The life course as a theoretical orientation:Sequences of person-situation interaction'William McKinley Runyan, Instiiufe of Homan Development,University of California, Berkeley

ABSTRACT

The life course is proposed as a theoretical orientation concernedwith the problems of describing, understanding, generalizing about,predicting, and intentionally changing the course of lives. A lifecourse orientation provides a framework for analyzing the causaland probabilistic structure of the course of experience in individuallives, groups of lives, and lives in general.

The life course can be conceptualized as a sequence of person-situation interactions, or as a sequence of behavior-determining,person-determining, and situation-determining processes. This per-spective is illustrated through an analysis of the careers of heroinusers, and through a critical examination of several common strategiesfor predicting behavior. The study of lives is distinguished from thestudy of personality, and the historical and theoretical backgroundfor a life course orientation is briefly reviewed.

In the history of the behavioral sciences, we can identify anumber of theoretical orientations which have each had a pro-found influence upon the questions that psychologists ask, thepuzzles they try to solve, and the empirical research which theyconduct. The function of a theoretical orientation is to "indicatetypes of variables which are somehow to be taken into accoimtrather than specifying determinate relationships between par-ticular variables" (Merton, 1968, p. 142). Psychodynamic, trait,

1. This paper is based, in part, on an unpuhlished doctoral dissertation inClinical Psychology and Public Practice, Harvard University, 1975. My wannestthanks to the members of my dissertation committee: Edwin Barker, LawrenceKohlberg, Alexander Leighton, David Ricks, and Zick Rubin; to Henry Murray;and to John Clausen, Kenneth Craik, Hardson Gough, Norma Haan, RavennaHelson, Richard Lazarus, Laurie Wermuth, and an anonymous reviewer for theircomments on an earlier draft of the manuscript.

Requests for reprints should be sent to William McKinley Runyan, Instituteof Human Development, 1203 Tolman Hall, University of California, Berkeley,Califomia 94720.

570 Runyan

behavioral, and phenomenological orientations have each directedour attention to a specific range of phenomena, suggested a kindof order which may be discovered in the world, and had distinctimplications for theory construction, research design, and practi-cal affairs.

The intentions of this paper are to outline the life course asa theoretical orientation, and to suggest that order may be dis-covered in the analysis of sequences of person X situation inter-action. The life course may be defined as the sequence of eventsand experiences in a life from birth imtil death, and the chain ofpersonal states and encountered situations which influence, andare influenced by this sequence of events. The generative ques-tions for a life course orientation are: What kinds of order orregularity may be found in the sequence of events and processesin individual lives, groups of lives, or lives in general? What isthe causal and probabilistic structure of the course of experience?What processes (cognitive, affective, physiological, social, andhistorical) are responsible for the flow of subjective experienceand overt behavior over time? Most simply, how do our liveswork?

A life course orientation is concerned with the enduring prob-lems of describing, imderstanding, making predictions about, anddesigning and evaluating efforts to change the course of lives.Specific generalizations about the life course may come and go(cf. Cronbach, 1975; Gergen, 1973), but these problems remain.To the extent that the structure of the life course varies acrosscultures and historical periods, our theories may need to be con-tinually revised and freshly constructed.

The objectives of a life course orientation are somewhat dif-ferent than those in the study of personality. Consider thefollowing question: What are the aspirations of personality psy-chologists? If personality psychologists could know anything inthe world, what would they most like to know? Answers willundoubtedly vary widely, but representative aspirations might bedescribed as (a) understanding the psychological mechanisms andprocesses which regulate the expression of impulses, the satisfac-tion of needs, and the contents of consciousness, (b) identifyingand measuring the major dimensions of personality, determiningtheir correlational structure, and assessing their relationship tooutcomes in the natural environment, (c) imderstanding the waysin which behavior is controlled by situational stimuli, (d) under-

The life course 571

standing how phenomenological awareness and self-concepts af-fect behavior, or (e) understanding how persons and situationsreciprocaiYy mteracf nr cfeiferrmimTg' iIft5iV<r*i\a". i r lasnvtaifi* ^isfdisfof the life course have more explicit temporal and ecologicaldimensions, and are concerned with questions about the causalstructure of sequences of experience, about the processes throughwhich persons and situations interact over the course of time,and about the movement of persons through social and historicalworlds. The analysis of persons or personalities is not the same asthe analysis of lives. Persons are one component of "the life struc-ture" (Levinson et al., 1978), or of the life history process. Thesedistinctions have not always been made, and it should be notedthat study of the life coiu se is congruent with several streamswithin the history of personality psychology, such as the person-ology of Henry Murray (1938), Robert White's (1966) study oflives, and recent investigations in interactional psychology (Mag-nusson & Endler, 1977).

The second section of this paper discusses sequences of personX behavior X situation interaction as an approach to conceptualiz-ing the life course. The third section illustrates this orientationthrough an analysis of careers of heroin use, while the fourthsection examines the implications of a life course orientation forthe prediction of behavior. The fifth section provides a shortoverview of the historical and theoretical background of a lifecourse orientation.

A Sequential Interactionist Conceptualization of the Life Course

The life course may be conceptualized as a sequence of personX situation interactions, or, more specifically, as a sequence ofperson X behavior X situation interactions. This is certainly notthe only way of conceptualizing the life course (Runyan, 1975),but it is a useful one, and one which relates the life course tocurrent discussions of interactionism in personality and socialpsychology (e.g., Alker, 1972; Argyle & Little, 1972; Bem & Allen,1974; Bowers, 1973; Ekehammar, 1974; Endler & Magnusson,1976; Magnusson & Endler, 1977; Mischel, 1973; and Wachtel,1973).

Olweus (1977) has distinguished at least four different mean-ings of the term interaction: (1) unidirectional interaction, or thecombination of person and situation variables in determining be-havior, (2) the interdependency of person and situation variables.

572 Runyan

as in the situation being defined by a person's perception of it,(3) reciprocal interaction, or the ways in which persons andsituations influence each other over time, and (4) statistical in-teraction, as used in the analysis of variance. The limitations ofinteraction terms in the analysis of variance for studying personX situation interaction have been convincingly argued (Colding,1975; Olweus, 1977). The fact that persons can perceive and con-ceptualize situations in a variety of ways, and that the perceivedsituation may be influential in determining behavior, does not, inmy view, make persons and situations analytically inseparable(sense 2). So for this paper, interaction will refer to the combina-tion of personal and situational variables in determining behavior(sense 1). This unidirectional interaction will be distinguishedfrom reciprocal interaction over time (sense 3).

A life course orientation is concerned with sequences of ex-perience in the natural environment, and thus with the processesthrough which persons, situations, and behaviors influence eachother over the course of time. This is interactionism with, an ex-plicit temporal dimension, and is consistent with what Olweus(1977) and Bandura (1977) describe as reciprocal interaction,or Magnusson and Endler (1977) describe as "dynamic" inter-action. It is also interaction with an explicit ecological dimen-sion, in its concem with persons' movements through a socialstructure and an historical world over the course of time. Thispaper furthers the discussion of interaction sequences by analyz-ing the empirical probabilities of changes in persons, behaviors,and situations over a four-stage sequence (in careers of heroinuse), by indicating the operation of different causal factors atdifferent stages in the process, and by briefly exploring the impli-cations of a sequential interactionist framework for prediction.

It is widely agreed (e.g., Magnusson & Endler, 1977; Overton& Reese, 1973; Pervin, in press) that recognizing interaction isnot sufficient, and that attention needs to be paid to the how ofinteraction, or to the cognitive, affective, physiological, and socialprocesses which determine how persons function within situ-ations.

The life course can be conceptualized as a sequence of pro-cesses, as well as a sequence of interactions (cf., Haan, 1977, whoargues that personality can best be conceptualized as processesand their organization). Three of the most general processes that

573

PERSONA SITUATION

BEHAVIOR-BIRTH

Figure 1. An interactional model of the life course.

need to be considered are (1) behavior-determining processes, re-sulting from the interaction of persons with situations, (2) person-determining processes, or the processes which create, maintain,and change personal states and characteristics, and (3) situation-determining processes, or the processes through which peopleselect, create, and influence the situations they encounter. Therelationship between these three processes over time is repre-sented in an interactional model of the life course in Figure 1.

Each of the arrows in Figure 1 represents one of the threecausal processes mentioned above. There are no arrows directlylinking persons and situations with each other, which representsthe assumption that the effects of persons upon situations areusually mediated through their behavior, and that the effects ofsituations upon persons are typically mediated by the person'sexperience within the situation. However, for a few analyticalptirposes, such as studying the effects of physical attractiveness, itwill be useful to consider causal arrows which directly link personsand situations with each other. It should be noted that the personcomponent of this diagram is not restricted to psychological

574 Runyan

Figure 2. Sequences of person X behavior X situation interaction.

variables, but also includes biological, social, and economic vari-ables, such as sex, appearance, credentials, and wealth.

Personal variables, such as intelligence or self-concept, andsituational variables, such as home or college, are frequently ofgreater duration than behavioral or experiential variables, suchas a specific act, thought, or feeling. At the risk of using a clumsymetaphor, the stream of behavior flows through banks of personaland situational variables which are of relatively greater stability.

An altemative way of representing a sequence of person Xsituation interactions is contained in Figure 2. In this diagramthere are no causal arrows which link behaviors directly toeach other, representing the assumption that behaviors cannot cause each other, except as mediated through changesin the organism, or less directly, through changes in the situa-tion. For some purposes, it would be useful to add arrowscoming from outside the system to indicate maturational or non-experiential sources of change upon person variables, and to in-dicate situational changes not caused by the person.

Figures 1 and 2 may be used for conceptualizing the temporalcontext of a specific behavior, for examining a particular se-quence, process, or career, or for conceptualizing the entire lifecourse. It should be clear that these diagrams of person X be-havior X situation interaction represent only one way of con-ceptualizing the life course, and may be expected to prove usefulfor many, but not all, analytic purposes.

An Analysis of Heroin Use from a Life Course Perspective

A specific substantive example, such as an analysis of heroinuse, may help to illustrate the implications of a life course orien-tation, and to compare it with other orientations. Using a traitor psychodynamic view, one is led to raise questions about the

The life course 575

relationships of heroin use to personality variables. The person-ahty characteristics of addicts have been assessed with theRorschach, TAT, interviews, MMPI, CPI, Rotter Intemal-Ex-temal Control Scale, the Edwards Personal Preference Schedule,etc. (Kurtines, Hogan, & Weiss, 1975; Reith, Crockett, & Craig,1975). Drug addicts have been characterized as maladjusted,hostile, immature, dependent, manipulative, and narcissistic(Feldman, 1968). One weakness of this approach is that mostpeople with these characteristics do not become heroin addicts.A second problem is that the causal relationships between per-sonality variables and heroin use are often unclear. Which per-sonality characteristics led to heroin use, and which are a con-sequence of heroin use? There is, for example, evidence thatelevations on the Hypochondriasis and Hysteria scales of theMMPI are a temporary consequence of the addict's stressful lifeon the street, rather than indicators of enduring personality char-acteristics which led to heroin use (Sheppard, Ricca, Fracchia, &Merlis, 1973; Sutker & Allain, 1973).

Using a situational orientation, one is inclined to ask questionsabout the contexts or environments associated with narcotics use.Heroin use in New York City has been foimd to be heaviest inneighborhoods with the highest poverty rates, the most crowdeddwelling imits, the highest incidence of broken families, and soon (Chein, Gerard, Lee, & Rosenfeld, 1964). A limitation of thisapproach is that most people exposed to these conditions do notbecome heroin users. A second problem is that some features ofthe environment may be a function of heroin use as well as a causeof it. A similar difficulty arises in ecological studies of mentalillness and criminal behavior, in which it is unclear to what ex-tent people with deviant tendencies move into high prevalenceareas, and to what extent deviant behavior is elicited by the socialenvironment (Moos, 1976).

Using an interactionist position, one is led to formulate ques-tions about which individual tendencies interacting with whatsituational conditions will lead to heroin use. This orientationpermits greater precision than the first two, but it does not pro-vide an imderstanding of which persons get into which situations,or of how their behavior affects them and affects their environ-ment, which will in tum affect future behavior.

A life course orientation builds on an interactionist perspec-

576 Runyan

tive and suggests a more complex set of questions about thetemporal course of behavior. The questions about heroin usesuggested by a life course orientation are: (1) What kinds ofpersons in interaction with what kinds of situations lead to initialuse of heroin? (2) How do these person X situation configura-tions come about? (3) What are the consequences of initial usefor the person and for the situation? (4) Which new person Xsituation configurations will result in continuation vs. termina-tion of opiate use? (5) If heroin use continues and leads toaddiction, what are the further consequences for the person andthe situation? (6) What person X situation processes will leadto continuation or termination of addiction? (7) If the personbreaks the habit, what new person X situation configurationsmay lead to relapse?

The temporal course of heroin use can be usefully dividedinto the four stages of experimental use, occasional use, addictiveuse, and termination of use. How can the initial or experimentaluse of heroin be explained? In Chein et al.'s (1964) study ofheroin users in New York City, the first opportunity for heroinuse came through offers of a friend, or in a group setting. Fewaddicts actively sought this first opportunity. Even so, exposureto these situations was not random. The Chein study comparedthe experiences of four groups: nondelinquent addicts, delin-quent addicts, nonusing delinquents, and controls with similarsocial characteristics living in the same neighborhoods. Fifty per-cent of the control group had an opportunity to try heroin, com-pared to two-thirds of the delinquent nonusers, and to 100 per-cent of the delinquent and nondelinquent users.

Differential exposure to situations where heroin was availableis a partial explanation of initial use but not a complete one. Re-sponses to this opportunity differed radically among the fourgroups. Of the nonusing delinquents, only 4 out of the 36 whohad the opportunity tried it.

How can the differential response of these groups to the op-portunity to try heroin be explained? Only 17 percent of the users(interviewed in 1953) claimed that they knew anything beforeage sixteen about the harmful effects of heroin, while 79 percentof the delinquent nonusers said that they were aware of its nega-tive consequences in areas such as health, pressure toward crim-inal activity, and effect on character. Although part of these dif-

The life course 577

ferences may be due to retrospective distortion, the data stillsuggest that the anticipated consequences of heroin use dif-ferentiated those who responded to the opportunity from thosewho refused it.

Second, what explanation can be given of the transition fromexperimental to occasional use of heroin? One variable is theeffect of initial use on the person. Those who experience a posi-tive reaction are more likely to progress to occasional use thanthose who do not. In Chein's sample, about half of the usersreported a positive reaction upon first use, while a third reportednegative reactions. This first reaction influenced the timing ofsubsequent use. Of those with favorable reactions, two-thirdscontinued immediately; while of those with imfavorable reactions,only two-fifths continued immediately. Given the appropriateindividual predispositions, occasional use of a drug may continuelargely as a function of its easy availability (Clausen, 1966).

Third, the transition from occasional to addictive use hasbeen conceptualized in a variety of ways. When heroin is usedregularly, physical tolerance develops, and increased doses areneeded in order to obtain the same subjective effects. A majoraspect of the transition to addiction is the growth of physicaldependence and the desire to avoid the pain of withdrawal(Lindesmith, 1947). Recent evidence indicates that habitualnarcotics users are motivated by a continuing desire for euphoria,as well as by avoidance of withdrawal symptoms (McAuliffe &Gordon, 1974).

Another theory emphasizes that the person may become la-belled as a deviant, perhaps be fired from his job, and be cut offfrom participation in more conventional groups (Becker, 1963).This labelling process can be accompanied by changes in self-conception and changes in attitudes toward society. In addition,the heroin user may choose to disengage himself from people whodisapprove of drugs and to select new friends primarily amongother users (Feldman, 1968). Thus, the heroin user can createan environment for himself in which narcotics are more readilyavailable, and where heroin use is not looked upon with suchdisapproval.

Heroin has fewer direct physical effects than is generallyrealized (Goode, 1973), but depending upon social, legal, andeconomic conditions, addictive use of heroin may radically alter

578 Runyan

the person's whole way of life. For physicians, who have rela-tively free access to narcotics, addiction does not necessarilycause major life changes. There are wide variations across cul-tures and historical periods in the social, medical, and criminolog-ical correlates of heroin use, but for the contemporary streetaddict in the United States, addiction frequently leads to healthproblems, to crime, to alienation from family and friends, andto prison.

Fourth, heroin use may be terminated or reduced through awide variety of person X situation processes. The breaking ofphysical dependency does not, however, ensure the end of ad-diction. Out of 1,912 addicts living in New York City who weredischarged from the drug rehabilitation program in Lexington,Kentucky, more than 90 percent were re-addicted six monthsafter discharge (Hunt & Odoroff, 1962).

In spite of high re-addiction rates, there is a tendency forheroin use to decrease over time and to decrease with increasingage. In a sub-sample of the Hunt and Odoroff group, 9 percentwere voluntarily abstinent six months after discharge, 17 percenttwo years after discharge, and 25 percent five years after dis-charge (Duvall, Locke, & Brill, 1963). Behavior change maycome through a change in the situation (no drugs available),through changes in the person (physical detoxification, a de-cision to change), or through some interaction of personal andsituational factors (losing a supplier and deciding not to makethe effort to locate a new one). However, the various sequencesof changes in person X situation interaction which lead to thetermination of heroin use are not well understood.

In review, let us formally conceptualize a sequence of processesinvolved in the four stages of a drug use career. At each stage,rough estimates of the state of selected person variables, situationvariables, and behavior probabilities can be given. Althoughsomewhat schematic, this should suggest a general approach toanalyzing sequences of person X behavior X situation inter-action.

At stage one, the period of experimental use, heroin is avail-able infrequently, the person's desire for it is low-to-moderate,and the probability of taking the drug is moderately low. Atstage two, the period of occasional use, heroin is available oc-casionally, but perhaps not much more frequently than at stage

The life course 579

one. If the individual has found his first experience pleasurable,the probability of taking heroin rises.

At stage three, the period of regular or heavy use, the wholeperson-behavior-situation configuration may change dramatically.The addict may be in a situation where heroin is readily avail-able, and where most of his friends are drug users. His desirefor heroin may be so intense that he is willing to do anythingin order to obtain it. In situations where drugs are available, theprobability that he will take them rises to near certainty. As aproduct of his increased desire, and of the increased availabilityof drugs, the frequency of heroin use may jump from once amonth or once a week to three or four times a day.

At stage four, the period of discontinued use, change may oc-cur in a variety of ways, such as a decision to withdraw on one'sown, methadone treatment, or involvement with Synanon. Theprobability of taking heroin, if it is available, may shrink to near-zero. Environmental changes, such as being placed in prison,may make temporary withdrawal inevitable. However, termina-tion of use is not necessarily permanent, and depending upon anindividual's characteristics and later environments, the wholeprocess may begin again.

In summary, a history of heroin use is characterized by a se-quence of person X behavior X situation interactions. Psycho-dynamic, trait, behavioral, sociological, or interactive orientationsmay all shed light on some phases of narcotics use, but the careerof a heroin-user cannot be adequately understood without con-sidering sequences of behavior-determining, person-determining,and situation-determining processes.

Similar analyses may be made of histories of aggressive be-havior, sexual behavior, altruistic behavior, abnormal behavior,creative behavior and other forms of behavior which vary withboth changing personal states and encountered situations. A se-quential process orientation can also be used for analyzing thetemporal trajectory of personal characteristics which are sub-ject to influence by situations and by experience, such as self-concept, level of anxiety, values, or attitudes. Finally, a life courseorientation can be used to study "environmental histories" or"situational careers," and to explore the processes responsible foran individual's movement through a sequence of situations. Insummary, a life course orientation provides a common conceptual

Runyan580

framework for analyzing sequences of behavior, sequences ofpersonal states, and sequences of encountered situations.

The utihty of a life course or sequential process orienta-tion can be suggested by considering research examples from anumber of other substantive areas. In a study of disturbed pre-adolescent boys, Raush (1965) examined the "contingenciesthrough which the antecedent action of one child exerts controlon the subsequent actions of another" (p. 488). Normal andhyperaggressive boys, aged 9-12, in groups of 6, were systemat-ically observed in 6 different situations (e.g., breakfast, or struc-tured games). The two control groups produced 85 percent and89 percent antecedent acts that could be classified as friendly,while the hyperaggressive group produced 58 percent friendlyacts In response to unfriendly antecedent acts, the conbrol anddisturbed boys both responded in an unfriendly way about 80percent of the time. However, in response to antecedent actswhich the staff saw as friendly, the normals responded in an un-friendly way only 8 percent of the time, while the hyperaggres-sive group responded in an unfriendly way 45 percent of the timeearly in treatment, and 19 percent of the time after L5 years ottreatment. The hyperaggressive and control groups did not ditterin their response to hostile acts, but did differ in response tofriendly acts. Putting together a sequential chain of five be-haviors, the hyperaggressive boys were behaving in a friendly way40 percent of the time by the fifth behavior, while the two normalgroups were behaving in a friendly way approximately 80 percent

of the time.Wachtel (1977), in an effort to integrate features of psycho-

dynamic and behavioral theory, has spoken of "interactioncycles," in which early psychodynamic processes affect subse-quent experiences and later environments, which in tum affectthe continuation or change of earlier psychodynamic processesand so on. Within this interactional-cyclical view, "the criticalrole of childhood is understood in terms of the way in which theparticular pattems of behavior one develops skew the kinds oflater experience one is likely to encounter and hence create anidiosyncratic environment of a sort likely to maintain the verypattem which produced that kind of environment in the firstplace" (Wachtel, 1977, p. 320). As an example, Wachtel con-siders the case of a man with a reaction formation against in-

The life course 581

tense rage. He is frightened by feelings of anger, so acts in ex-cessively meek or conciliatory ways, which lead others to takeadvantage of him. This mistreatment arouses further feelings ofresentment and anger. His defensive strategy has led to ex-periences which generate anger, and to the continued need todefend against it.

As a third example. Elder (1974) examined the life coursesof a group of 167 persons bom in 1920-21 in Oakland, Califomia,and studied periodically since 1932 at the Institute of HumanDevelopment, University of Califomia, Berkeley. He comparedthe life experiences of middle-class and working-class childrengrowing up in famiUes that were deprived (35 percent or morereduction in income) or nondeprived during the Depression. Onesurprising finding was that Vidthin each social class (middle vs.working class), economic deprivation in childhood did not leadto substantially lower occupational attainment at age 47. Withinthe working class, economic deprivation in childhood did reduceeducational attainment, but this was largely offset by a sequenceof events which went roughly as follows: men from deprivedworking-class families were faced earlier with adult-like re-sponsibilities in the family, which led to earlier crystallization ofvocational goals, and to earlier establishment of a stable careerline, which led to higher occupational attainment. In this ex-ample, and in others throughout the book. Elder's work illustratesthe possibilities for studying sequences of person X situation in-teraction in the natural environment.

The importance of studying sequences and processes of personX situation interaction has recently been stated in a number ofgeneral theoretical formulations (e.g., Bandura, 1977; Magnusson& Endler, 1977; Pervin, in press), and in analyses of interpersonalrelationships (Peterson, 1977), marital interaction (Raush,Barry, Hertel, & Swain, 1974), and processes of stress and coping(Lazarus & Launier, in press).

Implications for the Prediction of Behavior

The limitations of atemporal perspectives in personality theorybecome particularly apparent when one is concemed with theissue of prediction. A life course orientation, considering the in-teraction of personal, behavioral, and situational variables overtime, provides a more adequate foundation for predictive efforts

582 Runyan

than do orientations focussing primarily on traits, psychodynamicprocesses, or situational influences on behavior.

The utility of predictions based on inferences about globaltraits has been seriously questioned. It has been argued thattrait-based predictions have not been very successful, except inthe domain of intelligence and ability variables (Mischel, 1968;Peterson, 1968; Vemon, 1964).

According to one view, the disappointing results of predictivestudies indicate that the trait paradigm is faulty (Mischel, 1968,1973), while another view is that predictive failures are often aresult of seriously defective research, and that adequate researchdoes provide examples of predictive success (Block, 1977). Al-tematively, as will be argued here, the trait paradigm may be ap-propriate under a limited and specifiable set of conditions.

If one assumes that behavior is a function of stable individualtraits, then it makes sense to measure these traits, and to usethese measurements in predicting future behavior. This strategycan be effective only under a restricted set of conditions. If atrait, such as intelligence or introversion-extraversion, is relativelystable over time, and is fairly consistent across situations, thentrait assessments can yield valid predictions. Conversely, to theextent that behavior is situationally or interactively determined,and to the extent that person variables change as a result of ex-perience, a trait approach to prediction cannot be expected to bewidely effective.

A behavioral or social leaming orientation is frequently pro-posed as an altemative to trait approaches. The implications ofa social leaming orientation for the assessment and change of be-havior are fairly clear, but what are its implications for the pre-diction of behavior? A social behavior orientation emphasizes thespecificity of behavior, or its dependence upon the environmentalconditions which maintain and change behavior. The predictiveimplications of this view are that, most generally, future behaviorwill depend upon future situations. In conditions where tiiefuture situation is known, assessment and prediction proceed bytaking "samples" of a person's behavior in similar situations. "Aperson's relevant past behaviors tend to be the best predictors ofhis future behavior in similar situations" (Mischel, 1968, p. 135).If behavior is determined largely by situations, the strategy ofpredicting future behavior from samples of past behavior canbe successful only to the extent that future situations are similar

The life course 583

to past situations. The prediction of future situations has neverreceived much attention, but if social behaviorists, or interac-tionists, seek to accurately predict future behavior, research onsituation-determining processes in the natural environment seemsto be a necessity.

Even in those circumstances where future situations are known,the social leaming paradigm will be useful under only a limitedset of conditions. To the extent that relevant person variableschange during the time interval between assessment and criterionperiods, predictions from behavior samples are hkely to be in-accurate. For example, suppose that we want to predict the re-sponse of a woman with a snake phobia to handling a live snakethree weeks from now. As part of a systematic desensitizationtreatment, she has constructed an anxiety hierarchy rankingsituations related to snakes from least to most anxiety provoking,with "handling a live snake" as the most frightening situation.The strategy of predicting future behavior on the basis of past be-havior in similar situations would lead one to assess her responseto this situation at Ti, and predict that if exposed to the samesituation at T3, her behavior would most hkely be the same. Theweakness in this predictive strategy is obvious. If the subjectspends the intervening time period working her way through thestimulus hierarchy, then her response at T3 is likely to be quitedifferent. Knowledge of likely intervening experiences enablesus to improve over predictions based solely on past behavior insimilar situations.

In itself, this is a minor example, chosen for its transparency.But in principle, is the situation any different in the prediction ofvocational performance, criminal behavior, or interpersonal be-havior? The accuracy of behavioral prediction depends upon as-suming that person variables remain stable. To the extent thatrelevant person variables change over the predictive interval, pre-dictions based on behavior samples are likely to be inaccurate.

If an interactionist position is taken seriously, what are itsimplications for the prediction of behavior? An interactionistorientation implies that future behavior will be determined byan interaction of future person states with future situations. Thissuggests that the problems of assessment and prediction cannotbe solved by the consideration of personal or situational variablesalone.

A life course orientation suggests a strategy for integrating

584 Runyan

person-centered and situation-centered assessment, enabling usto avoid some of the limitations inherent in either approachtaken individually. This strategy focusses on the assessment ofperson-situation configurations, and on the likelihood of alter-native person-situation configurations developing out of thisinitial system. For example, in an effort to predict the termina-tion of heroin addiction, assessments would be made of personalcharacteristics and of the person's location in an environmentalnetwork. For men living in New York City, the probability ofvoluntary heroin abstinence one year after treatment for ad-diction was about lO-to-15 percent (Duvall et al., 1963). Foraddicts in the army in Viet Nam, the probability of abstinenceone year after retum to the United States was greater than 90percent (McGlothlin, 1975). Our best predictions of the termina-tion of heroin use would depend upon assessment of subjects'locations in environmental networks, and upon their likelihoodof moving to altemative future environments, as well as uponassessment of their personal characteristics.

In summary, trait-based approaches to prediction are usefulonly when person attributes can be assumed to be stable overtime, and only to the extent that behavior is stable across pre-dictor and criterion situations. Situational or behavioral ap-proaches are useful only when relevant person variables can beassumed to be stable, and only to the extent that future situa-tions are knovra or that behavior is invariant across the range ofexpectable future situations. If behavior is determined by theinteraction of persons with situations, then a comprehensiveapproach to prediction must assess person-behavior-situation con-figurations, and estimate the probability of altemative configura-tions developing out of this initial system. This more complexview of the prediction problem suggested by a life course orien-tation will, in some domains, enable us to increase our predictivecapacities, but in other areas, it can help us to recognize exag-gerated predictive claims, and to better understand those condi-tions which inevitably limit the possibilities of accurate predic-tion.

Historical and Theoretical Background

Life histories have been studied, with waxing and waningdegrees of enthusiasm, by a wide range of social scientists

The life course 585

throughout the twentieth century. The history of life historystudies in psychology is discussed by Allport (1942) and Murray(1955); in anthropology by Kluckhohn (1945) and Langness(1965); in sociqlogy by Angell (1945) and Becker (1966); andin political science by Greenstein (1969) and Glad (1973).

Some authors (e.g., Denzin, 1970) have spoken of a "lifehistory method," by which they mean the procedure of havinga subject tell his or her own story. However, the term life historyalso refers to the sequence of events in a life. In this latter sense,it should be clear that there is no single life history method, anymore than there is a single personality research method, and thatlife histories may be studied through phenomenological reports,the case study method, survey research, longitudinal research,and experimental research. The term life history applies mostreadily to a past sequence of events, while life course is a moregeneral term, which may refer to either past or future sequencesof life events.

Focussing on the disciplines of psychology and sociology,we can roughly identify three periods in the study of life his-tories. From approximately 1920 through the 1940s, there wasa substantial and growing interest in the study of life histories,much of it associated with the analysis of personal documents(e.g., Thomas & Znaniecki, 1920; Shaw, 1930; Buhler, 1933;DoUard, 1935; Murray, 1938; Blumer, 1939; Allport, 1942;Gottschalk, Kluckhohn, & Angell, 1945; Leighton & Leighton,1949).

In the 1950s and early 1960s, the energies of psychologistsseemed directed more toward laboratory experimentation, andrelatively little was published in the study of life histories, al-though there were some significant exceptions (e.g., Pressey &Kuhlen, 1957; Erikson, 1958). As dissatisfaction with laboratoryexperimentation as an exclusive methodology increased in thelate sixties and seventies, more attention was devoted to problem-oriented research, to field experiments, to cognitive processes, andto interactive paradigms. During this period, studies of the lifecourse and of life transitions increased dramatically (e.g., Buhler& Massarik, 1968; Block with Haan, 1971; Dailey, 1971; Clausen,1972; Elder, 1974; Lowenthal, Thumher, & Chiriboga, 1975;Sheehy, 1976; Vaillant, 1977; Levinson et al., 1978).

An understanding of the course of lives and of the processes

586 Runyan

through which persons interact with situations over the course oftime is currently pursued in a remarkable range of fields. It is,I believe, worth identifying the range of fields having a commonconcern with analyses of the life course or of sequences of personX situation interaction, and not proceeding as if the issues werean exclusive concern of a single field. The study of person Xsituation interaction in personality psychology, although sug-gested in the work of Murray (1938) and Lewin (1951), has onlyrecently received a great deal of attention (e.g., Argyle & Little,1972; Bowers, 1973; Ekehammar, 1974; Endler & Magnusson,1976; Magnusson & Endler, 1977). A wide-ranging and sophisti-cated discussion of the interaction of personal and situationaldeterminants of behavior, which seems to have been overlookedin this literature, was previously presented by sociologist J. M.Yinger (1965).

In clinical psychology and psychiatry, life history researchersin psychopathology are investigating the interacting biological,psychological, and sociological antecedents and consequents ofdifferent forms of deviance, such as schizophrenia, neuroses, anddelinquency (e.g., Roff & Ricks, 1970; Ricks, Thomas, & Roff,1974). They are also concemed with "constmcting a life historyframework for the interpretation of intervention and outcomestudies" (Ricks, Thomas, fit Roff, 1974, p. viii).

Within developmental and hfe-span developmental psychol-ogy, there is an increasing emphasis on interactive analyses ofdevelopmental processes, and on the ecology of human develop-ment (Bronfenbrenner & Mahoney, 1975). Nesselroade andBaltes (1974) argue that a substantial amount of influence onthe developmental process lies outside the person, and that studiesof human development cannot be confined within the disciplineof psychology. They believe that "future research should focuson conjoint measurement and analysis of organism-environmentsystems" (p. 71).

There are significant similarities between a life course orien-tation and a life-span developmental orientation in that bothare concemed with processes of person X situation interactionover time, yet there are also substantial differences. It is diffi-cult to attain consensus on a precise definition of developmentalpsychology, yet it is widely agreed that developmental changesoccur in a relatively common sequence, and are generally ir-

The life course 587

reversible. Developmental changes are most significant in child-hood and in old age, the two areas that have received the greatestresearch attention. However, not all changes over time are de-velopmental, and particularly not in the lengthy period betweenmaturation and degeneration. The effects of psychotherapy, or ofheroin use, or of getting a new job cannot, in any meaningfulsense of the term, be considered developmental. If developmentalchanges are defined as changes due to maturation or aging, thenclearly, developmental changes are only one kind of change overtime. Even though events and changes in adult life are studiedby developmental psychologists, it is awkward trying to fit theminto any systematic definition of developmental psychology. An-other way of making the distinction is to say that a person isfaced with a wide variety of ways of living, of possible lifecourses, whUe facing a limited number of ways of developing.Persons are free to create the course of their lives, or to direct thecourse of their experience, in ways that they are not able to alterthe course of their physical or psychological development.

Interest in the life course is not confined to psychology, butis also emerging within several sub-fields of sociology, such associal mobility research (Blau & Duncan, 1967), socializationresearch (Clausen, 1968), and the sociology of aging (Elder,1975), all of which share a concem with the interaction of chang-ing persons with the social structure over the course of time.Aspects of the life course are also being investigated in familyhistory and quantitative historical biography (Shorter, 1975;Stone, 1972; Thernstrom, 1973), and in psychohistory and psycho-biography (Lifton & Olson, 1974; Mazlish, 1971).

One additional field related to the study of life courses isevaluation research or outcome research (Bergin & Garfield,1971; Struening & Guttentag, 1975). A common endeavor ofthese studies is to trace the effects of a given intervention, suchas psychotherapy, compensatory education, or job training uponthe subsequent characteristics or experiences of its participants.The net value of an intervention is determined by its effects on asequence of future experiences, and not by its effects on a singledependent variable at a single point in time. Thus, some formsof outcome research must investigate sequences of experience,and can be informed by an understanding of life courses. In tum,the systematic evaluation of social action programs can provide

588 Runyan

us with experimental or quasi-experimental studies of the lifecourse.

In summary, studies of the life course, or studies of the pro-cess of hving, are not an exclusive concem of personality psy-chologists, but are also related to selected lines of work in de-velopmental, social, and clinical psychology, to several subfieldswithin sociology, to psychohistory and quantitative history, andto longitudinal outcome research.

CONCLUSION

There is little doubt that order can be found, and has beenfound, by those pursuing psychodynamic, trait, behavioral, andphenomenological orientations. By using a criterion of "statisticalsignificance," some empirical support can be found for each ofthese orientations. No single orientation has a monopoly uponempirical support. By using the more rigorous criterion of pro-portion of variance accounted for, or of complete predictability,it also becomes apparent that no single orientation provides acompletely adequate theory of behavior.

Each orientation seems to find its strongest support in differ-ent methodological foundations, and to be particularly suited fordifferent applications. For example, a behavioral view seems toreceive its strongest support on experimental groimds, and to beparticularly suited for problems of behavior change. A psycho-dynamic view seems to receive its strongest support from clinicalobservation and from personal experiences in therapy, and to bemost useful for the interpretation of disturbed thought, affect,and behavior. A phenomenological view seems to receive somesupport from both clinical and scientific contexts, and to be mostuseful in the conduct of interpersonal relationships. A trait-factor approach seems to receive its strongest support from cor-relational research, and to be most useful for purposes of selec-tion and decision-making. (This comparative assessment is notintended to imply that all orientations are equal in scope, power,or utiHty, but rather is an effort to comprehend the relationshipsof the different orientations to each other, and to understandhow intelligent and well-meaning scientists can be so passionatelydivided in their theoretical loyalties.)

Each orientation may appear unimpressive when viewed fromdifferent epistemological bases, or when applied to different

The life course 589

uses. For example, a psychoanalytic view does not receive verystrong support when tested experimentally, or, a trait-factor ap-proach is not that useful for the conduct of psychotherapy. Al-though the strengths and weaknesses of these orientations mightwell be characterized in other ways, it is unlikely that this wouldlead to disagreement with the view that each of these orienta-tions has discovered some form of order in the world, each hasrelied upon somewhat different procedures for discovering thisorder, and each has proven useful for some practical concems,but not for others. Similarly, a life course orientation (1) suggestsa type of order that may be found in the world, (2) requiresmethods and strategies appropriate for discovering this kind oforder, and (3) has implications for a particular range of prac-tical concems.

To expand briefly on these points, a life course orientationsuggests that order may be found in the causal structure of theflow of experience, and in the accompanying processes of personX behavior X situation interaction. It directs our attention to afundamental and enduring set of problems in describing, under-standing, making predictions about, and intentionally changingthe course of lives.

The methodological problems are enormous and demand ourcreative attention. Understanding of the cotirse of lives may bepursued through (a) quantitative experimental and correlationalresearch, (b) historical and biographical research, (c) analysesof personal experience, and (d) clinical and professional prac-tice.

We need research designs capable of studying sequences ofperson X behavior X situation interaction, in both natural and con-trolled settings. However, understanding the course of lives de-mands more than the development of more complex researchdesigns; it also requires a shift in our strategies of explanationand prediction away from purely nomothetic-deductive ones inwhich individual cases are subsumed imder general laws. Any-one who has attempted to interpret a life history using solelygeneral theories will appreciate the hmitations of this approach.Due to the enormous variety and change in the course of ex-perience across cultures, across historical periods, and even acrossindividuals, greater attention must be paid to the development ofidiographic methods, N = 1 methodology, and explanation via

590 Runyan

patteming or collagation. A valuable alternative model is outlinedby Bromley (1977), who advocates a "quasi-judicial method"for the collection and interpretation of evidence in case studies.

Increased understanding of the course of lives, and the de-velopment of appropriate research strategies should prove useful(a) for social scientists in their theoretical concems, (b) forpsychobiographers and psychohistorians, in their efforts to de-scribe and interpret the lives of public and historical figures, (c)for clinicians and professionals, in those tasks of assessment, pre-diction, intervention, or evaluation which involve the course ofexperience (e.g., Runyan, 1977), and (d) for individuals, in try-ing to understand, predict, and make decisions within the courseof their OWTI lives.

Skeptics may feel that the study of lives is impractical, orhopelessly complex. It is granted that research on the life coursemay be methodologically complex, and that absolute certaintyin the understanding of lives may be forever beyond us. However,assumptions about the causal structure of the course of experienceand about the long-range effects of different courses of actionupon this experience are inextricably bound into our personallives, out intellectual and professional lives, and even our socialpolicies. Demands for interpreting, making predictions about,and intentionally changing the course of lives are ever beforeus. Is there any viable course other than to pursue these tasks asrigorously and intelligently as we can?

REFERENCES

Alker, H. A. Is personality situationally specific or intrapsychically consistent?Joumal of Personality, 1972, 40, 1-16.

Allport, G. The use of personal documents in psychological science. New York:Social Science Research Council, 1942.

Angell, R. A critical review of the development of the personal documentmethod in sociology, 1920-1940. In L. Gottschalk et al. (Eds.), The use ofpersonal documents in history, anthropology, and sociology. New York: SocialScience Research Council, 1945.

Argyle, M., & Little, B. Do personality traits apply to social behavior? Joumalfor the Theory of Social Behavior, 1972, 2, 1-35.

Bandura, A. Sodcd learning theory. Englewood Cliffs, N.J.: Prentice-Hall,1977.

Becker, H. Outsiders: Studies in the sociology of deviance. New York: FreePress, 1963.

Becker, H. Introduction. In C. Shaw, The jack-roller. Chicago: U. of ChicagoPress, 1966.

Bem, D., & Allen, A. On predicting some of the people some of the time: Thesearch for cross-situational consistencies in behavior. Psychological Review,1974, 81, 506-520.

The life course ^'^

Bergin, A., & Garfield, S. (Eds.). Handbook of psychotherapy and behaviorchange': An empirical analysis. New York: WUey, 1971.

Blau, P., & Duncan, O. The American occupational structure. New York:Wiley, 1967. .

Block, J. Advancing the psychology of personality: Paradigmatic sbm orimproving the quality of research? In D. Magnusson & N. Endler (Lds.),Personality at the crossroads. Hillsdale, N.J.: Erlbaum, 1977.

Block J. with Haan, N. Lives through time. Berkeley, Cal.: Bancroft, 197LBlumer, H. An appraisal of Thomas and Znaniecki's the Polish peasant in Europe

and America. New York: Social Science Research Council, 1 9 ^ .Bowers, K. Situationism in psychology: An analysis and a critique. Psycholog-

ical Review, 1973, 80, 307-336.Bromley, D. Personality description in ordinary language. New York: WUey,

1977Bronfenbrenner, U., & Mahoney, M. Influences on human development, 2nd ed.,

Hinsdale, 111.: Dryden, 1975.Buhler, C. Der menschUche lebenshuf ah psychologisches problem (The hu-

man course of life as a psychological problem). Leipzig: Hirael, 1933.Buhler, C , & Massarik, F. The course of human life. New York: Springer,

1968.Chein, I., Cerard, D., Lee, R., & Rosenfeld, E. The road to H: Narccrtics, de-

linquency, and social policy. New York: Basic, 1964.Clausen, J. Dmg addiction. In R. Merton & R. Nisbet (Eds.), Contemporary

social problems, 2nd ed. New York: Harcoxirt, Brace & World, 1966.Clausen, J. The life course of individuals. In M. Riley et al. (Eds.), Aging and

society. Vol. 3. New York: Russell Sage, 1972.Clausen, J. (Ed.). Socialization and socnety. Boston: Little, Brown, and Com-

pany, 1968.Cronbach, L. Beyond the two disciplines of scientific psychology. American

Psychologist, 1975, 30, 116-127.Dailey, C. The assessment of lives. San Francisco: Jossey-Brass, 1971.Denzin, N. The research act. Chicago: Aldine, 1970.DoUard, J. Criteria for the life history. New Haven: Yale, 1935.Duvall, H., Locke, B., & Brill, L. ,Follow-up study of narcotic drug addicts

five years after hospitalization. Public Health Reports, 1963, 78, 185-193.Ekehammar, B. Interactionism in personality from a historical perspective.

Psychologicd BuUetin, 1974, 81, 1026-1048.Elder, G. Children of the Great Depression. Chicago: U. of Chicago, 1974.Elder, C. Age differentiation and the life course. In A. Inkeles et al. (Eds.),

Annual Review of Sociology. Palo Alto: Annual Reviews, 1975.Endler, N., & Magnusson, D. (Eds.). Interactional psychology and personality.

Washington, D.C.: Hemisphere, 1976.Erikson, E. Young man Luther. New York: Norton, 1958.Feldman, H. Ideological supports to becoming and remaining a heroin addict.

Joumal of Health and Social Behavior, 1968, 9, 131-139.Gergen, K. Social psychology as history. Jourrud of Personality and Social

Psychology, 1973, 26, 309-320.Clad, B. Contributions of psychobiography. In J. Knutson (Ed.), Handbook

of political psychology. San Francisco: Jossey-Bass, 1973.Golding, S. Flies in the ointment: Methodological problems in the analysis of

the percentage of variance due to persons and situations. PsychologicalBulletin, 1975, 82, 278-288.

Goode, E. The major drugs of use among adolescents and young adults. InE. Harms (Ed.), Drugs and youth: The challenge of today. Ehnsford, N.Y.:Pergamon, 1973.

Gottschalk, L., Kluckhohn, C , & Angell, R. The use of personal documents inhistory, anthropology and sociology. New York: Social Science ResearchCouncil, 1945.

592 Runyan

Greenstein, F. PersonaUty and politics. Chicago: Markham, 1969.Haan, N. Coping and defending: Processes of sdf-environment organization.

New York: Academic, 1977.Hunt, G., & Odoroff, M. Follow-up study of narcotic drug addiction after hos-

pitalization. Public Health Reports, 1962, 77, 41-54.Kluckhohn, C. The personal document in antluropological science. In L. Gott-

schalk et al. (Eds.), The use of personal documents in history, atUhropology,and sociology. New York: Sodal Science Research Council, 1945.

Kurtines, W., Hogan, R., & Weiss, D. Personality dynamics of heroin use.Joumal of Abnornud Psychology, 1975, 84, 87-89.

Langness, L. The life history in anthropologtcd science. New York: Holt, Rine-hart & Winston, 1965.

Lazarus, R., & Launier, R. Stress-related transactions between person and en-vironment. In L. Pervin & M. Lewis (Eds.), Intemal and external determinantsof hehavior. New York: Plenum, in press.

Leighton, A., & Leighton, D. Gregorio The Hand-Trembler: A psychobiologicdstudy of a Navaho Indian. Papers of the Peabody Museum, Harvard Uni-versity, 40, 1949.

Levinson, D., et al. The seasons of a man's life. New York: Knopf, 1978.Lewin, K. Field theory in social science. New York: Harper Torchbooks, 1951.Lifton, R., & E. Olson (Eds.). Explorations in psychohistory. New York: Simon

and Schuster, 1974.Lindesmith, A. Opiate addiction. Bloomington, Ind.: Principia, 1947.Lowenthal, M., Thumher, M., & Chiriboga, D. Four stages of hfe. San Fran-

cisco: Jossey-Bass, 1975.Magnusson, D., & Endler, N. PersonaUty at the crossroads: Currera issues in

interactional psychology. Hillsdale, N.J.: Erlbavun, 1977.Mazlish, B. (Ed.). Psychoanalysis and history, rev. ed. New York: Grosset &

Dunlap, 1971.McAuliffe, W., & Gordon, R. A test of Iindesmith's theory of addiction: The

frequency of euphoria among long-term addicts. American Joumal of Sociol-ogy, 1974, 79, 795-840.

McGlothlin, W. Dmg use and abuse. Annual Review of Psychology, 1975, 26,45-64. *

Merton, R. Social theory and social structure. New York: Free Press, 1968 ed.Mischel, W. PersonaUty h- assessment. New York: Wiley, 1968.Mischel, W. Toward a cognitive social leaming reconceptualization of person-

ality. Psychological Review, 1973, 80, 252-283.Moos, R. The human context: Environmental determinants of behavior. New

York: Wiley-Interscience, 1976.Murray, H. et al. Explorations in personaUty. New York: Oxford, 1938.Murray, H. Introduction. In A. Burton & R. Harris (Eds.), Clinical studies of

personaUty, Vol. 1. New York: Haiper & Row, 1955.Nessekoade, J., & Baltes, P. Adolescent personality development and historical

change: 1970-1972. Monographs of the Society for Research in Child De-velopment, 1974, 39 (1 , Serial No. 154).

Olweus, D. A critical analysis of the "modem" interactionist position. In D.Magnusson & N. Endler (Eds.), PersonaUty at the crossroads. Hillsdale,N.J.: Erlbaum, 1977.

Overton, W., & Reese, H. Modeb of development: Methodological implications.In J. Nesselroade & H. Reese (Eds.), Life span developmental psychology:Methodological issues. New York: Academic, 1977.

Pervin, L. Some perspectives on the dichotomy between intemal and externaldeterminants of behavior. In L. Pervin & M. Lewis (Eds.), Intemal and ex-temal determinartts of behavior. New York: Plenum, in press.

Peterson, D. The cUnical study of social behavior. New York: Appleton-Century-Crofts, 1968.

The life course 593

Peterson, D. A functional approach to the study of i)erson-person interactions.In D. Magnusson & N. Endler, (Eds.), Personaltty at the crossroads. Hillsdale,N. J.: Erlbaum, 1977.

Pressey, S., & KiJJen, R. Psychologicd development through the life span.New York: Harper and BroAers, 1957.

Raush, H. Interaction sequences. Joumal of Persondity and Social Psychology,1965, 2, 487-499.

Raush, H. et al. Communication, conflict and marriage: Explorations in thetheory and study of intimate relationships. San Francisco: Jossey-Bass, 1974.

Reith, G., Crockett, D., & Craig, K. Personality characteristics in heroin addictsand non-addicted prisoners using the Edwards Personal Preference Schedule.Intemational Joumal of the Addictions, 1975, 10, 97-112.

Ricks, D., Thomas, A., & Roff, M. (Eds.). Life history research in psycho-pathology. Vol. 3. Minneapolis: U. of Minnesota, 1974.

Roff, M., & Ricks, D. (Eds.). Life history research in psychopathology. Vol. 1.Minneapolis: U. of Minnesota, 1970.

Runyan, W. Life histories: A field of inquiry and a framework for intervention.Ph.D. diss.. Harvard, 1975.

Runyan, W. How should treatment recommendations be made? Three studiesin the logical and empirical bases of clinical decision making. Joumal ofConsulting and Clinical Psychology, 1977, 45, 552-558.

Shaw, C. The jack-roller: A delinquent boy's own story. Chicago: U. ofChicago Press, 1930.

Sheehy, G. Passages: Predictable crises of adult life. New York: Dutton, 1976.Sheppard, C , Ricca, E., Fracchia, J., & Merlis, S. Personality characteristics of

urban and suburban heroin abusers: More data and another reply to Sutkerand Allain (1973). Psychological Reports, 1973, 33, 999-1008.

Shorter, E. The making of the modem family. New York: Basic Books, 1975.Stone, L. Prosopography. In F. Gilbert & S. Graubard (Eds.), Historicd

studies today. New York: Norton, 1972.Struening, E., & Guttentag, M. Handbook of evaluation research. Beverly HiUs,

Cal.: Sage, 1975.Sutker, P., & Allain, A. Incarcerated and street heroin addicts: A personality

comparison. Psychologicd Reports, 1973, S2, 243-246.Themstrom, S. The other Bostonians: Poverty and progress in the American

metropolis, 1880-1970. Cambridge: Harvard Univ. Press, 1973.Thomas, W., & Znaniecki, F. The Polish peasata in Europe and America. New

York: Knopf, 1920.VaiUant, G. Adaptation to life. Boston: Little, Brown, and Company, 1977.Vemon, P. Persondity assessment: A critical survey. New York: Wiley, 1964.Wachtel, P. Psychodynamics, behavior therapy, and the implacable experi-

menter: An inquiry into the consistency of personality. Joumal of AbnomidPsychology, 1973, 82, 323-334.

Wachtel, P. Interaction cycles, unconscious processes, and the person-situationissue. In D. Magnusson & N. Endler (Eds.), Personality at the crossroads,Hillsdale, N.J.: Erlbaum, 1977.

White, R. Lives in progress, 2nd ed. New York: Holt, Kinehart & Winston,1966.

Yinger, J. M. Toward a field theory of behavior. New York: McGraw-Hill,1965.

Manuscript received January 31, 1977.