitting outside the cafeteria entrancehalf an hour before lunchtime. To agroup of young residents he indicatedthat such behavior was characteristicof the oral-acquisitive pature of thesyndrome. It seemed not to occur tohim that there were very few things toanticipate in a psychiatric hospital be-sides eating.A psychiatric label has a life and an

influence of its own. Once the impres-sion has been formed that the patient isschizophrenic, the expectation is thathe will continue to be schizophrenic.When a sufficient amount of time haspassed, during which the patient hasdone nothing bizarre, he is consideredto be in remission and available for dis-charge. But the label endures beyonddischarge, with the unconfirmed expec-tation that he will behave as a schizo-phrenic again. Such labels, conferred

253

by mental health professionals, are asinfluential on the patient as they are onhis relatives and friends, and it shouldnot surprise anyone that the diagnosisacts on all of them as a self-fulfillingprophecy. Eventually, the patient him-self accepts the diagnosis, with all ofits surplus meanings and expectations,and behaves accordingly (5).The inferences to be made from

these matters are quite simple. Muchas Zigler and Phillips have demon-strated that there is enormous overlapin the symptoms presented by patientswho have been variously diagnosed(16), so there is enormous overlap inthe behaviors of the sane and the in-sane. The s.- ne are not "sane" all ofthe time. We lose our tempers "for nogood reason." We are occasionally de-pressed or anxious, again for no goodreason. And we may find it difficult toget along with one or another person-again for no reason that we can specify.Similarly, the insane are not always in-sane. Indeed, it was the impression ofthe pseudopatients while living withthem that they were sane for long pe-riods of time-that the bizarre behav-iors upon which their diagnoses wereallegedly predicated constituted only asmall fraction of their total behavior.If it makes no sense to label ourselvespermanently depressed on the basis ofan occasional depression, then it takesbetter evidence than is presently avail-able to label all patients insane orschizophrenic on the basis of bizarrebehaviors or cognitions. It seems more

useful, as Mischel (17) has pointedout, to limit our discussions to behav-iors, the stimuli that provoke them, andtheir correlates.

It is not known why powerful impres-sions of personality traits, such as"crazy" or "insane," arise. Conceivably,when the origins of and stimuli thatgive rise to a behavior are remote orunknown, or when the behavior strikesus as immutable, trait labels regardingthe behaver arise. When, on the otherhand, the origins and stimuli are knownand available, discourse is limited tothe behavior itself. Thus, I may hallu-cinate because I am sleeping, or I mayhallucinate because I have ingested apeculiar drug. These are termed sleep-induced hallucinations, or dreams, anddrug-induced hallucinations, respective-ly. But when the stimuli to my hallu-cinations are unknown, that is calledcraziness, or schizophrenia-as if thatinference were somehow as illuminatingas the others.

Te Experience of

Psychiatric Hospitaization

The term 'mental illness" is of re-cent origin. It was coined by peoplewho were humane in their inclinationsand who wanted very much to raise thestation of (and the public's sympathiestoward) the psychologically disturbedfrom that of witches and "crazies" toone that was akin to the physically ill.And they were at least partially success-ful, for the treatment of the mentallyill has improved considerably over theyears. But while treatment has im-proved, it is doubtful that people reallyregard the mentally ill in the same waythat they view the physically ill. Abroken leg is something one recoversfrom, but mental illness allegedly en-dures forever (18). A broken leg doesnot threaten the observer, but a crazyschizophrenic? There is by now a hostof evidence that attitudes toward thementally ill are characterized by fear,hostility, aloofness, suspicion, and dread(19). The mentally ill are society'slepers.

That such attitudes infect the generalpopulation is perhaps not surprising,only upsetting. But that they affect theprofessionals-attendants, nurses, phy-sicians, psychologists, and social work-ers-who treat and deal with the men-tally ill is more disconcerting, bothbecause such attitudes are self-evidentlypernicious and because they are unwit-ting. Most mental health professionalswould insist that they are sympathetictoward the mentally ill, that they areneither avoidant nor hostile. But it ismore likely that an exquisite ambiv-alence characterizes their relations withpsychiatric patients, such that theiravowed impulses are only part of theirentire attitude. Negative attitudes arethere too and can easily be detected.Such attitudes should not surprise us.They are the natural offspring of thelabels patients wear and the places inwhich they are found.

Consider the structure of the typicalpsychiatric hospital. Staff and patientsare strictly segregated. Staff have theirown living space, including their diningfacilities, bathrooms, and assemblyplaces. The glassed quarters that con-tain the professional staff, which thepseudopatients came to call "the cage,"sit out on every dayroom. The staffemerge primarily for caretaking pur-poses-to give medication, to conduct atherapy or group meeting, to instruct orreprimand a patient. Otherwise, staff

keep to themselves, almost as if the dis-order that afflicts their charges is some-how catching.So much is patient-staff segregation

the rule that, for four public hospitalsin which an attempt was made to mea-sure the degree to which staff and pa-tients mingle, it was necessary to use"time out of the staff cage" as theoperational measure. While it was notthe case that all time spent out of thecage was spent mingling with patients(attendants, for example, would occa-sionally emerge to watch television inthe dayroom), it was the only way inwhich one could gather reliable dataon time for measuring.The average amount of time spent

by attendants outside of the cage was11.3 percent (range, 3 to 52 percent).This figure does not represent onlytime spent mingling with patients, butalso includes time spent on such choresas folding laundry, supervising patientswhile they shave, directing ward clean-up, and sending patients to off-wardactivities. It was the relatively rare at-tendant who spent time talking withpatients or playing games with them. Itproved impossible to obtain a "percentmingling time" for nurses, since theamount of time they spent out of thecage was too brief. Rather, we countedinstances of emergence from the cage.On the average, daytime nurses emergedfrom the cage 11.5 times per shift,including instances when they left theward entirely (range, 4 to 39 times) .Late afternoon and night nurses wereeven less available, emerging on theaverage 9.4 times per shift (range, 4 to41 times). Data on early morningnurses, who arrived usually after mid-night and departed at 8 a.m., are notavailable because patients were asleepduring most of this period.

Physicians, especially psychiatrists,were even less available. They wererarely seen on the wards. Quite com-monly, they would be seen only whenthey arrived and departed, with the re-maining time being spent in their officesor in the cage. On the average, physi-cians emerged on the ward 6.7 timesper day (range, 1 to 17 times). Itproved difficult to make an accurateestimate in this regard, since physiciansoften maintained hours that allowedthem to come and go at different times.The hierarchical organization of the

psychiatric hospital has been com-mented on before (20), but the latentmeaning of that kind of organization i5worth noting again. Those with the'"

SCIENCE, VOL. 179*254

Table 1. Self-initiated contact by pseudopatients with psychiatrists and nurses and attendants, compared to contact with other groups.

Psyciatrchmitab University capu University medical center

Psychiatric hospitals Uniicalu PhysiciansContact (2)

Psychiatrists aNurses (3) "Looking for a "Looking for No additionalPsychitrists attndat Fcly psychiatrist"" an internist"l comment

ResponsesMoves on, head averted (%) 71 88 0 0 0 0Makes eye contact (%) 23 '10 11 0 0Pauses and chats () 2 2 0 ll 0 1OStops and talks (%) 4 0.5 100 78 tOO 90

Mean number of questionsanswered (out of 6) * * 6 3.8 4.8 4.5

Respondents (No.) 13 47 14 18 15 10Attempts (No.) 185 1283 14 18 15 10* Not applicabk.

most power have least to do with pa'tients, and those with the least power,are most involved with them. Recall,khowever, that the acquisition of role-appropriate behaviors occurs mainlythrough the observation of others, withthe most powerful having the most in-fluence. Consequently, it is understand-able that attendants not only spendmore time with patients than do anyother members of the staff-that is re-quired by their station in the hierarchy-but also, insofar as they learn fromtheir superiors' behavior, spend as littletime with patients as they can. Attend-ants are seen mainly in the cage, whichis where the models, the action, andthe power are.

I turn now to a different set ofstudies, these dealing with staff re-sponse to patient-initiated contact. Ithas long been known that the amountof time a person spends with you canbe an index of your significance to him.If he initiates and maintains eye con-tact, there is reason to believe that heis considering your requests and needs.If he pauses to chat or actually stopsand talks, there is added reason to inferthat he is individuating you. In fourhospitals, the pseudopatient approachedthe staff member with a request whichtook the following form: "Pardon me,Mr. (or Dr. or Mrs.] X, could you tellme when I will be eligible for groundsprivileges?" (or " . . . when I will bePresented at the staff meeting?" or ". . .when I am likely to be discharged?").While the content of the question variedaccording to the appropriateness of thetarget and the pseudopatient's (appar-ent) current needs the form was al-Ways a courteous and relevant requestfor information. Care was taken neverto approach a particular member of theStaff more than once a day, lest theStaff member become suspicious or ir-19 JANUARY 1973

ritated. In examining these data, re-member that the behavior of thepseudopatients was neither bizarre nordisruptive. One could indeed engage ingood conversation with them.The data for these experiments are

shown in Table 1, separately for physi-cians (column 1) and for nurses andattendants (column 2). Minor differ-ences between these four institutionswere overwhelmed by the degree towhich staff avoided continuing contactsthat patients had initiated. By far, theirmost common response consisted ofeither a brief response to the question,offered while they were "on the move"and with head averted, or no responseat all.The encounter frequently took the

following bizarre form: (pseudopatient)"Pardon me, Dr. X. Could you tell mewhen I am eligible for grounds priv-ileges?" (physician) "Good morning,Dave. How are you today?" (Moves offwithout waiting for a response.)

It is instructive to compare thesedata with data recently obtained atStanford University. It has been allegedthat large and eminent universities arecharacterized by faculty who are sobusy that they have no time for stu-dents. For this comparison, a younglady approached individual faculty mem-bers who seemed to be walking pur-posefully to some meeting or teachingengagement and asked them the fol-lowing six questions.

1 ) "Pardon me, could you direct meto Encina Hall?" (at the medicalschool: ". . . to the Clinical ResearchCenter?").

2) "Do you know where Fish Annexis?" (there is no Fish Annex at Stan-ford).

3) "Do you teach here?"4) "How does one apply for admis-

sion to the college?" (at the medical

school: ". . . to the medical school?").5) "Is it difficult to get in?"6) "Is there financial aid?"Without exception, as can be seen in

Table 1 (column 3), all of the questionswere answered. No matter how rushedthey were, all respondents not onlymaintained eye contact, but stopped totalk. Indeed, many of the respondentswent out of their way to direct or takethe questioner to the office she wasseeking, to try to locate "Fish Annex,"or to discuss with her the possibilitiesof being admitted to the university.

Similar data, also shown in Table I(columns 4, 5, and 6), were obtainedin the hospital. Here too, the younglady came prepared with six questions.After the first question, however, sheremarked to 18 of her respondents(column 4), "I'm looking for a psy-chiatrist," and to 15 others (column5), "I'm looking for an internist." Tenother respondents received no insertedcomment (column 6). The general de-gree of cooperative responses is con-siderably higher for these universitygroups than it was for pseudopatientsin psychiatric hospitals. Even so, differ-ences are apparent within the medicalschool setting. Once having indicatedthat she was looking for a psychiatrist,the degree of cooperation elicited wasless than when she sought an internist.

Powerlessness and Depersonalization

Eye contact and verbal contact re-flect concern and individuation; theirabsence, avoidance and depersonaliza-tion. The data I have presented do notdo justice to the rich daily encountersthat grew up around matters of deper-sonalization and avoidance. I have rec-ords of patients who were beaten bystaff for the sin of having initiated ver-

bal contact. During my own experience,for example, one patient was beaten inthe presence of other patients for hav-ing approached an attendant and toldhim, "I like you." Occasionally, punish-ment meted out to patients for misde-meanors seemed so excessive that itcould not be justified by the most radi-cal interpretations of psychiatric canon.Nevertheless, they appeared to go un-questioned. Tempers were often short.A patient who had not heard a call formedication would be roundly excori-ated, and the morning attendants wouldoften wake patients with, "Come on,you m----f----s, out of bed!"

Neither anecdotal nor "hard" datacan convey the overwhelming sense ofpowerlessness which invades the indi-vidual as he is continually exposed tothe depersonalization of the psychiatrichospital. It hardly matters which psy-,fiiatric hospital-the excellent public,*s and the very plush private hospitalwere better than the rural and shabby-ones in this regard, but, again, thefeatures that psychiatric hospitals hadin common overwhelmed by far theirapparent differences.

Powerlessness was evident every-where. The patient is deprived of manyof his legal rights by dint of his psy-chiatric commitment (21). He is shornof credibility by virtue of his psychiatriclabel. His freedom of movement is re-stricted. He cannot initiate contact withthe staff, but may only respond to suchovertures as they make. Personal pri-vacy is ininimal. Patient quarters andpossessions can be entered and ex-amined by any staff member, for what-ever reason. His personal history andanguish is available to any staff member(often including the "grey lady" and"candy striper" volunteer) who choosesto read his folder, regardless of theirtherapeutic relationship to him. His per-sonal hygiene and waste evacuation areoften monitored. The water closets mayhave no doors.

At times, depersonalization reachedsuch proportions that pseudopatientshad the sense that they were invisible,or at least unworthy of account. Uponbeing admitted, I and other pseudo-patients took the initial physical exami-nations in a semipublic room, wherestaff members went about their ownbusiness as if we were not there.On the ward, attendants delivered

verbal and occasionally serious physicalabuse to patients in the presence ofother observing patients, some of whom(the pseudopatients) were writing it all

down. Abusive behavior, on the otherhand, terminated quite abruptly whenother staff members were known to becoming. Staff are credible witnesses.Patients are not.A nurse unbuttoned her uniform to

adjust her brassiere in the presence ofan entire ward of viewing men. One didnot have the sense that she was beingseductive. Rather, she didn't notice us.A group of staff persons might point toa patient in the dayroom and disctisshim animatedly, as if he were not there.One illuminating instance of deper-

sonalization and invisibility occurredwith regard to medications. All told,the pseudopatients were administerednearly 2100 pills, including Elavil,Stelazine, Compazine, and Thorazine,to name but a few. (That such a varietyof medications should have been ad-ministered to patients presenting identi-cal symptoms is itself worthy of note.)Only two were swallowed. The restwere either pocketed or deposited inthe toilet. The pseudopatients were notalone in this. Although I have no pre-cise records on how many patientsrejected their medications, the pseudo-patients frequently found the medica-tions of other patients in the toiletbefore they deposited their own. Aslong as they were cooperative, theirbehavior and the pseudopatients' own

in this matter, as in other importantmatters, went unnoticed throughout.

Reactions to such depersonalizationamong pseudopatients were intense. Al-though they had come to the hospitalas participant observers and were fullyaware that they did not "belong," theynevertheless found themselves caughtup in and fighting the process of de-personalization. Some examples: a grad-uate student in psychology asked hiswife to bring his textbooks to the hos-pital so he could "catch up on hishomework"-this despite the elaborateprecautions taken to conceal his profes-sional association. The same student,who had trained for quite some timeto get into the hospital, and who hadlooked forward to the experience, "re-membered" some drag races that hehad wanted to see on the weekend andinsisted that he be discharged by thattime. Another pseudopatient attempteda romance with a nurse. Subsequently,he informed the staff that he was ap-plying for admission to graduate schoolin psychology and was very likely to beadmitted, since a graduate professorwas one of his regular hospital visitors.The same person began to engage in

psychotherapy with other patients-allof this as a way of becoming a personin an impersonal environment.

The Sources of Depersonalization

What are the origins of depersonali-zation? I have already mentioned two.First are attitudes held by all of ustoward the mentally ill-including thosewho treat them-attitudes character-ized by fear, distrust, and horrible ex-pectations on the one hand, and benev-olent intentions on the other. Ourambivalence leads, in this instance asin others, to avoidance.

Second, and not entircly separate,the hierarchical structure of the psy-chiatric hospital facilitates depersonali-zation. Those who are at the top haveleast to do with patients, and their be-havior inspires the rest of the staff.Average daily contact with psychia-trists, psychologists, residents, andphysicians combined ranged from 3.9to 25.1 minutes, with an overall meanof 6.8 (six pseudopatients over a totalof 129 days of hospitalization). In-cluded in this average are time spentin the admissions interview, ward meet-ings in the presence of a senior staffmember, group and individual psycho-therapy contacts, case presentation con-ferences, and discharge meetings.Clearly, patients do not spend muchtime in interpersonal contact with doc-toral staff. And doctoral staff serve asmodels for nurses and attendants.There are probably other sources.

Psychiatric installations are presently inserious financial straits. Staff shortagesare pervasive, staff time at a premium.Something has to give, and that some-thing is patient contact. Yet, whilefinancial stresses are realities, too muchcan be made of them. I have the im-pression that the psychological forcesthat result in depersonalization aremuch stronger than the fiscal ones andthat the addition of more staff wouldnot correspondingly improve patientcare in this regard. The incidence ofstaff meetings and the enormousamount of record-keeping on patients,for example, have not been as sub-stantially reduced as has patient con-tact Priorities exist, even during hardtimes. Patient contact is -not a signifi-cant priority in the traditional psychia-tric hospital, and fiscal pressures do notaccount for this. Avoidance and de-personalization may.

Heavy reliance upon psychotroP;'.SCLENC-E. VOL. 179,i256

mnedication tacitly contributes to deper-sonalization by convincing staff thattreatment is indeed being conductedand that further patient contact maynot be necessary. Even here, however,caution needs to be exercised in under-standing the role of psychotropic drugs.If patients were powerful rather thanpowerless, if they were viewed as inter-esting individuals rather than diagnosticentities, if they were socially significantrather than social lepers, if their an-guish truly and wholly compelled oursympathies and concerns, would wenot seek contact with them, despite theavailability of medications? Perhaps forthe pleasure of it all?

The Consequences of Labeling

and Depersonalization

Whenever the ratio of what is knownto what needs to be known approacheszero, we tend to invent "knowledge"and assume that we understand morethan we actually do. We seem unableto acknowledge that we simply don'tknow. The needs for diagnosis andremediation of behavioral and emo-tional problems are enormous. Butrather than acknowledge that we arejust embarking on understanding, wecontinue to label patients "schizo-phrenic," "manic-depressive," and "in-sane," as if in those words we hadcaptured the essence of understanding.The facts of the matter are that wehave known for a long time that diag-noses are often not useful or reliable,but we have nevertheless continued touse them. We now know that we can-Inot distinguish insanity from sanity. Itis depressing to consider how that in-formation will be used.Not merely depressing, but frighten-

ing. How many people, one wonders,are sane but not recognized as such inour psychiatric institutions? How manyhave been needlessly stripped of theirprivileges of citizenship, from the rightto vote and drive to that of handlingtheir own accounts? How many havefeigned insanity in order to avoid thecriminal consequences of their behav-ior, and, conversely, how many wouldrather stand trial than live interminablyin a psychiatric hospital-but arewrongly thought to be mentally- ill?How many have been stigmatized bywell-intentioned, but nevertheless erro-neous, diagnoses? On the last point,recall again that a "type 2 error" inPsychiatric diagnosis does not have the19 JANUARY 1973

same consequences it does in medicaldiagnosis. A diagnosis of cancer thathas been found to be in error is causefor celebration. But psychiatric diag-noses are rarely found to be in error.The label sticks, a mark of inadequacyforever.

Finally, how many patients might be"sane" outside the psychiatric hospitalbut seem insane in it-not becausecraziness resides in them, as it were,but because they are responding te abizarre setting, one that may be uniqueto institutions which harbor netherpeople? Goffman (4) calls the processof socialization to such institutions"mortification"-an apt metaphor thatincludes the processes of depersonali-zation that have been described nere.And while it is impossible to knowwhether the pseudopatients' responsesto these processes are characteristic ofall inmates-they were, after all, notreal patients-it is difficult to believethat these processes of socialization toa psychiatric hospital provide usefulattitudes or habits of response for liv-ing in the "real world."

Summary and Conclusions

It is clear that we cannot distinguishthe sane from the insane in psychiatrichospitals. The hospital itself imposes aspecial environment in which the mean-ings of behavior can easily be misunder-stood. The consequences to patientshospitalized in such an environment-the powerlessness, depersonalization,segregation, mortification, and self-labeling-seem undoubtedly counter-therapeutic.

I do not, even now, understand thisproblem well enough to perceive solu-tions. But two matters seem to havesome promise. The first concerns theproliferation of community mentalhealth facilities, of crisis interventioncenters, of the human potential move-ment, and of behavior therapies that,for all of their own problems, tend toavoid psychiatric labels, to focus onspecific problems and behaviors, and toretain the individual in a relatively non-pejorative environment. Clearly, to theextent that we refrain from sending thedistressed to insane places, our impres-sions of them are less likely to be dis-torted. (The risk of distorted percep-tions, it seems to me, is always present,since we are much more sensitive to anindividual's behaviors and- verbaliza-tions than we are to the subtle con-

textual stimuli that often promote them.At issue here is a matter of magnitude.And, as I have shown, the magnitudeof distortion is exceedingly high in theextreme context that is a psychiatrichospital.)The second matter that might prove

promising speaks to the need to in-crease the sensitivity of mental healthworkers and researchers to the Catch22 position of psychiatric patients.Simply reading materials in this areawill be of help to some such workersand researchers. For others, directlyexperiezicing the impact of psychiatrichospitalization will be of enormous use.Clearly, further research into the socialpsychology of such total institutionswill both facilitate treatment anddeepen understanding.

I and the other pseudopatients in thepsychiatric setting had distinctly nega-tive reactions. We do not pretend todescribe the subjective experiences oftrue patients. Theirs may be differentfrom ours, particularly with the pas-sage of time and the necessary processof adaptation to one's environment. Butwe can and do speak to the relativelymore objective indices of treatmentwithin the hospital. It could be a mis-take, and a very unfortunate one, toconsider that what happened to us de-rived from malice or stupidity on thepart of the staff. Quite the contrary,our overwhelming impression of themwas of people who really cared, whowere committed and who were uncom-monly intelligent. Where they failed,as they sometimes did painfully, itwould be more accurate to attributethose failures to the environment inwhich they, too, found themselves thanto personal callousness. Their percep-tions and behavior were controlled bythe situation, rather than being moti-vated by a malicious disposition. In amore benign environment, one that wasless attached to global diagnosis, theirbehaviors and judgments might havebeen more benign and effective.

References and Notes

1. P. Ash, J. Abnorm. Soc. PsychoL 44, 272(1949); A. T. Beck. Amer. J. Psychiat. 119,210 (1962); A. T. Boisen, Psychlairy 2, 233(1938); N. Kreitman, J. Ment. Sci. 107, 976(1961); N. Kreitmsn, P. Salnsbury, J. Morrisey,J. Towers. J. Scrivener, Ibid., p. 887; H. 0.Schmitt and C. P. Fonda, J. Abnorn. Soc.Psychol. 52, 262 (1956); W. Seeman, J. Nerv.Ment. DIs. 118, 541 (1953). For an analysisof these artifacts and sunumarles of the dis-putes, see 3. Zubin, Annu. Rev. Psycho. i18,373 (1967); L. Phillips and J. G. Dragms,Ibid. 22, 447 (1971).

2. R. Benedict, J. Gen. Psychol. 10. 59 (1934).3. See in this regard H. Becker, Outsiders:

Studies In the Sociology of Deviance (FreePress, New York, 1963); B. M. Brginsky.

257

D. D. Braginsky, K. Ring, Methods ofMadness: The Mental Hospital as a LastResort (Holt, Rinehart & Winston, NewYork. 1969); G. M. Crocetti and P. V.Lemnkau. Amer. Socol. Rev. 304, 577 (1965);E. Goffman. Behavior In Pubtk Places (FrePres New York, 1964); R. D. Laing. TheDivided Sell: A StudY of Sanity and Madness(Quadrangle, Chicago, 1960); D. L PhiWps,Amer. SocJol. Rev. 28. 963 (1963); T. R.Sarbin, Psychol. Today 6. 18 (1972); E. Schur,Amer. J. Soclol. 7S. 309 (1969); T. SZasz,Law, Liberty and Psychiatry (Macmillan,New York, 1963); The Myth of Mental Illness:Foutdatlons ot a Theory o1 Mental Illness(Hoeber-Harper. New York, 1963). For acritique of some of these views, see W. R.Gove, Amer. SocIol. Rev. 3S, 873 (1970).

4. E. Gormnan, Asylums (Doubleday, GardenCity, N.Y., 1961).

S. T. J. Sceff, Bent Mentally Ill: A Soclologi-cat Theory (Aldine, Chicago. 1966).

6. Data from a ninth pseudopatient are notincorporated in this report because. althoughhis sanity went undetected, he falsified aspectof his personal history, including his maritalstatus and parental relationships. His exped-mental behaviors therefore were not Identicalto those of the other pseudopatients.

7. A. Barry, Bellevue Is a State of Mind (Har-court Brace Jovanovich, New York, 1971);I. Belknap, Human Probkms of a State MentalHospital (McGraw-Hill, New Yoric. 1956);W. Caudill, F. C. Redlich, H. R. Gilmore,E. B. Brody, Amer. J. Orthopsychiat. 22, 314(1952); A. R. Goldman. R. H. Bohr, T. A.Steinberg, Prof. Psychol. 1, 427 (1970); un-authored, Roche Report I (No. 13), 8(1971).

8. Beyond the personal difficuldes that thepseudopatient is likely to experience in thehospital, there are legal and social ones that,combined require considerable attention be-fore entry. For exampke, once admitted to apsychiatric institution, it is difficult, If notimpossible, to be discharged on short notice,

state law to the contrarY notwithstanding. Iwas not senidve to thse difficulties at theo'tet of the project, nor to the personal andsituadonal emergencies that can arise, butlater a writ of habeas corpus was preparedfor each of the entering pseudopatients andan attomey was kept "on call" during everyhospitalization. I am grateful to John Kaplanand Robert Barels for legal advice andassistance in these matters.

9. However distasteful such concealment is. Itwas a necessary first step to examining thesequestions. Without concealment, there wouldhave been no way to know how valid theseexperiences were; nor was there any way ofknowing whether whatever detections oc-curred were a tribute to the diagnosticacumen of the staff or to the hospital'srumor network. Obviously. since my con-cerns are general ones that cut across indi-vidual hospitals and staffs, I have respeaedtheir anonymity and have eliminated cluesthat might lead to their identil;ation.

10. Interestingly, of the 12 admissions, 11 werediagnosed as schizophrenic and one, with theidendcal symptomatology, as mranc.depressivepsychosis. This diagnosis has a more favorableprognosis, and it was given by the onlyprivate hospic-l in our sample. On the rela-tions between social class and psychiatricdiagnosis, see A. deB. Hollingshead andF. C. Redlich, Social Class and Mental Illness:A Community Study (Wiley, New York,1958).

11. It is possible, of course, that patients havequite broad latitudes in diagnosis and there-fore are Inclined to call many people sane. eventhose whose behavior is patently aberrant.However, although we bave no hard data onthis matter, it was our distinct impression thatthis 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. Cumming and E. Cumming, CommunityMent. Health 1, 135 (1965); A. Farina andK. Ring, J. Abnorm. PArchol. 70. 47 (1965):

H. E. Freeman and 0. G. Simmons, TheAfental Patient Comes Home (Wiley. NewYork, 1963): W J. Johannsen, Metn. Hygtene53. 218 (1969); A. S. LUnsky, Soc. Psychlat. S.166 (1970).

13. S. E. Asch, J. Abnorm. Soc. Psychol. 41, 25S(1946); Sociafl Psychology (Prentice-Halt NewYork, 1952).

14. See also 1. N. Mensh and J. Wishner, JPersonality 16, 188 (1947); J. Wislnr,Psychol. Rev. 67, 96 (1960); J. S. Bnmer andR. Tagiuri. in Handbook ol Social PsychologyG. Lindzey. Ed. (Addison-Wesley, CambridgMass.. 1954). vol. 2, pp. 634-654; J. S. Bruner.0. Shapiro. R. Tagiui, in Person Perceptioand Interpersonal Behavior, R. Tagiuri andL Petrullo. Eds, (Stanford Univ. Press, Stan-ford, CaLif., 1958), pp. 277-288.

IS. For an example of a similar self-fulfiUingprophecy. in this instance dealing with the"central" trait of inteligence, see R. Rosen.thal and L. Jacobson, Pygmalion In theClassroom (Holt. Rinehart & Winston. NewYork, 1968).

16. E. Zigier and L. PhilipS, J. Abnorm. Soc.Psychot. 63, 69 (1961). See also R. K.Freudenberg and J. P. Robertson, A.M.A.Arch. Neurol. Psychlatr. 76, 14 (1956).

17. W. Mischel, Personalty and Assessment(Wiley, New York, 1968).

18. The most recent and unfortunate instance ofthis tenet is that of Senator Thomas Eagleton.

19. T. R. Sarbin and J. C. Mancuso, J. Clin.Consult. Psychol. 3S. 159 (1970); T. R. Sarbin,ibid. 31, 447 (1967); J. C. NunnaUy, Jr.,Popular Conceptions of Mental Health (Holt.Rinehart & Winston, New York, 1961).

20. A. H. Stanton and M. S. Schwartz, TheMental Hospital: A Study of InstitutionalParticipation in Psychiatrtc Illness and Treat.ment (Basic, New York, 1954).

21. D. B. Wexler and S. E. Scoville, Arlz. LawRev. 13, 1 (1971).

22. 1 thank W. Mischel, E. Ome, and M. S.Rosenhan for comments on an earlier draftof this manuscript.

NEWS AND COMMENT

AAAS Council Meeting: VietnamResolutions; Bylaws Voted

In an unprecedented expression ofpolitical sentiment, the governing coun-

cil of AAAS adopted a strongly wordedresolution in its business meeting of 30December condemning the UnitedStates' continued involvement in theVietnam war and the application ofAmerican science and technology tothe "wanton destruction of man andenvironment."The council passed a second war-

related resolution urging Congress tosupport a major study, by the Na-tional Academy of Sciences, of thewar's impact on the people and the

environment of Indochina. At the same

time, the council in effect voted itsown termination by approving a newand much-discussed set of bylaws thatwill drastically reduce the size of thecouncil and allow the general member-

258

ship of the AAAS to elect it. TheAAAS thereby completed what formerchairman of the board Mina Rees andchief executive officer William Bevancalled "a major step toward becominga genuine membership organization."The council's antiwar resolution was

the first in which the AAAS has takenan unqualified stand in opposition toU.S. military involvement in Vietnam.Past councils have limited themselvesto expressions of "concern," particular-ly about the adverse effects of defoli-ants.

This year's bluntly phrased resolutionwas introduced as an "emergency mo-

tion" by seven council delegates, includ-ing Everett Mendelsohn, a Harvardhistorian of science and a AAAS vicepresident, and E. W. Pfeiffer, a Univer-sity of Montana zoologist who was in-

strumental in arousing the association'sinterest in the herbicide issue severalyears ago.

During a brief debate, the resolutionwas modified slightly at the suggestionof Lewis M. Branscomb, the formerhead of the National Bureau of Stan-dards and now the IBM Corporation'schief scientist. Branscomb urged thattwo critical references to U.S. militaryactivity in Thailand be deleted, on thegrounds that the American presencethere was not analogous to U.S. involve-ment in Vietnam. The council con-sented, and the midified resolution car-ried by a vote of 80 to 41 with a largebut uncertain number of abstentions,including those of Glenn Seaborg, theformer chairman of the Atomic EnergYCommission, and others seated at thedais. Only about 170 of the council'sapproximately 530 members were pres-ent.The full text of the resolution is as

follows:

The Council of the AAAS condemlulthe United States' continued participatiotlin the war in Vietnam, heightened inthe post-election bombing escalation.As scientists we cannot remain sid""

SCIENCE, VOL. 194i