Vulnerability: A new view of schizophrenia.

Journal of Abnormal Psychology1977, Vol. 86, No. 2, 103-126

Vulnerability—A New View of Schizophrenia

Joseph ZubinColumbia University and New York State

Psychiatric Institute, BiometricsResearch Unit

Bonnie SpringHarvard University and New York State

Psychiatric Institute, BiometricsResearch Unit

Although descriptive and etiological approaches to psychopathology have madenotable advances, they seem to have reached a plateau. After reviewing the sixapproaches to etiology that now preempt the field—ecological, developmental,learning, genetic, internal environment, and neurophysiological models—a sec-ond-order model, vulnerability, is proposed as the common denominator, andmethods for finding markers of vulnerability are suggested in the hope of re-vitalizing the field. It is assumed that exogenous and/or endogenous challengerselicit a crisis in all humans, but depending on the intensity of the elicited stressand the threshold for tolerating it, that is, one's vulnerability, the crisis willeither be contained homeostatically or lead to an episode of disorder. Vulner-ability and episode stand in a trait-state relation, and markers for each mustbe provided to distinguish between them.

The purpose of this article is to briefly re-view some of the classic approaches to under-

An abbreviated version of this paper was presentedby Zubin at the occasion of the Distinguished Scien-tist Award given by Section III (Section for theDevelopment of Clinical Psychology as an Experi-mental Behavioral Science) of Division 12 (Divisionof Clinical Psychology) of the American Psychologi-cal Association, Chicago, Illinois, August 30, 1975.

Portions of this article were also presented bySpring and Zubin at the First Vermont Conferenceon the Primary Prevention of Psychopathology,Burlington, Vermont, June 197S.

The preparation of this article has been partiallysupported by National Institute of Mental HealthGrant MH 27342, "Vulnerability to SchizophrenicEpisodes in Adults," U.S. Public Health Service.While working on this article the second author wassupported for varying periods of time by a New YorkState Department of Mental Hygiene traineeship, anNIMH predoctoral iellowship, and a Scottish RiteDissertation Research fellowship.

The authors are grateful for some of the materialincluded in this article to the following members ofthe Biometrics Research Unit in 197S: Muriel Ham-mer, Anthropology; Kurt Salzinger, Behavior Analy-sis and Modification; Joseph L. Fleiss, Biostatistics;Barry Gurland, Diagnosis and Psychopathology;Robert L. Spitzer, Evaluation; Denise Kandel, Familyand Youth Research; Ruth Bennett, Gerontology;Samuel Sutton, Psychophysiology; and David Wilder,Sociology.

Bonnie Spring is now at the Department of Psy-chology and Social Relations, Harvard University.

Requests for reprints should be sent to JosephZubin, who is now at the Veterans AdministrationHospital, Highland Drive 1S1R, Pittsburgh, Pennsyl-vania 15206.

standing schizophrenia and to indicate theemerging trends. This overview suggests thatafter tremendous strides made in the lastdecade in both description and etiology, thestudy of schizophrenia has come to a stand-still. Each of the various approaches seemsnow to provide only a partial answer. Perhapsfor this reason, many proponents of eachmajor theory indicate that they have gone asfar as they can go with present methods andinsights. A new, fresh view of the entire sub-ject seems to be necessary if we are to moveahead. We propose that such a view emergeswhen we look at the common elements sharedby all of the different theories.

One major stumbling block in studyingschizophrenia has always been the tendencyto intermesh description or phenomenologywith etiology, as was done in the Diagnosticand Statistical Manual (DSM II) of theAmerican Psychiatric Association (1968). Weshall try to keep these two perspectives apartand begin with the area of description.

Descriptive Psychopathology

Written descriptions of the disordered be-havior fitting our current category of schizo-phrenia have been available ever since theHindu elders' observations in the Ayur-vedasome 34 centuries ago. Moreover, such dis-orders were probably known to mankind longbefore they were recorded. We shall, however,

103

104 JOSEPH ZUBIN AND BONNIE SPRING

skip over the centuries and begin with thegreat systematizer of mental disorders—EmilKraepelin, who combined a variety of discretesymptoms (catatonia, hebephrenia, and para-noia) into the disease category of dementiapraecox. Kraepelin distinguished dementiapraecox from manic depressive psychosis onthe basis of ultimate deterioration in cases ofthe former. The Kraepelinian tradition per-sists today and is found in more elegant formin the concepts of "true" schizophrenia de-scribed by Feighner et al. (1972) and by Kety(Note 1). Spokesmen for this position main-tain that a definite diagnosis of schizophreniacan be made only if symptoms persist chroni-cally and if a deteriorating, unremitting dis-ease course is followed. True schizophrenia isthus considered to be synonymous with poorprognosis schizophrenia.

Departing from the Kraepelinian view, E.Bleuler (1911/1950) eliminated poor prog-nosis as a criterion and specified instead ahierarchy of primary symptoms (autism, asso-ciation and affective disturbances, and am-bivalence) and secondary symptoms (halluci-nations, delusions, negativism, and stupor).This second tradition, giving primacy tosymptomatology rather than to disease course,has been further developed by Schneider(1959), Langfeldt (1939, 1956), the Bio-metrics Research Unit (Gurland, cited inZubin et al., 1975; Spitzer & Endicott, citedin Zubin et al., 1975), the WHO Pilot Studyof Schizophrenia (World Health Organization,1973), and by cognate groups. These groupshave specified the psychopathological char-acteristics of schizophrenic behavior along thewell-known dimensions of perceptual dysfunc-tion, speech disorder, delusions, hallucinations,affective and cognitive dysfunction, and so on.However, these classic symptoms do not inthemselves indicate a particular diagnosis,since they tend to be manifested by a diverseand heterogeneous group of patients. To re-duce this heterogeneity and yet save thesymptom or syndrome approach, Feighner etal. (1972) and subsequently Spitzer andEndicott (cited in Zubin et al., 1975) haveproposed operational definitions of schizo-phrenia and the other mental disorders for thenew Diagnostic and Statistical Manual of the

American Psychiatric Association (DSM III,in press). Each disorder is operationally de-fined by explicit criteria. These criteria takethe form of sets of characteristics, a specifiednumber of which must be present before agiven diagnosis is warranted. Because the spe-cific psychopathological features associatedwith each official DSM III diagnosis can nowbe explicated, the reliability of diagnostic de-cisions can be enhanced (Spitzer & Endicott,cited in Zubin et al., 1975) and the hetero-geneity that characterized the DSM II diag-nostic categories can be greatly reduced.

Diagnostic improvements became possiblewith the availability of systematic structuredinterviews to replace the freewheeling clinicalinterviews that had previously preempted thefield. Structured interviews provide the greatadvantage of reliable and comprehensive cov-erage of all areas of psychopathology and be-havior germaine to a set of diagnostic cate-gories. These interviewing methods havedemonstrated their value in two major studies—the US-UK Diagnostic Project (Cooper etal., 1972) and the WHO Pilot Study inSchizophrenia (World Health Organization,1973). The US-UK project was initiated todetermine why the national statistics showsuch disproportionate frequency of affectivedisorders in the United Kingdom and ofschizophrenia in the United States. When thenewly developed systematic structured inter-views (Gurland, cited in Zubin et al., 1975)were applied to samples of patients admittedto hospitals in the two countries, the cross-national differences turned out to reflect dif-ferent diagnostic practices of psychiatristsrather than differing characteristics of patientsin the two countries. Whereas the US-UKproject limited itself to two cultures but in-vestigated the entire panoply of mental dis-orders, the WHO study limited itself to onedisorder but investigated its incidence, form,and course in nine different cultures. This in-vestigation found specific syndromes of schizo-phrenia ubiquitously distributed in developingand advanced cultures from Ibadan, Nigeria,to Washington, D.C.

Other advances in descriptive psychopathol-ogy have been (a) the use of computers thatsimulate the clinician's decision processes and

VULNERABILITY—A NEW VIEW OF SCHIZOPHRENIA 105

arrive at a diagnosis (Spitzer & Endicott,cited in Zubin et al., 1975; Wing, Cooper, &Sartorius, 1974); (b) the provision of mathe-matical methods for clustering individualswith shared psychopathological characteristicsinto more homogeneous subgroups (Fleiss &Zubin, 1969; Wing & Nixon, 1975); and (c)the development of behavior-analytic descrip-tions of patient characteristics (Kanfer &Saslow, 1969; Salzinger, cited in Zubin et al.,1975). Rather than emphasizing the presenceof symptoms per se, behavior-analytic descrip-tion relates deviant behavior to environmentalcontingencies that initiate and sustain it. De-scription is therefore focused on aspects of theindividual and his environment that are ofimmediate relevance to behavior modificationtherapy.

A dispassionate view of where we stand indescriptive psychopathology today leads tothe conclusion that we have made considerableprogress and perhaps gone as far as we can gobut that this is still not far enough. Throughthe use of systematic structured interviewsand operational criteria for selecting a diag-nosis, a high degree of agreement on diagnosiscan be attained. We can now objectively de-scribe the characteristics of patients we dealwith so that replication of basic research andtreatment findings on similar patients is pos-sible. Moreover, we can demonstrate that indi-viduals suffering from such syndromes of psy-chopathology as schizophrenia can be foundin parts of the world that differ widely in cul-tural and ecological conditions.

Where does this leave the allegation thatschizophrenia is a myth? We must bear inmind that science deals not only with osten-sive facts but also with mythlike concepts orfictions that are used to organize the facts.We can inquire whether the facts are true,but we cannot inquire whether the conceptsare true—we can only ask if they are useful.To the extent that trained psychologists andpsychiatrists all over the world can consensu-ally recognize the syndrome that we callschizophrenia and find this summary descrip-tion of psychopathological behavior beneficialfor selecting treatment strategies and inter-preting research findings, schizophrenia is stilla useful concept.

However, we cannot presuppose that re-liable description of a syndrome implies validunderstanding of its underlying cause. Weneed to do more than describe, since descrip-tion never cured anyone, nor did it by itselfreveal etiology. Although treatment can oftennot await a thorough knowledge of etiology,the discovery of an effective cure for any dis-order is less likely to result from chance thanfrom an understanding of factors that causethe disorder. As in the classification of or-ganisms, in which taxonomy looks to commonorigins of descent, so in the classification ofdisorders we look for common causes produc-ing the disorder. But how can progress bemade, when with but few exceptions, we areabysmally ignorant of the causes of mentaldisorder and even more poignantly ignorantof the efficacy of the treatment? When facedwith such ignorance, one can only contemplatepossible or "as if" causes, formulate them intoparsimonious scientific models, and proceedto test the hypotheses they generate.

Scientific Models of Etiology

What models have been proposed? The sixmodels depicted in Figure 1 have been de-scribed in detail elsewhere (Zubin, 1972) withevidence on the tenability and the limitationsof each. Here we will only present a shortoverview of all six models and briefly examinethe status of two of them—the genetic andthe ecological.

The etiological models can be classified intothe following three types in accordance withtheir etiological focus: (a) field theory modelsfocusing on forces emanating from the or-ganism's external environment—the ecologicalniche it occupies—(b) behavioral psychologi-cal models focusing on forces emanating fromthe experience of the organism through learn-ing and development; and (c) biologicalmodels focusing on forces emanating from theorganism's internal milieu—its genes, bio-chemistry, and neurophysiology.

One type of field theory model is the eco-logical approach, which sees man's health orillness as dependent on the physical, social,cultural, educational, and economic parame-ters of the ecological niche he occupies. Thedevelopmental model is concerned with exog-


LEARNING

AND

DEVELOPMENT

ENVIRONMENTALFIELD FORCES

SOCIAL- P8YCHOLOOICALEXPERIENCES

Figure 1. Scientific models of etiology.

INTERNALBIOLOQICAL FORCES

enous and endogenous factors that affectman's progression through maturationalphases and hence contribute to either buffer-ing or fostering the development of a pro-pensity to become mentally ill. The learningtheory model has been most concerned withcontingencies that elicit and sustain psycho-pathological behavior. This model differs fromthe developmental model in generally assign-ing more importance to current rather thanhistorical causes of behavior. It also tends toeschew investigation of how environmentalinfluences might act differentially on individu-als with different propensities.

, From a more biological perspective, thegenetic model suggests that health and illnessare predicated on the genetic equipment manis born with. The internal environment modelstipulates that the roots of man's illnesses areto be sought in his metabolism, body fluids,and body chemistry in general. The neuro-physiological model concentrates on the func-tioning of man's nervous system and its ca-pacity to take in and process information.

Genetic Model

The genetic model languished for a timebecause some of its claims to high concordancein monozygotic twins were deflated during the1960s. However, the separation of rearing

from hereditary influence in the studies ofadoptees (e.g., Kety, Rosenthal, Wender, &Schulsinger, 1968) and the monitoring of thedevelopment of high-risk children (e.g., Med-nick & Schulsinger, 1968) added new dimen-sions to genetic research. Findings from suchstudies have been exciting and striking. How-ever, they have also posed unforeseen chal-lenges to the genetic model. Compared to the85%-concordance rates found by Kallman(19S9) in monozygotic twins, it has beensomewhat disappointing that current studiesfind only 40%-60% 1 of monozygotic twinsconcordant for schizophrenia (Slater & Cowie,1971, pp. 18-19). Moreover, even when theeffects of genetic predisposition are added tothose of rearing by a schizophrenic parent, asis the case for home-reared offspring of twoschizophrenic parents, only 60% of their off-spring are likely to become schizophrenic. Thecurrent inability to explain or predict whichgenetically predisposed individuals will ulti-mately become schizophrenic has forced us toconfront our ignorance of the precise mecha-nism and mode of transmission of this dis-order.

i Recent studies (Slater & Cowie, 1971, p. 19) havefound only 27% concordance when both twins sufferfrom "strict" schizophrenia.


An additional problem has arisen from thefinding that offspring and relatives of schizo-phrenics may show a spectrum of behaviorranging from normalcy through the schizoidspectrum and from psychopathy to core schiz-ophrenia. The discovery of such a schizo-phrenic spectrum has a number of difficultimplications. Taken to its extreme, it maynullify the usual procedures for comparingschizophrenic patients with normal controlson specific biochemical or other markers, sincethe control group may contain an unknownproportion of unexpressed genotypes for schiz-ophrenia. Moreover, the spectrum findingraises the old question of the most appropriatemeans of conceptualizing the schizophrenicprocess. Should we adopt a dimensional ap-proach in which the disorders in the spectrumare seen as falling along a quantitative con-tinuum of severity? Or should we select atypological approach that delineates particularsubcategories based on different etiologies?

Many prominent spokesmen for the geneticmodel now seem to be adopting a typologicalapproach, which postulates that only somecases or types of schizophrenia arise fromgenetic proclivity:

Several studies taken together clearly indicate thatmany schizophrenics have an important genetic con-tributant. Further studies of this sort may moreclearly define the boundaries of the spectrum, butwill not elucidate its mechanism. Comparing the situ-ation to that of studies of retardation in the nine-teenth century, one must expect that further progresswill be made only by attempting to define homo-geneous subgroups, some of which have specific bio-logital etiologies, and exploring the biological attri-butes of those subgroups.

These adoption studies may be compared crudely tothat in which a kite was flown and an electric sparkwas discovered. They are provocative and, I think,exciting; but one can anticipate further progress onlyon the basis of careful study of electricity and notby the use of bigger and higher flying kites. (Wender,Rosenthal, Rainer, Greenhill, & Sarlin, in press)

Ecological Model

The ecological approach seeks the causes ofschizophrenia in the parameters of the en-vironmental niche that a person occupies.Cassel (1974) has pointed out that throughouthistory the environmental factors singled outfor study have reflected both the reigning

etiological theories and the existing level oftechnology. Thus the airs, waters, and placesbelieved to be etiological agents in Hippocratictimes have been succeeded by the microchemi-cals and microorganisms of today. Ironically,old etiological theories have often been per-manently preserved in the names of variousdiseases, for example, malaria ("bad air"),hysteria ("wandering uterus"), and melan-cholia ("black bile").

Some ecological parameters that are cur-rently proposed as causes of mental disorderinclude (a) low socioeconomic status; (b) dis-organization in the social milieu; (c) crowd-ing; and (d) minority status in the com-munity. Adequate investigation of theseparameters has been hampered in severalways. First, it has always been exceedinglydifficult to know whether an individual's situ-ation in an undesirable niche causes a disorderor results from a disorder. Does a high rate ofillness associated with an ecological character-istic mean that the environmental factor fa-cilitates the likelihood of a disorder? Does itmean that the social system affords especiallyefficient means of recognizing and selectingfor patienthood disordered individuals whoappear in a certain environmental context?Does it mean that schizophrenic individualsprefer and choose an environment with certaincharacteristics? Does it mean that schizo-phrenics are forced to inhabit certain typesof ecology as a consequence of their illness?To date there has never been an adequate wayof choosing among the competing hypotheses,although B. P. Dohrenwend and B. S.Dohrenwend (1969) have proposed some in-genious research strategies to test these issues.

Two other major stumbling blocks for theecological hypothesis have been (a) the lackof an adequate taxonomy for parameters ofthe ecology that influence good and badhealth and (b) the inability to do fully con-trolled studies on these parameters. No doubtthe effect of any ecological factor on health isinfluenced by a host of interactions withmoderator variables. For example, we knowthat population density takes a totally differ-ent toll depending on whether it involves over-crowding among strangers or among familiarpeers. It is likely that the link between popu-lation density and high risk of disorder is


mediated by the disorganized relations thatensue between individuals. The impact of adisordered relation on mental health is inturn dependent on (a) the importance orsalience of the relation that becomes disor-dered (e.g., mating, rearing of offspring); (b)the position of the individual experiencing thedisordered relation in the status hierarchy;(c) the degree of previous experience withsuch disorganization; and (d) the nature andstrength of available group support. But inthe final analysis, progress has been preventedby our inability to measure these factors ob-jectively or to find variables that transcendlocal or temporal biases.

Other Models

The developmental model, with its emphasison rearing, has also failed to provide theanswers we are seeking. Rosenthal et al.(1975) maintain that childhood rearing is lessimportant in the production of psychopathol-ogy than is genetic loading. M. Bleuler (1974)finds that childhood neglect and deprivationof the most dire type have little or no bearingon the development of schizophrenia.

Nearly all geneticists gesture graciouslytoward the developmental model by acknowl-edging that genes do not work in a vacuumbut rather require circumstances to elicitgenetically encoded propensities. But neithergeneticists nor developmentalists have beenable to specify the circumstances that are re-quired. Attempts have been made to delineateparticular family interaction patterns thatmight prove schizophrenogenic (e.g., Bateson,Jackson, Haley, & Weakland, 1956). How-ever, Jacob's (1975) careful review of thisliterature has shown how difficult it is todemonstrate that these patterns occur withany greater frequency in the families of dis-turbed compared with normal offspring.Rosenthal et al. (1975), commenting on thecomplexity of family interaction research,wrote the following:

The familial behavior must be examined over ap-preciable stretches of time. Reliability of behavioralcoding must be determined, and the selection of be-haviors to be observed will have to be limited. Andeven with the best of circumstances, investigatorswill still not be able to determine from the usual

high-risk study how much of the behaviors notedstemmed from familial genes or from mutual in-fluences of parents and children, (p. 476)

Thus, here too we have reached an impasse.The learning theory model has been pri-

marily concerned with the contingencies thattrigger a schizophrenic episode in an indi-vidual who has already developed a propensitytoward schizophrenia. Beyond studying rein-forcements that elicit and maintain schizo-phrenic behavior, this model has generally notdealt with basic causes of schizophrenic be-havior. Thus it has not explained why eventsand reinforcements that catapult one indi-vidual into an episode of schizophrenia leaveanother individual unaffected.

Despite the enthusiasm recently rekindledby Kety and his group (Kety et al., 1968),the internal environment model is still onlypromising rather than delivering. And theneurophysiological model, with its stress oninformation processing, has hardly gotten offthe ground.

Although space does not permit us to fullydescribe and analyze the contributions ofthese models, our general conclusion is that aformidable impasse has been reached. Howcan we now transcend it and start movingagain? It would be foolhardy to claim thatwe have found the definitive answer. We have,however, decided to take a fresh view and seeif a new approach is possible.

Vulnerability Model

Progress rarely precedes an understandingof historical pitfalls. Therefore, it is importantto try to determine the general reasons why allsix scientific models seem to have faltered.Perhaps the main difficulty is that each modelis framed so broadly that entire schools ofpsychopathology can pass through its portalswithout even rubbing shoulders. Just as nomodel has been in danger of being refuted byfindings that support another, neither has itbeen enriched by the interaction of differingviews. The trend for parochialism seems un-fortunate, since for too long it has permittedproponents of each model to exploit their ownresparch without noting the progress of others.We feel that it may now be useful to examine


the various models to see whether the partsmight illuminate the whole.

Our strategy is essentially pragmatic. Weare trying to pool the wisdom from all of themodels to analyze out a common factor thatcan best guide research across all the ap-proaches. Put more picturesquely, we shallsqueeze the juice out of all these models intoa goblet and see what the elixir consists of.This logical factor analysis suggests that thecommon denominator connecting all six mod-els is a second-order model that can be calledthe vulnerability model.

The vulnerability model proposes that eachof us is endowed with a degree of vulnerabilitythat under suitable circumstances will expressitself in an episode of schizophrenic illness.Each etiological model offers suggestionsabout the possible origins of such vulnerabil-ity. Moreover, many earlier publications(Gottesman & Shields, 1972; Meehl, 1962;Millon, 1969; Rosenthal, 1970) have antici-pated the vulnerability notion in discussionsof the concept of a schizophrenic diathesis.There are, however, three somewhat novelelements in our formulation. First, we ac-knowledge numerous contributions to an indi-vidual's degree of vulnerability, ranging fromhis genetic inheritance to his acquired pro-pensities. Second, we are concerned with mea-suring vulnerability. That is, we are attempt-ing to capture empirically those characteristicsof individuals that might predict the proba-bility that a schizophrenic episode will de-velop. Finally, we distinguish between vulner-ability to schizophrenia, which we regard as arelatively permanent, enduring trait, and epi-sodes of schizophrenic disorder, which arewaxing and waning states.

Further on in this article we will have moreto say about the episodic nature of schizo-phrenia. At the moment we will be concernedsimply with the concept of vulnerability. Letus preface our discussion by noting thatwhether such vulnerability extends to all ofmankind and whether it is the same sort thatpredisposes an individual to disorders otherthan schizophrenia remains an open question.But even if only one or two percent of hu-manity is capable of becoming schizophrenic,this still represents 30 million of the world'spopulation.

Corresponding to the two types of etiologi-cal models—the biological and the field theory—there are two major components of vulner-ability, the inborn and the acquired. Else-where (Spring & Zubin, in press-b) we havedescribed inborn vulnerability as that whichis laid down in the genes and reflected in theinternal environment and neurophysiology ofthe organism. The acquired component ofvulnerability is due to the influence of trau-mas, specific diseases, perinatal complications,family experiences, adolescent peer interac-tions, and other life events that either enhanceor inhibit the development of subsequentdisorder.

As Audy (1971) suggested, the preserva-tion of health requires the maintenance of adynamic equilibrium against insults continu-ally emanating from the chemical, physical,infectious, psychological, and social environ-ment. When this equilibrium is disturbed be-yond its capacity to reinstate its own ho-meostasis, a disorder ensues. An individual'svulnerability to any illness determines the easeand frequency with which suitable challengesto homeostasis will catapult him into that dis-order. The highly vulnerable person is one forwhom numerous contingencies encountered indaily living are sufficient to elicit an episode.Others have such a low degree of vulnerabilitythat nothing short of a rare and probablycatastrophic event would induce an episode,and even then only a very brief one.

The "challengers" that disrupt adaptationand are capable of provoking a temporarycrisis or even an episode of illness may beendogenous or exogenous in origin. The endog-enous events may be either biochemical orneurophysiological. Relatively little is knownof what causes such internal events, but thepossibilities are numerous: maturationalchanges within the organism, ingestion oftoxic substances, inadequate nutrition, patho-logical responses to infection or to stress, andso on. By and large, endogenous precipitantsof homeostatic disturbance are difficult tostudy without advanced technology, since theyare generally unreported and often unper-ceived by the individual. Exogenous chal-lengers are usually referred to as life events.We shall limit our discussion to these exog-enous stressors, not because they are more


MAXIMUM

MINIMUM

THRESHOLD-

WELL

LOW

VULNERABILITYHIGH

Figure 2. Relation between vulnerability and chal-lenging events.

important than endogenous challengers, butbecause they are more directly measured andhave been extensively studied.

There is considerable evidence that lifeevent stressors can play a major role in elicit-ing both physical and mental disorders(Brown, 1968; Dohrenwend & Dohrenwend,1972; Rahe, Meyer, Smith, Kjaer, & Holmes,1964). A life event stressor is an incidentsuch as bereavement, promotion, marriage, ordivorce that challenges adjustment and bringsin its wake a readjustment or reorganizationof a person's life. It is easiest to regard lifestresses as events that befall an individual re-gardless of his temperament or behavior. How-ever, it is interesting to speculate on the ex-tent to which life stressors really do occur asrandom happenstances. With the exception ofsuch natural calamaties as earthquakes, itmay be that a person's choice of life stylecontributes to his likelihood of encounteringstressful events. Thus it may be possible toidentify "stress-prone" patterns of living, justas we can discuss "accident-prone" indi-viduals. However, for the purpose of concep-tual clarity, we shall oversimplify and discusslife events as a dimension orthogonal to theindividual's other characteristics.

Figure 2 shows the hypothesized relationbetween life event stressors and vulnerability.As long as the stress induced by the challeng-ing life event stays below the threshold ofvulnerability, the individual responds to thestressor in an elastic homeostatic way andremains well within the limits of normality.When the stress exceeds threshold, the personis likely to develop a psychopathological epi-sode of some sort. Further, we postulate that

the episode is time limited. When the stressabates and sinks below the vulnerabilitythreshold, the episode ends and the patientreturns to a state similar to his pre-episodelevel of adaptation.

Our model is Selyean (Selye, 1973), inwhich life event stressors impinging on theorganism induce a state of strain. The strainin turn sets in motion adaptive capacities toovercome the stress or contain the strain. Weconcur with Mechanic's (1967) definition ofstress as involving "a discrepancy between thedemands impinging upon a person—whetherthese demands be external or internal, whetherchallenges or goals—and the way the indi-vidual perceives his potential responses tothese demands" (p. 201). As Wilkins (1974)and Gross (1970) suggest, strain occurs as aresult of "the failure of routine methods formanaging threats" (Gross, 1970, p. 55). Inbrief, stressful life events tax the organism'sadaptive capacities. But of what does adapta-tion consist?

Adaptation

Adaptation describes the extent to whichan organism responds adequately and appro-priately to life's exigencies. It is a biologicalconcept with a long history and tremendousliterature that cannot be summarized here.However, it may be useful to break downadaptation into its components to use the con-cept in our discussion of vulnerability. Asimplified model of adaptation may be bor-rowed from Hooke's law and Young's modulusin physics. This model deals with the stressimposed on a body by a load that produces ameasurable tensile strain. The classic exampleinvolves a string suspended from the ceilingto which a weight has been attached. As aresult of the load, the string may (a) with-stand the load without lengthening, thoughheat may be generated; (b) temporarilylengthen but return to its original length oncethe load is removed; (c) lengthen perma-nently; or (d) break under the load.

To extend the analogy, consider the stringto represent an organism and the weight torepresent a load imposed by a life eventstressor. The string's tendency to stretchmight be likened to the biological and


Piagetian concepts of accommodation, whichdescribe a temporary or a permanent struc-tural alteration of the organism to suit en-vironmental demands. Organisms, unlikestrings, have another way of responding toload, by adjusting it to suit their tolerancelimits. This pattern of altering environmentalpressures without altering the organism's levelof homeostasis might be called assimilation.It is analagous to the process by which de-partures from a predetermined level of heator cold activate a thermostat to regulate thetemperature of the surrounding room. Whena stressful life event threatens to impose toodamaging a strain, organisms may actually ad-just the stressful impact by avoiding, rein-terpreting, or distorting the event.

In applying the load-strain model borrowedfrom physics to the study of human behavior,we must exercise some caution. As Ayala andDobzhansky (1975) have written, the progressof evolution can be regarded as a gradualdeparture from the importance attached tophysicochemical laws as the sole determinersof behavior. We can only loosely specify thecorrespondence between physical and behav-ioral parameters. The intensity of a loadcorresponds roughly to the normatively per-ceived severity of a life event stressor, as as-sessed by Holmes and Rahe (1967). However,as Brown (19,68) has pointed out, it is anindividual's own perception of the stressful-ness of an event that ultimately defines theseverity of the load. Corresponding to thephysical concept of a measurable tensilestrain, an overt strain or displacement of thecharacteristics of an organism is induced bythe stressor. Strain in an organism may bemeasured by any number of response systems,including physiological response parameters,subjective reports of emotion, and overt be-havioral change. Moreover, it is doubtful thatall response systems will be identically af-fected by any stressful event.

The attempt to predict the individual's re-sponse to a stressor brings us to the limits ofthe reductionistic approach. The followingvariables are some that must be taken intoaccount: (a) the normatively perceived se-verity of the life event stressor, (b) the indi-vidual's perception of the stressfulness of the

load, (c) the capacities or general competencelevel of the individual, (d) the coping effortsexerted in dealing with the stressful situation,and (e) the vulnerability of the individual.

What progress has been made in predictingthe response to stressful situations? What de-termines whether adaptation will take theform of assimilation or accommodation? Whenwill an episode of illness develop? We shalllimit our discussion to two possible models ofadaptation to stress—Murphy's (1974),which is primarily oriented to childhood, andFrench and Steward's (1976), which describesmore mature stages of development.

Murphy (1974) proposes that adaptation

involves (1) reflexes (built-in mechanisms) and in-stincts (broader built-in patterns); (2) coping ef-forts (to deal with situations not adequately man-aged by reflexes); (3) mastery resulting fromeffective and well-practiced coping efforts; (4) com-petence as the congeries of skills resulting fromcumulative mastery achievement, (p. 77)

Reflexes are exemplified by such responsesas the dilation of the pupil in darkness, thetonic-neck reflex of the baby, and so on. In-stincts, or broader built-in or developed pat-terns, are exemplified by nest building in ani-mals and possibly by some aspects of maternalbehavior in humans. To the extent that re-flexes and instincts do not go far in dealingwith the demands, threats, and opportunitiesthat life offers, man must engage in copingefforts. Coping efforts consist of the constantinvention and incessant trial-and-error stabsat adapting that supplement the limited, ifimpressive, repertoire of reflex capacities. Thecoping process may lead to direct solutions ofa presented problem or to circumventions viadefense mechanisms. Coping efforts are ex-emplified by the persistent application ofenergy toward problem solving and abstractthinking in situational dilemmas. Competenceis developed by exerting coping efforts andconsists of social skills, intellectual strategies,and other acquired capacities that equip theindividual to deal with life exigencies.

Coping efforts should not be confused withcompetence. The former refers to the atti-tudinal, motivational stance of an individualfaced with a task; the latter refers to his abili-ties, skills, and accumulated know-how insolving life problems. Whereas coping repre-


sents the motive power of an organism, com-petence represents its capacities. Coping ef-forts are analogous to the voltage of an elec-tric current that sets a motor into operation.The motor may be intact in every respect, andits interconnection with the machine may beunchanged, but unless the voltage is main-tained within proper limits the machine willnot operate adequately. Competence, on theother hand, is analogous to the capacities ofthe machine, the functions for which it wasequipped. If the machine is applied to a jobfor which its circuitry is inadequate, no workwill be accomplished no matter how adequatethe power supply.

In the growth of the individual, developingcapacities are put to use in efforts to cope withroutine daily events. Well-practiced copingstrategies crystallize as effort and competenceare rewarded with success in life. However,when challenging new situations arise, routinecoping strategies may be inadequate and newones may be needed. If adaptation falters insuch circumstances, it will be very difficult todetermine whether the fault lies in copingefforts that are sluggish or overzealous or inan inadequate repertoire of competencies.

Some encounters with stress-producingevents are routinized by cultural sanctionsthat proscribe a wide range of institutionalforms for coping. In many cultures, for ex-ample, extreme and "deviant" forms of griefare tolerated and expected; however, not allforms of stressful challenges are matched withinstitutionalized forms to channel coping be-havior. Lazarus, Averill, and Opton (1974)have distinguished between the

healthy, positively adaptive problem-solving effortscharacteristic of low-stake situations and the primi-tive and maladaptive forms of coping typically as-sociated with conditions of high stake (that is,severe threat, frustration, conflict and great chal-lenge), (p. 30S)

Most of us, if observed closely, probablydisplay some disturbance of coping when cata-strophic situations arise. In the resilient per-son the "mini-episode" of coping breakdownpasses, and after a temporary period of dis-tress, routine coping strategies regain the field.In a highly vulnerable person, however, atemporary breakdown in adaptation may pro-vide an occasion for more fundamental prob-

lems to manifest themselves. The comparisonmay be clarified by an analogy. If a healthyperson overexerts himself by running a mara-thon, he will be stressed, flushed, breatheharder, and be very tired. The next day, asidefrom some sore muscles, he will have re-covered. However, a person who has recentlysuffered a heart attack will probably not re-cover so easily. The disturbances directly re-lated to running the race may soon vanish, butdifficulties due to his weakened heart mayonce again be manifest. So, in the personvulnerable to schizophrenia, episodes of psy-chiatric disorder may often follow close onthe heels of periods of acute stress and copingbreakdown. We need not conclude that copingbreakdown causes schizophrenia, but only thatit provides an opportunity for vulnerabilityto germinate into disorder.

This relation between deviant coping pat-terns and the development of a schizophrenicepisode goes back historically to AdolphMeyer (Arieti, 1974, p. 16). He became con-vinced by his longitudinal studies thatdementia praecox results from an accumula-tion of faulty habits of reaction to life'sexigencies. Confronted with failure after fail-ure, the patient gradually develops substitu-tive reactions instead of efficient adjustmentto actual difficulties. These coping failuresthemselves do not constitute an episode ofschizophrenia, but according to Meyer, thetransition from maladaptation to schizo-phrenia is continuous and insidious. Thus,maladaptation, with its coping failures, mayprepare for either the gradual transition to aschizophrenic episode or set the stage for asudden onset of an episode following a cata-strophic event. Intervention at this point, be-fore an episode develops, may be most propi-tious for preventive therapeutic efforts.

French and Steward (1976) have presenteda model of adaptation to a stressor that isless developmentally oriented than Murphy's(1974). They describe successful adaptationas a balance between accommodation and as-similation and suggest that a feedback loopfor maintaining homeostasis might consist ofthree components. The first is a force F hav-ing magnitude but no direction and experi-enced subjectively as an "ought." Althoughthe individual may or may not know the ob-


Table 1Patterns of Response to Stressors in Relation to Components of Adaptation and Its Outcome

Pattern

Components of adaptation

Coping effort Felt discrepancy Competence Outcome

1 +2 +3 +4 +567

—

+ + joy+ — frustration-anxiety— + obsessive behavior— — obsessive rumination+ + inhibition+ — ineffectual complaints— + unstressed— — passivity (?)

Note. Table adapted with modifications from French and Steward (1976, p. 469, Table 1).Plus = presence of component; minus = absence.

ject or survival value of the "ought," he feelsa drive to do something about correcting animbalance. The "ought" therefore correspondsto the subjective triggering event for a copingeffort. The second component is a cognitiveperception of the discrepancy between theobserved stressed state of the organism, itsprior state, and the state of adaptation towardwhich it strives. The felt discrepancy corre-sponds to the organism's ability to identifythe disparity between its own present stateof distress and the remembered or imaginedhappier circumstances. It helps define a direc-tion for coping efforts to return the organismto its original state or to some other adaptivestate. Put simply, the felt discrepancy entailsthe identification of "what's wrong." The thirdcomponent is the biological and/or cognitivestructural capacity to correct the felt dis-crepancy. This corresponds to our concept ofcompetence and refers to the repertoire ofskills and capacities at the organism's disposalto formulate coping strategies that will re-attain adaptation.

French and Steward (1976) provide anelaborate typology for all possible patterns ofpresence or absence of the three independentcomponents and the type of mini-episode andresidual state that will ensue in response to astressor. We have presented their scheme inmodified form in Table 1. Thus, if all threecomponents are present in adequate amount,only a brief mini-episode occurs in responseto the stressor, successful adaptation ensues,and the organism experiences a state of joy

(Pattern 1). If competence is insufficient eventhough the other components are present insufficient amount, a state of frustration ensues(Pattern 2). If coping efforts are insufficientbut the other two components are adequate,an inhibitory state sets in and no solution issought (Pattern 5).

The other patterns give rise similarly to avariety of other responses, and whether acatastrophic situation ensues or a mini- ormaxi-episode develops depends on the indi-vidual's assets and degree of vulnerability.But the important point is that the initialresponse to the stressor—the crisis—is abreakdown in coping that is either transient,or if it persists, one that provides the matrixfor the development of an episode.

Relation Among Vulnerability, Life EventStressors, and Adaptation

What is the relation between vulnerabilityand the various components of adaptation,and what is the association between failuresin adaptation and the onset of episodes ofpsychiatric disorder?

First, we have already suggested thatstresses of a catastrophic nature for an indi-vidual generally produce some measurabledisruption in adaptation. The disruption neednot be gross; it might perhaps be describedas a mini-episode of failure in adaptation. Weshall take up the problem of how to index amini-episode at a later point. Here it must beremembered that the stress value of various


life events depends on the perception of threatby the individual. And although Holmes andRahe (1967) have demonstrated group-basedsimilarities in the stress value of various lifeevents, there is still considerable room forindividual differences.

When adaptation goes awry and a mini-episode appears, the individual's degree ofvulnerability determines the severity andlongevity of the consequences. Phrased posi-tively, invulnerability describes the organism'sultimate resilience to the load imposed on itby life event stressors. Failures in adaptationmay take the form of temporary disability—the pain and distress that will reduce effi-ciency and happiness—but stop short of be-coming enduring disruptions of the organ-ism's functioning. One form of transient fail-ure in adaptation is the "bad day" or "roughweek" familiar to everyone. Such normal fluc-tuations in adaptation are ephemeral and donot usually require therapeutic intervention.In a vulnerable individual, however, what be-gins as a mini-episode may develop into anepisode of psychiatric disorder that is far moreoverwhelming to the person, is longer lasting,and involves a more total disabling of be-havior.

In both mini-episodes and episodes of psy-chiatric disorder, the outcome may be eitherassimilation or accommodation. In assimila-tion, adaptation is regained primarily by alter-ing the environment. When the stress is elimi-nated, either by the organism's owninstrumental behavior, by the passage of timeand circumstance, or by the assistance of someexternal agency, the organism returns towardits preepisode status. Perhaps the primaryconcept behind the development of asylumsfor the psychiatrically ill was to remove pa-tients from the stresses of life and allow theassimilative process to take its course. In epi-sodes of psychiatric disorder, as well as inmini-episodes of failures of adaptation, wehave suggested that a full return to preepisodestatus can occur. However, the process maytake longer in episodes of psychiatric disorderand may be complicated or even prevented byother adverse effects of institutionalizationand stigmatization. The other outcome of amini-episode or major episode is accommoda-tion, in which the organism alters its internal

state to meet the needs of the environment.Optimally, such alterations are adaptive andrepresent growth that may prove to be pre-ventive of future episodes. Alternatively, theymay be regressive and lead finally to an en-trenchment of psychopathology and a loss ofbasic capabilities.

Return for a moment to the French andSteward (1976) model: In which componentof adaptation may we expect to see signs ofan encroaching episode of failure of adapta-tion? Is it a drop in coping effort, an inabilityto perceive the discrepancy between thestressed and optimum states, a loss in com-petence, or a patterned combination of allthree components? At present, it would be fartoo complicated to deal with all three elementsand their interactions. Hence we will place ahunch on one component—the coping effort.

This tentative decision is based on an as-sumption that competence and the cognitiveability to perceive discrepancy are stable,structural characteristics of the organism, de-veloped over long experience with previoussituations requiring adaptation. The exertionof coping efforts, on the other hand, seems tobe a more labile, dynamic force that can waxand wane with time and occasion. It is func-tional rather than structural, serving to suc-cessfully evoke the capacities of the organismto contend with problems or to inhibit andfail to elicit adaptive strategies. Again, onemight liken the cognitive and competencecomponents to the structure of a machine,whereas the coping-effort component is thepower needed to make the machine function.

Although we propose that a diminution,exacerbation, or inappropriate application ofcoping efforts may perhaps be the ideal mark-ers for an episode, we should point out thedifficulty of separating the three componentsof adaptation in actual measurement. Pre-cisely because the structural cognitive andcompetence components can only be energizedby coping efforts, it would be difficult to elicitand demonstrate their intactness during anepisode. It is apparent that individuals of themost excellent premorbid status display extra-ordinary degrees of incompetence and grosslapses of cognitive insight during the heightof an episode of psychiatric disorder.


Premorbid Competence and Vulnerability

If one were to question laymen, clinicians,and researchers about vulnerability to psy-chiatric disorder, one would probably unearththe implicit assumption that there is a greater-than-chance association between incompetenceand high vulnerability or between high com-petence and invulnerability to psychopathol-ogy. Such an implicit assumption has becomemore explicit with the advent of programplanning to prevent psychiatric disturbanceamong children at some risk for psychopathol-ogy. Quite frequently such programs are de-signed to incorporate competence-enhancinginterventions on the grounds that these inter-ventions should simultaneously prove to re-duce vulnerability.

Let us therefore examine the relation be-tween vulnerability and preepisode compe-tence. It is of course clear that during an epi-sode, a patient's competencies appear to be ata low ebb. However, our concern is not withthe patient's capacity to display his skills andabilities during his illness, but rather with hispre-episode demonstrations of competence andtheir relation to vulnerability. In other words,is a person less vulnerable to psychiatric dis-order because he is generally competent atmaking use of many response options andmeeting life exigencies successfully? Or iscompetence independent of vulnerability?

There are several reasons why we shalltentatively assume that competence is or-thogonal to vulnerability. First, there are butfew data available, and until more adequatedata are brought forth it is best to adopt thenull hypothesis. Second, the evidence for in-terdependence stems largely from studies ex-amining the prevalence of psychiatric disorderas a function of socioeconomic status. Thegeneral finding in this area has been the nega-tive correlation between socioeconomic statusand rates of schizophrenia. However, thereare numerous difficulties in generalizing fromthis result to the conclusion that competenceand vulnerability are interrelated. First, insmaller towns, the negative correlation doesnot generally hold. Second, the generalizationentails the ecological fallacy, which infersfrom correlations between group means aninterpretation that should only be inferred

from the total correlations, including within-group correlations. Third, it is unclear whetherthe negative correlation is due to (a) a causalrelation between socioeconomic status andschizophrenia or to (b) the downward drift ofschizophrenics on the socioeconomic-statusscale. Fourth, and very fundamental, iswhether socioeconomic status can be inter-preted as a valid index of competence. Thereare many who might question whether finan-cial income, occupational level, or level ofeducation accurately reflect an individual'sgeneral level of success in dealing with lifeexigencies. Even when we put aside this al-most imponderable and value-laden question,however, can we gain any better understand-ing of the relation between competence andvulnerability by examining one individualcomponent of socioeconomic status?

There is no certainty that occupational level—one of the chief elements of socioeconomicstatus—is negatively related to schizophrenia.Findings by Turner (in press) indicate thatnearly two thirds of the premorbid schizo-phrenics in the sample drawn from the Mon-roe County Case Register (New York) wereemployed before entering the hospital, a figureonly 20 percentage points lower than that forthe general population. Furthermore, the pro-portion of these employed patients who wereworking at skilled manual levels or higherpremorbidly corresponds to the similar pro-portion in the general population—about70%. It is also interesting to note Serban's(1975) finding that the proportion of thoseschizophrenic admissions to Bellevue Hospitalin New York City who were never employedwas no greater than the proportion in theBellevue catchment area (3.2% to 3.5%).Christensen (1974) found that in a follow-upof a cohort of admissions, the rehospitalizedpatients had a better employment record buta more severe level of psychopathological dis-turbance at follow-up than did their peers whowere not rehospitalized. Thus, occupationalcompetence seems to provide no protectionagainst initial admission to a psychiatric hos-pital or against rehospitalization.

One of the most incisive studies of the com-patibility between mental disorder and theworking world is based on an investigation ofunion members of the Amalgamated Clothing


Workers of America (Weiner, Akabas, &Somner, 1973). A special clinic was estab-lished for the purpose of preventing mentaldisorder from interfering with worker pro-ductivity. Taken as a whole, the patients wereable to work at a level almost as productiveas that of their demographic and occupationalpeers. The neurotics even excelled their peers,whereas the psychotics fell somewhat belowin earning power. Only 20% of the patientshad not returned to their jobs by the end ofthe 6-month therapeutic endeavor. Thus, 92%of the patients were at work immediately priorto their entry into the clinic; from 73% to87% worked during the period of treatment;and fully 80% were back at work at the endof treatment.

Two tentative conclusions might be drawnfrom all of these findings. First, occupationalcompetence generally confers no immunity toschizophrenia. Second, the highly vulnerableindividual, even when ill, may be able todemonstrate satisfactory levels of work com-petence under suitable conditions.

All things considered, perhaps we should be-gin to consider the possibility that the pre-sumed interdependence between competenceand invulnerability to schizophrenia is a myth,and indeed a comforting one to many of us.One must bear in mind a number of circum-stances that might give such a myth the ap-pearance of truth. First, most of the data arebased on cases of patients hospitalized in pub-lic institutions. Private hospitals or clinicswhere more affluent members of the popula-tion send family members have often beenexcluded from our statistics. Second, it islikely that highly competent schizophrenics,often with the help of supportive families, areable to find ways of maintaining themselves inthe community. Autobiographical accounts(Barnes & Berke, 1971; MacDonald, 1960;Wallace, 1965) amply demonstrate the in-genuity with which some psychotic individualscan avoid unpleasant periods of hospital con-finement. This more talented group of patientsmay therefore be particularly unlikely to berepresented in our statistics for hospitalizedcases. If we were to include schizophrenics atlarge who do not, or have not yet, reached thehospital, a much less distorted picture of therelation between competence and vulnerability

would probably emerge in the direction of re-ducing or even eliminating patient-nonpatientdifferences.

Another distorting element in our data isthe tendency for low-socioeconomic-status pa-tients to remain in the hospital longer, per-haps even after their episodes have ended.Since most of our data are based on preva-lence rather than incidence figures, anothersource of bias enters. There is also some evi-dence suggesting that low socioeconomic statusor low competence per se may occasionallyenhance the likelihood of admission to psy-chiatric hospitals. Thus, as Meyers and Bean(1968) have noted, lower-class patients hadmore frequent admissions to hospitals eventhough they exhibited less psychopathologythan upper-class patients. The so-called mar-ginal persons—individuals who are unmarried,living alone, unemployed, or of the lowestsocioeconomic status—show higher rates ofseeking and receiving mental health care thanwould be expected from the overall prevalenceof symptoms of psychopathology in thesedemographic groups in the community (Tisch-ler, Henisz, Meyers, & Boswell, 197Sa,197Sb).

In concluding this section we might merelynote that an individual's tendency to be hos-pitalized in a psychiatric institution may beinfluenced by many factors other than thepresence of psychopathology. These factorsmay tend to inflate the probability that lessskilled individuals suffering from schizophre-nia will appear in the hospitalized population,whereas more competent individuals withschizophrenia may not. It is interesting to notethat during the great depression of the early1930s, the marginal persons described abovetended to be herded into the mental hospitals(Zubin & Burdock, 1965). The decline in theresident patient population might be attrib-uted in part to the development of welfareand social programs that make possible thesubsistence of this cohort in the community.

Premorbid Coping Ability, Life Events,and Vulnerability

Coping ability is the resultant of the copingeffort, or initiative, and the competence, orskill, that an organism brings to bear in for-


mulating strategies to master life situations.When an individual is described as a goodcoper, it is generally meant that he maintainsadaptation on an even keel (a) most of thetime and (b) in the face of life events that arewidely recognized to be quite stressful. Thepoor coper, in contrast, seems to be thrownoff balance (a) rather frequently and (b)more often by events that are not consensuallyregarded as threatening or challenging. Majoror minor periods of coping breakdown or fail-ure in adaptation will, of course, be observedin the good coper as well as in his less adeptcounterpart. These will be marked by suchtypical signs of strain as subjective discomfortor distress, feelings of helplessness or tension,withdrawal, inappropriate or ineffectual be-havior, and physiological disequilibrium orvegetative disturbances, but not necessarily byepisodes of mental disorder.

In general, the dimensions of coping abilityand vulnerability to psychiatric disorder areprobably independent of each other. Thechronic ne'er-do-well, who seems to be upsetby the slightest stress, is not necessarily theperson most likely to develop an episode ofschizophrenia. On the other hand, there maybe one sense in which coping ability andvulnerability are interrelated.

It seems likely that the person in a state ofcoping breakdown enters a period of risk, astate of lowered psychological resistance. Ifthe individual's vulnerability is sufficientlylow, this period will pass with relatively smallconsequence. On the other hand, if the vul-nerability is high, such sensitive points arelikely opportunities for vulnerability to de-velop into the expression of psychopathology.The poor coper will encounter more dailyevents that lower his resistance and may con-sequently spend a greater proportion of hislifetime at risk for the manifestation of what-ever vulnerability he possesses.

We might liken the occurrence of a copingbreakdown to the development of a fault inthe earth's surface. Ordinarily, a brief tremorwill occur before a stable formation of theearth is reestablished, just as in the person'slife a slight jar in the spheres of work andpersonal affairs will precede a new state ofclosure. On the other hand, if a bed of moltenlava is seething below the crack in the earth's

surface, the development of the fault willcreate an opportunity for the emergence of avolcano. In the vulnerable individual, a copingbreakdown may open the way for the eruptionof a psychopathological disorder.

Episodic Nature of Schizophrenia

We have postulated that the primary per-sistent characteristic of the schizophrenic ishis vulnerability, not his disorder. This, how-ever, flaunts the traditional view that regardsschizophrenia as a permanent condition lead-ing to chronic deterioration or unremittingimpairment. A recent review (Spring & Zubin,in press-b) finds only sparse evidence for thechronic, unremitting nature of most schizo-phrenic illness. Is the episodic hypothesis atenable alternative?

Hospital statistics indicate that the averageduration of hospitalization has dropped fromseveral years in the custodial period before1956 to 37 days in 1975. This trend is aworldwide phenomenon. Most chronic patientscurrently in our hospitals were admitted aslong as 20 years ago. We must remember thatthese cases are the product of the custodial-care era, and their chronicity may reflect theiatrogenic influences of long-term incarcera-tion as much if not more than an unremittingdisease course. Parallel with the reduction inchronic cases has been a considerable increasein the rate of readmission to hospitals, andalthough some of these readmissions may rep-resent the unimproved patients who were re-leased in error, or rehospitalization for lackof other living facilities, many of them mayrepresent new episodes.

M. Bleuler's (1974) lifetime follow-up of208 schizophrenic probands provides the mosttelling comment on the gradual replacementof the prototypical chronic, unremitting schiz-ophrenia by a more episodic pattern of dis-order. Bleuler found that only 10% of hispatients showed the type of disease course de-scribed as typical in Kraepelin's writings andE. Bleuler's early work. Half of the patientsachieved an adequate adjustment in the com-munity, whereas 40% lived most of their livesin the community but were hospitalized oc-casionally.


Even the 10% of unremitted cases in M.Bleuler's (1974) sample were products of thecustodial-care era, and it remains to be seenwhether even this percentage will be reducedin today's more benign treatment milieu.Moreover, it will also be important to investi-gate whether the apparently continuous courseof chronic illness may actually consist of aclosely spaced series of episodes. It is possiblethat the chronic group of patients may includeindividuals of such high vulnerability thatthey almost continuously pass in and out ofepisodes with relatively slight provokingstress.

At present, the nuclear type of schizophre-nic patient seems to have become increasinglydifficult to find. For example, Hawk, Carpen-ter, and Strauss (1975) applied Schneider's(19S9) first-rank symptoms; Langfeldt's(1939, 1956) criteria; and Carpenter, Bartko,Langsner, and Strauss's (1976) discriminatingsigns to patients at intake, but they could notfind any nuclear types. Moreover, Hawk et al.were entirely unsuccessful in predicting out-come. To their surprise, 40% of their schizo-phrenic patients had the best outcome whencompared with other psychiatric patients ona composite index of symptom severity, dura-tion of hospitalization, and social adjustment.

With the decline in the prevalence ofchronic hospitalization, a new pattern is be-ginning to emerge as characteristic of theschizophrenic disorder. This course seems toinvolve brief episodes of illness, recovery, re-lapse, and recovery. Recently, groups of pa-tients randomly assigned to brief therapy last-ing only 4 weeks have shown results as goodor better than groups consigned to long-termtherapy (Herz, Endicott, & Spitzer, 1976).That patients are responsive to therapy afterso short a time lends support to the episodicnature of schizophrenia. Just as relativelyrapid improvement is one side of the episodicpicture, relapse is the other. The problem ofrelapse into new episodes has emerged as oneof the most prominent issues in the mainte-nance treatment of patients today. Davis(1975) reviewed 24 double-blind placebo anddrug maintenance studies of schizophrenicswho had been released after suffering an epi-sode of illness. He noted that 60% of thepatients never relapsed during the 2-year fol-

low-up. However, another 40% (65% in theplacebo group and 30% in the drug group)did relapse and probably underwent new epi-sodes during the 2-year follow-up. In anotherstudy of maintenance treatment, Goldberg,Schooler, Hogarty, and Roper (in press)found that 35% of schizophrenics never re-lapsed during the follow-up period, whereas65% did relapse (&0% in the placebo groupand 48% in the drug group). At least somepatients in the relapsed group recovered again.Goldberg suggests that the efficacy of mainte-nance medications may reflect their prophy-lactic value in aborting new episodes of illnessrather than their effects on a continuouslypresent disease process.

The factors influencing relapse, as well asnew episodes, are clearly complex. Reportsfrom England (Leff, 1976) indicate that therelapse rate for patients returned to hostilehome environments is far greater than thatfor patients returned to benign home environ-ments. It is also interesting to note that therelapse rate in developing countries is lessthan in developed nations, despite the factthat the proportions of deteriorating patientsare comparable (Jablensky & Sartorius,1975). It may be that a lesser degree oflabeling and stigmatization in the developingcountries affords the returned patient a lesshostile ecological niche in the community. Itmay take a more "advanced" culture to stig-matize an individual into permanent deviance!

In summary, it seems that chronic unre-mitting schizophrenics are becoming increas-ingly rare. The many follow-up studies thatfound a pattern of disorder succeeded by cy-cles of improvement and relapse support thehypothesis that what is permanent aboutschizophrenia is the vulnerability, rather thanthe episodes of psychosis.

How does this hypothesis contribute to un-derstanding the case of the schizophrenic pa-tient who does appear to be chronically ill?There are at least four possible interpreta-tions: (a) There are probably some schizo-phrenics who do remain unremittingly in anepisode of illness, although they are in theminority, (b) As suggested earlier, others areprobably patients so vulnerable that they re-cover only very briefly before being catapultedback into an episode. Their periods of emer-


gence from disorder are probably so brief asto go unnoticed, (c) Still others in this cate-gory may have been of such poor premorbidability that their recovery from illness is diffi-cult to discern. In contrast to the good coperwho recovers and resumes his place in society,these patients recover most unremarkably:They remain unable to cope with life's exi-gencies, just as they were unable to do sopremorbidly. (d) Finally, there is a group ofchronic patients whose coping ability deterio-rates far below its premorbid level as a conse-quence of the onslaught of psychopathology,isolation in the hospital or community, anddisuse of social skills. Gruenberg (1967) hasdescribed a social breakdown syndrome thatmay characterize patients in backwards andisolated individuals. The resultant losses incoping ability may indeed be chronic, andthey may be mistaken for permanent psycho-pathology.

Indexing the Onset, Duration, andOffset oj an Episode

To test the vulnerability model and its hy-pothesis regarding the episodic nature ofschizophrenia, it will first be necessary to de-velop objective indexes of the onset, duration,and end of an episode.

Timewise documentation of the onset,course, and termination of a schizophrenic epi-sode will not be an easy task. Ordinarily therewill be a considerable time lag between thepoint at which a patient first enters an episodeand the time when he is admitted to a hos-pital. Once the patient comes to our attention,he is already well into an episode and mayeven be beginning to emerge from his periodof disorder with the help of stabilization onmedication. Retrospective interviewing to es-tablish the time, if not the behavioral corre-lates, of onset generally reveals considerablediscrepancy between different informants. In-deed, onset has always posed one of the mostbaffling problems in diagnosis. Type of onset,whether insidious or sudden, is on the onehand of great prognostic importance and yeton the other hand notoriously difficult to de-fine objectively. The difficulty of determiningonset is magnified by the subtlety and com-plexity of behaviors that have been proposed

as signposts of the beginning of a schizophre-nic episode. One of the earliest suggestionswas that of Berze (Berze & Gruhle, 1929),who described an early and pervasive generallowering of psychic activity. Another proposal,by Jaspers (1963), pointed to the emergenceof "ununderstandable" behavior as the ear-mark of the onset of the disorder. Clearlythese phenomena do not easily lend themselvesto the formulation and quantification of ob-jective indicators, Moreover, we can mostlikely not expect them to be accurately re-ported by patients whose disorder is oftencharacterized by lack of insight.

Even though the patient usually comes tous too late to determine indexes of onset, per-haps we can expect greater success at observ-ing the ending of his episode. Here again, how-ever, several difficulties should be anticipated.Just as the date of admission to the hospitalbears no necessary isomorphism with the timeof episode onset, neither does time of dis-charge necessarily correspond to the recoveryand emergence from the episode. Often pa-tients who seem to have made adequate re-coveries remain in the hospital awaiting theavailability of appropriate placements in thecommunity. Likewise, patients who still ap-pear quite ill may be discharged to the careof their families and followed on an outpa-tient basis. Another problem with evaluatingepisode offset in hospitalized patients is thatwe have no opportunity to observe them inthe significant contexts in which their psycho-pathology might be displayed. Given all ofthese stumbling blocks, what recourse havewe?

In searching for indexes of episodes ofschizophrenia, we might do well to look ini-tially at such conditions as epilepsy and de-pression, in which there is some evidence ofstate- and trait-related markers. Thus, thebeginning and end of episodes of epilepsy canbe indexed by electroencephalographic char-acteristics before and after seizure. Similarly,in depression, there have been findings indi-cating that changes in sleep characteristics(Kupfer, 1976) and motility (Kupfer et al.,1974) serve to mark the beginning and end ofepisodes. In schizophrenia, too, a reduction in


rapid-eye-movement sleep accompanies acuteepisodes (Kupfer & Foster, 1975).

Another strategy may be to monitor mini-episodes or acute periods of symptom exacer-bation in patients during their hospital stay.These may, in microcosm, reveal some proper-ties of onset and offset of maxi-episodes ofschizophrenic disorder. Some progress hasbeen made in this area. Although most of ithas concerned a search for indicators of theonset and offset of mini-episodes in depressedpatients, it is likely that comparable indexescan be found for schizophrenic mini-episodesinduced by visitors, peer stresses, and so on.In a study of the rise and fall of corticosteroidlevels in reactive depression, by Sachar, Mac-Kenzie, Binsback, and Mack (1968), it wasobserved that hormonal stress levels rise whenthe patient actively struggles with the con-frontation of the loss of a love object or whenhis defense mechanisms falter temporarily. Onthe other hand, when his defenses are operat-ing adequately and he seems clinically com-fortable and adjusted, the hormonal leveldrops to normal. After the patient has ad-justed to the hospital regime and achievedstable hormonal levels, such perturbationsmay therefore index the eruption of mini-episodes.

Another example is due to Schmale (1972),who suggests that giving in to feelings ofhelplessness and hopelessness is the earmarkof a beginning episode. Luborsky and Auer-bach (1969) have derived another potentialmini-episode index by examining speech sam-ples of patients in psychoanalysis taken justbefore instances of momentary forgetting orjust prior to reports of migraine headaches orstomach pain. When these speech passages arerated for content of expression, helplessnessratings are found to be much higher than forsamples taken at other times during therapysessions. It is not entirely outside the realmof possibility that mild transient mini-episodesmight also be provoked by biochemical chal-lenges. The ethics of such experimental pro-cedures must be carefully investigated so asnot to interfere with the civil rights andfreedom of the patient, but under skilled per-sonnel, mindful of the ethical issues involved,perhaps some salutory solution can be found.

As we have suggested earlier, the beginningof an episode of psychopathology often fol-lows close on the heels of a period of copingfailure. Similarly, once the episode of disorderhas subsided, one may expect the patient toonce again be able to resume coping at hischaracteristic premorbid level and style. Theconcurrence of episodes of coping dysfunctionand restoration with episodes of psychiatricdisorder and recovery opens the way for in-vestigation of another type of episode marker.If we were to have knowledge of the patient'spremorbid level of coping ability, we coulduse evidence of a sharp plummeting of copingeffectiveness and a return to the usual ca-pacity for efficiency as rough boundaries toindex the course of a psychopathological epi-sode. Following Goldfried and D'Zurilla's(1969) behavior-analytic model for assessingcompetence, Goldsmith and McFall (197S)have experimented with simulated interper-sonal contexts to evaluate the effectiveness ofpatients' social coping strategies. The entirespectrum of role performance should actuallybe sampled to see whether the capacity tocope has returned to its premorbid level.Weissman (1975) has recently reviewed 15scales available for assessing performance inoccupational, marital, extended family, andcommunity roles. These might serve to probefor fluctuations in coping capacity that occurduring hospitalization, particularly if it werepossible to simulate the significant role con-texts.

The first step in developing indexes ofvulnerability and of the onset, duration, andoffset of episodes is to turn to available tech-niques that objectively and reliably differenti-ate schizophrenic patients from normal sub-jects. Experimental approaches to developingsuch indexes have been described elsewhere(Spring & Zubin, in press-a). The next stepis to begin to sort out these differences intothose that characterize the patient only aslong as his episode persists, those that turnout to be permanent effects of an episode ofdisorder, and those that characterize thevulnerable individual regardless of whetherhe is in an episode. Those indexes that char-acterize the patient only during the episodemay eventually turn out to be good markersof the onset and offset of the episode. Those


that characterize the person before, during,and after an episode may turn out to bevulnerability markers. Finally, those char-acteristics that result from the illness mayhelp in studying the vulnerability to relapse.

Studies of identical twins and relatives canhelp in discovering indexes that mark thevulnerable individual regardless of the pres-ence of an episode. If both the proband andhis nonaffected monozygotic twin or his non-affected siblings and relatives have the samecharacteristic in question, that characteristicis a good candidate for becoming a vulnera-bility index. If the proband has the character-istic when he is in an episode but not when hehas recovered, and if his monozygotic dis-cordant twin or other blood relatives do nothave the characteristic, it is a good candidatefor marking the beginning and end of epi-sodes.

Utility of the Vulnerability Hypothesis

In recent years there has been great concernwith the civil rights of patients suffering frommental disorders. Many problems have beenraised concerning whether civil rights are infact abrogated when therapeutic interventionis exerted despite the patient's unwillingnessor incompetence to give his informed consentto treatment. Moreover, there is growing sus-picion that the consequences of being labeledand stigmatized as mentally ill may be far-reaching, dehumanizing, and injurious to civilrights. In the final analysis, attacks have oftenfocused on the so-called medical model be-cause it is claimed that this model adds insultto injury by considering mental disordersdiseases in the same way that cancer andturberculosis are diseases.

There can be little doubt that the "disease"conception of mental disorder, regardless ofits validity, influences the expectancies thatthe community, the hospital staff, and indeedthe patients themselves hold toward an indi-vidual who has been hospitalized for psychi-atric disorder. Perhaps the proposal to regardthe schizophrenic as a vulnerable individualwho will develop a temporary episode onlyunder certain provocations can help to placethis controversy in proper perspective. Byshifting our view from regarding a person as

suffering from a continuing mental disorder toregarding him as suffering from a temporaryepisode, and further by regarding him as being•vulnerable rather than diseased, some of thecontroversy might become superfluous.

In a recent study by Doherty (1975), acomparison was made among patients who atfirst agreed (but only reluctantly) to assumethe patient role and to consider themselvesmentally ill but who later rejected the label(rejectors) and two other groups—label ac-ceptors and label deniers. It is interesting tonote that the staff expected the patients toaccept the patient role if they were to benefitfrom treatment. Despite this staff expectancy,the label rejectors fared much better than theother two groups in outcome. If staff could betaught to regard patients as vulnerable ratherthan sick and could proceed to protect themagainst the stressors that elicit episodes, therewould be no need to label them as mentalpatients or induce them to accept the label.The vulnerability label is perhaps easier toaccept and live with, since it presages a time-limited episode from which the patient willrecover sooner or later without being labeledschizophrenic for the rest of his life.

There are several other baffling problemsthat the vulnerability hypothesis may helpexplain. The apparently spontaneous recoveryof long-standing patients falls into this cate-gory. The patient may actually have emergedfrom his episode much earlier, but when recog-nition of this fact suddenly occurs the recov-ery is regarded as miraculous. Placebo reac-tors may also be explicable on the basis of thevulnerability model. If all episodes are timelimited, it is to be expected that a certainproportion of patients will in time show im-provement without treatment. Placebo reac-tors may represent the unearned increment ofspontaneous improvement.

The vulnerability hypothesis might also beuseful in explaining the well-known statisticthat only 10% of schizophrenics have simi-larly affected parents. Moreover, it might elu-cidate Kety et al.'s (1968) finding of low ratesof schizophrenia in the biological relatives ofacute schizophrenic probands, particularly incomparison with the higher rates of schizo-phrenia in the biological relatives of chronicschizophrenic probands. It is possible that the


familial genes of the chronic patients set thevulnerability threshold so low that ordinarylife events, despite their minimal stress impactfor the average person, are sufficient to triggeran episode.

On the other hand, the low rate of con-cordance in the blood relatives of acute schizo-phrenics may be due to the higher thresholdof vulnerability in their families. Thus, unlessthe lives of blood relatives and schizophrenicprobands have involved similar stressful inci-dents, one would not expect high concordancefor schizophrenia in these groups. An analogymay perhaps be drawn between the chronicand acute forms of schizophrenia and twoconditions of sickle cell anemia: sickle celldisease (the homozygous condition) and sicklecell trait (the heterozygous condition).Whereas homozygous individuals will fall illregardless of circumstances, heterozygotes willdevelop the disease only if they are exposedto such specific triggering events as high alti-tudes, acute alcoholism, or deep anesthesia.Genetically comparable sickle cell trait car-riers who avoid these situations will remainfree of the disorder, and their vulnerability tosickle cell anemia may never be detected(Harris, 1963). However, the usual exigenciesof living in our particular ecological surround-ings are sufficient to trigger an episode in thehomozygotes. Perhaps acute schizophrenia,like sickle cell anemia, will develop only inthose vulnerable individuals who undergo spe-cific adverse circumstances. Unfortunately, thespecificity of these circumstances is still to bedetermined.

If we accept the vulnerability hypothesis, itwould seem that therapeutic interventions cantake one of two avenues. Vulnerability can bereduced or inhibited from full-blown expres-sion through psychopharmacological interven-tion. On the other hand, if it is correct, as wesuggested earlier, that periods of coping break-down provide fertile ground for vulnerabilityto germinate into disorder, psychological in-tervention might be applied to restore copingability or reduce the threatening nature of lifeevents that produced the breakdown. Eventoday, the patients who continue to come toour attention are those who fail to improveand who frequently relapse, rather than themajority who make a successful recovery. It

is this group of patients—the failures—whoseem to merit the most concerted efforts atpsychological intervention. Instead of shunt-ing them, as we now do, through the revolvingdoors of our hospitals and clinics into the ob-scure wards of the hospital or the back alleysof the community, we should devote our bestefforts and psychological skills to rehabilitat-ing them by enhancing their competence andcoping ability. We argue that the challenge ofthe 1970s is not to find a cure for schizophre-nia, since episodes of schizophrenia are in themajority of cases self-curing! Rather, thechallenge is to find ways of reducing vulner-ability or improving the coping abilities andcompetence of the vulnerable poor premorbidsso that the likelihood of future episodes canbe reduced. Even if an episode does occur, therehabilitated patient will have a better levelof coping to return to when the episode passes.

Summary of the Vulnerability Model

We can now summarize our model of vul-nerability to schizophrenic episodes, as shownin Table 2. The picture of the schizophrenicindividual that emerges from our vulnerabilitymodel is highly speculative and controversial,but it has the virtue of being consistent withthe data. In other words, it is a tenable if notyet cogent model. According to our findings,the schizophrenic is drawn from the entirerange of human variation with regard to intel-ligence, competence, coping ability, achieve-ment, and all other aspects of the human con-dition. He may be a top executive, a giftedartist, an outstanding scientist, a blue-collarworker, or an unskilled ne'er-do-well bowerybum. No stratum of mankind is spared fromthis blight. The one feature that all schizo-phrenics have in common is not the everpresence of their illness, but rather the everpresence of their vulnerability. Some of themare highly vulnerable and have repeated epi-sodes. Others are relatively invulnerable andhave but one brief episode or none at all.When episodes develop they are not lifelong.They terminate sooner or later with or with-out therapeutic intervention. The majority ofschizophrenics today spend the major part oftheir lives in the community, self-supportingand indistinguishable from the rest of the


Table 2Vidner ability Model

Vulnerability

Coping effort

Competence

Coping ability

Component

The empirical probability that an individual will experience an episodeof psychiatric disorder; an enduring traitThe energy exerted in situations not adequately dealt with by reflexes;orthogonal to vulnerabilityThe skills and abilities needed to achieve success in significant role con-texts of everyday life; orthogonal to vulnerabilityThe resultant of the initiative and skill that an organism brings to bearin formulating strategies to master life situations; orthogonal to vul-nerability

Episodes of coping breakdown

Episodes of psychiatric disorder

Effect of stressors

Occur in all individuals when catastrophic situations arise and renderroutine coping strategies ineffectual; do not necessarily lead to episodesof psychiatric disorderTend to develop in vulnerable individuals who are in a state of copingbreakdown; time-limited states of illness

population except in the eyes of those whohave labeled them. To the best of our knowl-edge, half the patients recover fully, another40% have relapses but still manage to spendmost of their lives in the community, and only10% appear to remain chronically ill. Whethertheir chronicity is endogenous or induced byiatrogenic or custodial factors remains to bedemonstrated. At the onset of an episode, thepatient's competence and ability to cope ade-quately with life's exigencies appear to go un-derground. At the end of the episode, theseattributes tend to reappear at their premorbidlevels, and the patient can resume his formerplace in society. In general, the good pre-morbid patient returns to his formerly goodadjustment and the poor premorbid to hispoor adjustment. The generally assumed nega-tive relation between vulnerability and com-petence seems to be largely an artifact of ourlimited information regarding the full spec-trum of schizophrenic illness. Our informationis biased by the fact that many episodes inhighly competent individuals are not recordedin our statistics and by the fact that thosewho fill our hospitals and clinics are largelythe poor premorbid, relapsing patients.

The picture of schizophrenia presented inthis article is not that which dominates ourtextbooks, and it may consequently be viewedwith suspicion. But let us remember M.Bleuler's comments (cited in Rosenthal, 1974)

on the observational biases that led hisfather's generation to regard schizophrenia asa chronic, unremitting, and even deterioratingcondition:

That which my father had to a certain extent donein cross section, I have investigated longitudinally.He could stay with his patients only as long as theyremained in his clinic. When they left the clinic,they were thereafter out of sight and lost to him,and this was the case with most psychiatrists of hisgeneration. For this reason, an unfavorable pictureof the course of illness had to be inferred: The im-proved and the healed patients disappeared beyondthe horizon of the clinic, and he saw above all thosewho were unimproved or relapsed, (p. 92)

It is also interesting that several otherworkers have entertained the possibility thatschizophrenia occurs in self-limiting episodes.After concluding that pre-drug-era follow-upshowed as good results as drug-era follow-up,Bockoven and Solomon (1975) discuss thecommon philosophy that permeated the clini-cal and social management of patients in thetwo eras.

This philosophy is based on the idea that the ma-jority of mental illnesses, especially the most severe,are largely self-limiting in nature if the patient isnot subjected to demeaning experiences or loss ofrights and liberties. Therapeutic management con-sists first and foremost of removing these negativeinfluences and replacing them with a positive atti-tude of respect for the patient's needs for humancompanionship and interest-holding activity. Somatictreatments are prescribed in this context to relieve


specific kinds of suffering and thereby to expeditethe spontaneous healing process, (p. 796)

Apparently, Slater and Slater (1944) haveproposed a concept of vulnerability that isclosely similar to the concept promulgated inthis paper. Even more cogently, Falconer(1965) proposed a similar mathematicalmodel under the term "liability to diseases"that has been applied to schizophrenia byKidd (1975).

Why then is this article necessary, if itsmain thrust has been anticipated by severalearlier publications (Gottesman & Shields,1972; Meehl, 1962; Millon, 1969; Rosenthal,1970)? The answer lies in the fact that al-though the data on which vulnerability wasbased are known, their organization and ap-plication have lagged. Further, although vul-nerability has been proposed several times, ithas never caught on in schizophrenia. We havearrived at this formulation independently ofour predecessors, driven to it not by theweight of evidence that has accumulated butby the way in which it illuminates the meagerevidence. In addition, we have suggested bio-metric approaches for testing the tenability ofthe vulnerability model by indicating the needfor measures of life event stressors, compe-tence, coping, and vulnerability to schizophre-nia as well as markers of the onset and offsetof schizophrenic episodes (Spring & Zubin, inpress).

Reference Note

1. Kety, S. Genetic and environmental factors in theetiology of schizophrenia. Paper presented at theMcLean Hospital Symposium on the BiologicalSubstrates of Mental Illness, Belmont, Massa-chusetts, February 1975.

References

American Psychiatric Association. Diagnostic andstatistical manual of mental disorders (DSM-II).Washington, D.C.: American Psychiatric Associa-tion, 1968.

American Psychiatric Association. Diagnostic and sta-tistical manual of mental disorders (DSM-III).Washington, D.C.: American Psychiatric Associa-tion, in press.

Arieti, S. Interpretation of schizophrenia (2nd ed.).New York: Basic Books, 1974.

Audy, J. R. Measurement and diagnosis of health.

In R. Shepard & D. McKinley (Eds.), Environ-mental. Boston: Houghton Mifflin, 1971.

Ayala, F. J., & Dobzhansky, T. (Eds.). Studies inthe philosophy of biology. Berkeley: University ofCalifornia Press, 1975.

Barnes, M., & Berke, J. Mary Barnes: Two accountsof a journey through madness. New York: Har-court Brace Jovanovich, 1971.

Bateson, G., Jackson, D. D., Haley, J., & Weakland,J. H. Towards a theory of schizophrenia. Behav-ioral Science, 1956, 1, 251-264.

Berze, J., & Gruhle, H. W. Psychologic der schizo-phrenic. Berlin: Springer, 1929.

Bleuler, E. [Dementia praecox: Or the group of schiz-ophrenias] (J. Zinkin, trans.). New York: Interna-tional Universities Press, 1950. (Originally pub-lished, 1911.)

Bleuler, M. The offspring of schizophrenics (S. M.Clemens, trans.). Schizophrenia Bulletin, 1974, 8,93-108.

Bockoven, J. S., & Solomon, H. C. Comparison oftwo five-year follow-up studies: 1947 to 1952 and1967 to 1972. American Journal of Psychiatry,1975, 132, 796-801.

Brown, G. W., & Birley, J. L. T. Crises and lifechanges and the onset of schizophrenia. Journal ofHealth and Social Behavior, 1968, 9, 203-214.

Carpenter, W. T., Bartko, J. J., Langsner, C. A., &Strauss, J. S. Another view of schizophrenic sub-types: A report from the International Pilot Studyof Schizophrenia. Archives of General Psychiatry,1976, 33, 508-516.

Cassel, J. C. Psychiatric epidemiology. In G. Caplan(Ed.), American handbook of psychiatry. (Vol. 2,2nd ed.). New York: Basic Books, 1974.

Christensen, J. K. A 5-year follow-up study of maleschizophrenics: Evaluation of factors influencingsuccess and failure in the community. A eta Psychi-atrica Scandinavica, 1974, SO, 60-72.

Cooper, J. E., Kendell, R. E., Gurland, B. J., Sharpe,L., Copeland, J. R. M., & Simon, R. J. Psychiatricdiagnosis in New York and London: A compara-tive study of mental hospital admissions (MaudsleyMonograph #20). London: Oxford UniversityPress, 1972.

Davis, J. M. Overview: Maintenance therapy in psy-chiatry of schizophrenia. American Journal of Psy-chiatry, 1975, 132, 1237-1245.

Doherty, E. G. Labelling effect in psychiatric hos-pitalization. Archives of General Psychiatry, 1975,32, 562-568.

Dohrenwend, B. P., & Dohrenwend, B. S. Socialstatus and psychological disorder. New York:Wiley, 1969.

Dohrenwend, B. S., & Dohrenwend, B. P. Social classand the relation of remote to recent stressors. InM. Roff, L. Robins, & M. Pollack (Eds.), Lifehistory research in psychopathology (Vol. 2).Minneapolis: University of Minnesota Press, 1972.

Falconer, D. S. The inheritance of liability to certaindiseases, estimated from the incidence among rela-tives. Annals of Human Genetics, 1965, 29, 51-76.


Feighner, J. P., Robins, E., Guze, S. B., Woodruff,R. A., Winokur, G., & Munoz, R. Diagnostic cri-teria for use in psychiatric research. Archives ofGeneral Psychiatry, 1972, 26, 57-63.

Fleiss, J. L., & Zubin, J. On the methods and theoryof clustering. Multivariate Behavioral Research,1969, 4, 23S-2SO.

French, A. P., & Steward, M. S. Adaptation andaffect: Towards a synthesis of Piagetian and psy-choanalytic psychologies. Perspectives in Biologyand Medicine, 1976,18, 464-474.

Goldberg, S. C., Schooler, N. R., Hogarty, G. E., &Roper, M. Prediction of relapse in schizophrenicpatients treated by drug and sociotherapy. Archivesof General Psychiatry, in press.

Goldfried, M. R., & D'Zurilla, T. J. A behavior-analytic model for assessing competence. In C. D.Spielberger (Ed.), Current topics in clinical andcommunity psychology (Vol. 1). New York: Aca-demic Press, 1969.

Goldsmith, J. B., & McFall, R. M. Development andevaluation of an interpersonal skill-training pro-gram for psychiatric inpatients. Journal of Ab-normal Psychology, 1975, 84, 51-58.

Gottesman, I. I., & Shields, J. Schizophrenia andgenetics. New York: Academic Press, 1972.

Gross, E. Work organization and stress. In S. Levine& N. Scotch (Eds.), Social stress. Chicago: Aldine,1970.

Gruenberg, E. The social breakdown syndrome—someorigins. American Journal of Psychiatry, 1967, 123,1481-1489.

Harris, J. W. The red-cell: Production, metabolism,destruction: Normal and abnormal. Cambridge,Mass.: Harvard University Press, 1963.

Hawk, A. B., Carpenter, W. T., & Strauss, J. S.Diagnostic criteria and five-year outcome: A reportfrom the International Pilot Study of Schizophre-nia. Archives of General Psychiatry, 1975, 32, 343-347.

Holmes, T. H., & Rahe, R. H. The social readjust-ment rating scale. Journal of Psychosomatic Re-search, 1967, 11, 213-218.

Herz, M. 1, Endicott, J., & Spitzer, R. L. Briefhospitalization: A two-year follow-up. AmericanJournal of Psychiatry, in press.

Jablensky, A., & Sartorius, N. Editorial: Culture andschizophrenia. Psychological Medicine, 1975, 5, 113-124.

Jacob, T. Family interaction in disturbed and nor-mal families: A methodological and substantiveview. Psychological Bulletin, 1975, 82, 33-65.

Jaspers, K. General psychopathology. Chicago: Uni-versity of Chicago Press, 1963.

Kallmann, F. J. The genetics of mental illness. InS. Arieti (Ed.), American handbook of psychiatry(Vol. 1). New York: Basic Books, 1959.

Kanfer, F. H., & Saslow, G. Behavioral diagnosis. InC. M. Franks (Ed.), Behavior therapy: Appraisaland status. New York: McGraw-Hill, 1969.

Kety, S. S., Rosenthal, D., Wender, P. H., &Schulsinger, F. The types and prevalence of mentalillness in the biological and adoptive families of

adopted schizophrenics. In D. Rosenthal & S. S.Kety (Eds.), The transmission of schizophrenia.Elmsford, N.Y.: Pergamon Press, 1968.

Kidd, K. K. On the possible magnitude of selectiveforces maintaining schizophrenia in the population.In R. R. Fieve, D. Rosenthal, & H. Brill (Eds.),Genetic research in psychiatry. Baltimore, Md.:Johns Hopkins University Press, 1975.

Kupfer, D. J. REM latency: A psychobiologic markerfor primary depressive disease. Biological Psychia-try, 1976, 11, 159-174.

Kupfer, D. J., & Foster, F. G. The sleep of psychoticpatients: Does it all look alike? In D. X. Freed-man (Ed,), Biology of the major psychoses: Acomparative analysis. (Association for Research inNervous and Mental Diseases, Research Publica-tions, Vol. 54.) New York: Raven Press, 1975.

Kupfer, D. J., Weiss, B. L., Foster, F. G., Detre,T. R., Delgado, J., & McPartland, R. Psychomotoractivity in affective states. Archives of GeneralPsychiatry, 1974, 30, 765-768.

Langfeldt, G. The schizophreniform states. Copen-hagen: Munksgaard, 1939.

Langfeldt, G. The prognosis of schizophrenia. Copen-hagen: Munksgaard, 1956.

Lazarus, R. S., Averill, J. R., & Opton, E. M. Thepsychology of coping: Issues of research and assess-ment. In G. V. Coelho, D. A. Hamburg, & J. E.Adams (Eds.), Coping and adaptation. New York:Basic Books, 1974.

Left, J. P. Assessment of drugs in schizophrenia.British Journal of Clinical Pharmacology, 1976,Supplement, 75-78.

Luborsky, L., & Auerbach, A. The symptom-contextmethod: Quantitative studies of symptom forma-tion in psychotherapy. Journal of the AmericanPsychoanalytic Association, 1969, 17, 68-99.

MacDonald, N. Living with schizophrenia. CanadianMedical Association Journal, 1960, 32, 218-221,678-681.

Mechanic, D. Invited commentary on self, social en-vironment and stress. In M. H. Appley & R. Trum-bull (Eds.), Psychological stress. New York: Apple-ton-Century-Crofts, 1967.

Mednick, S. A., & Schulsinger, F. Some premorbidcharacteristics related to breakdown in childrenwith schizophrenic mothers. In D. Rosenthal &S. S. Kety (Eds.), The transmission of schizophre-nia. Elmsford, N.Y.: Pergamon Press, 1968.

Meehl, P. E. Schizotaxia, schizotypy, schizophrenia.American Psychologist, 1962, 17, 827-838.

Meyers, J. K., & Bean, L. L. A decade later: Afollow-up of social class and mental illness. NewYork: Wiley, 1968.

Millon, T. Modern psychopathology. Philadelphia,Pa.: Saunders, 1969.

Murphy, L. B. Coping, vulnerability and resiliencein childhood. In C. V. Coelho, D. A. Hamburg, &J. E. Adams (Eds.), Coping and adaptation. NewYork: Basic Books, 1974.

Rahe, R. H., Meyer, M., Smith, M., Kjaer, G., &Holmes, T. H. Social stress and illness onset. Jour-nal of Psychosomatic Research, 1964, 8, 35-44.


Rosenthal, D. Genetic theory and abnormal behav-ior. New York: McGraw-Hill, 1970.

Rosenthal, D. Introduction to Manfred Bleuler's "Theoffspring of schizophrenics." Schizophrenia Bulletin,Spring, 1974, 8, 91-93.

Rosenthal, D., Wender, P. H., Kety, S. S., Schulsinger,F., Welner, J., & Reiber, R. D. Parent-child rela-tionships and psychopathological disorders in thechild. Archives of General Psychiatry, 197S, 32,466-476.

Sachar, E. J., MacKenzie, J. M., Binsback, W. A., &Mack, J. E. Corticosteroid responses to the psy-chotherapy of reactive depressions: II. Furtherclinical and physiological implications. Psycho-somatic Medicine, 1968, 30, 23-44.

Schmale, A. H. Giving up as a final common path-way to changes in health. Advances in Psycho-somatic Medicine, 1972, 8, 20-40.

Schneider, K. [Clinical psychopathology] (M. W.Hamilton, trans.). New York: Grune & Stratton,1959.

Selye, H. The evolution of the stress concept. Ameri-can Scientist, 1973, 61, 692-699.

Serban, G. Stress in schizophrenics and normals.British Journal of Psychiatry, 1975, 126, 397-407.

Slater, E., & Cowie, V. The genetics of mental dis-orders. London: Oxford University Press, 1971.

Slater, E., & Slater, P. A heuristic theory of neurosis.Journal of Neurology, Neurosurgery and Psychia-try, 1944, 7, 49-55.

Spring, B., & Zubin, J. Attention and informationprocessing as indicators of vulnerability to schizo-phrenic episodes. In L. Wynne, R. Cromwell, &S. Matthysse (Eds.), Nature of schizophrenia:New findings and future strategies. New York:Wiley, in press, (a)

Spring, B., & Zubin, J. Vulnerability to schizophrenicepisodes and their prevention in adults. In E.Albee (Ed.), Proceedings of the First VermontConference on the Primary Prevention of Psycho-pathology, in press, (b)

Tischler, G. L., Henisz, J. E., Meyers, J. K., & Bos-well, P. C. Utilization of mental health services.II. Mediators of service allocation. Archives ofGeneral Psychiatry, 1975, 32, 416-418. (a)

Tischler, G. L., Henisz, J. E., Meyers, J. K., & Bos-well, P. C. Utilization of mental health services. I.

Patienthood and the prevalence of symptomatologyin the community. Archives of General Psychiatry,1975, 32, 411-415. (b)

Turner, R. J. Unemployment and psychological dis-order: An assessment of the degree and nature oftheir association. Social Policy, in press.

Wallace, C. M. Portrait of a schizophrenic nurse.London: Hammond, Hammond, 1965.

Weiner, H. J., Akabas, S. H., & Somner, J. J.Mental health care in the world of work. NewYork: Association Press, 1973.

Weissman, M. The assessment of social adjustment:A review of technique. Archives of General Psy-chiatry, 1975, 32, 357-365.

Wender, P. H., Rosenthal, D., Rainer, J., Greenhill,L., & Sarlin, B. Schizophrenics' adopting parents:Psychiatric status. Archives of General Psychiatry,in press.

Wilkins, W. L. Social stress and illness in industrialsociety. In E. K. Gundersen & R. H. Rahe (Eds.),Life stress and illness. Springfield, 111.: Charles CThomas, 1974.

Wing, J. K., Cooper, J. E., & Sartorius, N. The mea-surement and classification of psychiatric symp-toms. Cambridge, England: Cambridge UniversityPress, 1974.

Wing, J., & Nixon, J. Discriminating symptoms inschizophrenia. Archives of General Psychiatry, 1975,32, 853-859.

World Health Organization. Report of the Interna-tional Pilot Study of Schizophrenia (Vol. 1).Geneva: WHO, 1973.

Zubin, J. Scientific models for psychopathology inthe '70s. Seminars in Psychiatry, 1972, 4, 283-296.

Zubin, J., & Burdock, E. I. The revolution in psy-chopathology and its implications for public health.Ada Psychiatrica Scandinavica, 1965, 41, 348-359.

Zubin, J., Salzinger, S., Fleiss, J. L., Gurland, B.,Spitzer, R. L., Endicott, J., & Button, S. Biometricapproach to psychopathology: Abnormal and clini-cal psychology—statistical, epidemiological, anddiagnostic approaches. Annual Review of Psychol-ogy, 1975, 26, 621-671.

Received September 30, 1975Revision received November 12, 1976

Date post:	25-Dec-2016
Category:	Documents
Upload:	bonnie
View:	212 times
Download:	0 times

Vulnerability: A new view of schizophrenia.

Documents