Psychological Bulletin 1980, Vol. 88, No. I. 60-77

Work and Nonwork: A Review of Models, Methods, and Findings
Boris Kabanoff
School of Social Sciences, Flinders University of South Australia, Bedford Park, South Australia This article examines theory and research in the field of work/nonwork relations. Three different theories of work/leisure relations are examined—compensation, generalization, and segmentation. All three theories have received some support; however, the review indicates that much of the available research evidence is constrained by conceptual and methodological problems. A number of paradigms for describing work/leisure patterns are reviewed, and it is concluded that most of these paradigms are more suitable for classifying leisure definitions than for guiding empirical research. Research in this field requires objective definitions of attributes common across life spheres. A task-based description of work and nonwork is discussed, and its research utility is illustrated. It is recommended that future research be concerned with describing different work/leisure patterns, the processes underlying these patterns, and the life consequences associated with different patterns.

The meaning and the relationships between labor and leisure, work and contemplation, or in the most general sense, work and nonwork have been a source of intellectual, political, and religious debate for a considerable period in Western history. However, the view that the interplay between labor and leisure forms a major social and intellectual problem is a largely postindustrial phenomenon (Wilensky, I960). 1 Wilensky proposed that there are a number of social trends linked to industrialization that account for this development. Industrialization resulted
An earlier version of this was included in a report prepared for the Australian Social Welfare Commission, which funded the Labour and Social Activities project (O'Brien, G. E., Dowling, P., & Kabanoff, B. Work, health and leisure [Working Paper No. 28]. Bedford Park: Flinders University of South Australia, National Institute of Labour Studies, 1978). Principal investigators were Gordon O'Brien, Philip Bentley, and Barry Hughes. I would like to thank Gordon O'Brien and Leon Mann of Flinders University for their valuable comments on an earlier version of this article. Requests for reprints should be sent to Boris Kabanoff, School of Social Sciences, Flinders University of South Australia, Bedford Park, South Australia 5042.

in a major segregation of roles between the economic and noneconomic. Work became distinct spatially and, to some extent, socially from other role systems of kinship, religion, politics, and education. Thus the previously highly integrated system of personal roles became disrupted, and this separation resulted in a clearer recognition of the dualism of work and leisure. A number of factors linked with this dualist view have begun to attract the interest of psychologists (e.g., Dunnette, 1973; Porter, Lawler, & Hackman, 1975). Porter et al. suggested two major reasons for studying work and nonwork. First, they stated that in the past, work often has had the connotation of a chore and nonwork the connotation of play. They predicted that in the future this splitting of work and nonwork will decrease and that for many people work will increasingly take on the appearance of nonwork. One can infer from their view that to design work that has the engaging and stimulating
1 1 will use the terms work/leisure and work/nonwork interchangeably to avoid undue repetition, unless there is a distinction to be made between the two.

Copyright 1980 by the American Psychological Association, Inc. 0033-2909/80/8801-0060$00.75

60

WORK AND NONWORK.

61

qualities of nonwork, we may need to study the nature of leisure as well as of work. Also, although the future may see a new fusion between work and leisure, the present inferred separation of the two has some immediate implications. Since for some groups in society, leisure is probably more important to the general quality of life than is work (Dubin, 1956; Goldthorpe, 1968) and since there has been increasing concern with the quality of life (Campbell, 1976; Seashore & Taber, 1975; Spink, 1975), we may need to show greater concern with improving the quality of nonwork life. Furthermore, improving the quality of work life for those who are nonwork centered may result in no improvement in either life quality or organizational performance (Hulin & Blood, 1968; London, Crandall, & Seals, 1977), and this possibility has received little attention. A second reason for studying work and nonwork advanced by Porter et al. (1975) is the likelihood that leisure time will increase in the society of the future. As automation eliminates many jobs and as routine jobs tend to disappear, the balance between work and nonwork will change. Not only will leisure time increase, but the nature of work experiences will also change, and, as these authors pointed out, we know relatively little about either the social science of leisure or the relationship between work and leisure. A number of other authors have subscribed to the view that in the future, the time allotted to leisure will increase and the time allotted to work will decrease (e.g., Caldwell, 1974; Dumazedier, 1967; Ellul, 1964; Fourastie, 1965; Kahn & Weiner, 1967), yet at the same time it seems likely that work will retain a central role in the future (Emery, 1977; Emery, Emery, Caldwell, & Crombie, 1974; Entwistle, 1970; Parker, 1971). Such predictions emphasize the importance of studying work/leisure relationships. Current experiences with increased nonwork time due to the 4-day week have also highlighted the complex processes involved. Studies by Steele and Poor (1970), Nord and Costigan (1973), Goodale and Aagaard (1975) have shown that in general the reaction of employees to the 4-day week is

overwhelmingly favorable, with perceived favorable effects on home, work, and leisure. However, all of the studies identified specific problem areas associated with the shortened working week. For example, Nord and Costigan's longitudinal study showed that favorability toward the changed workweek decreased with time for a significant number of respondents, apparently because of certain negative effects on family life. Such a result seems to support the fears expressed by a number of authors that shortening work time leads not to an increase in ideal leisure but to an increase in maintenance time, that is, time absorbed by personal hygiene, transport, household duties, and care of children (Robinson, 1969; Szalai, 1972). Under (1970) suggested that more nonwork time does not result in greater personal growth but rather in shopping for more commodities, using and caring for these commodities, following advertising, or in Marcuse's (1964) terms pursuing the gratification of false needs. Thus leisure has emerged as a social psychological problem area. A common theme is a concern with the interaction between the work and nonwork spheres of human experience. Research relevant to this problem is examined in the next section. This review is confined to examining how work experiences may be related to nonwork experiences rather than how changes in working time may affect leisure. Work and Leisure: Theories and Empirical Findings Wilensky (1960) identified the beginning of two general streams of thought on the interaction of work and nonwork in the writings of two classic social philosophers— Engels (1892) 2 and de Tocqueville (1954).
Nothing is more terrible than being constrained do some one thing every day from morning night against one's will. And the more a man worker feels himself, the more hateful must work be to him because he feels the constraint,
2

to till the his the

An excellent critique of Engels's book can be found in Marcus (1974).

62

BORIS KABANOFF

aimlessness of it for himself. . . . In most branches the worker's activity is reduced to some paltry, purely mechanical manipulation repeated minute after minute, unchanged year after year. (Engels, 1892, pp. 118-119)

Engels continues:
Next to intemperance in enjoyment of intoxicating liquor, one of the principal faults of English working men is sexual license. But this too follows with relentless logic [from the fact] that the working men, in order to get something out of life, concentrate their whole energy upon these two enjoyments, carry them to excess, (p. 128)

Wilensky (1960) distinguished two major hypotheses within Engels's analysis: the compensatory leisure hypothesis and, by implication, the spillover hypothesis. The compensatory hypothesis represents explosive "letting off steam" and excessive attempts to make up for the deprivations experienced at work, whereas spillover posits a carry-over or generalization of alienation from work into alienation from leisure. Although Engels condemned the brutalizing influence of work, de Tocqueville fought what he saw to be the trivializing nature of leisure in the welfare state—its lack of challenge, its quiet servitude, and its dependence on orderly regulation. His work has been continued by a number of influential authors (e.g., Riesman, 1953), but it is with the first of these two streams that this review is concerned. Wilensky (1961) proposed that where the technical and social organization of work offers much freedom—for example, discretion in methods, pace or schedule, and opportunity for interaction with fellow workers . . . then work attachments will be strong, work integrated with the rest of life, and ties to community and society solid. Conversely, if the task offers little workplace freedom . . . then work attachments will be weak, work sharply split from leisure, and ties to community and society uncertain, (p. 522)

This proposal is clearly for a generalization from work to leisure; unfortunately, Wilensky provided no evidence about the nature of the work his sample actually performed but simply classified them according to occupation.

The enduring classic of the work/leisure field is Kornhauser's (196S) "Mental Health of the Industrial Worker." The major conclusion from this interview study of over 400 Detroit factory workers was that mental health varied consistently with the level of job held. The higher the job with respect to skill and associated attributes of variety, responsibility, and pay, the better the average mental health. He also concluded that routine work was associated with narrow and routine leisure activities with little evidence of self-development, self-expression, or interest in larger social purposes. Kornhauser interpreted these findings as indicating a generalization from work to leisure. However, there are a variety of conceptual and methodological problems in his study. These include the controversial nature of his mental health measure, the question of a middle-class bias in determining what is a healthy life orientation (Hulin & Blood, 1968), the ad hoc way in which leisure activities were classified as lacking in "selfdevelopment" and "self-expression," the absence of any statistic more sophisticated than correlation, and the use of a limited specific sample without adequate comparison groups from other occupational strata. Nevertheless, despite its possible inadequacies, the study remains one of the few attempts at a rigorous, empirical investigation of the work/nonwork problem. A number of more recent studies (Coburn, 1975; Gechman & Wiener, 197S; Kasl, 1973) have also shown moderate positive correlations between job satisfaction and mental health, although these studies did not provide information about actual leisure activities. Of course, it is also an open question as to whether job satisfaction causes good mental health or poor mental health results in job dissatisfaction. Meissner (1971) provided some data on work/nonwork relations. Having reviewed past research, Meissner suggested that previous studies did not really succeed in showing that work affects leisure because they attempted to do so by eliciting preferences, evaluations, and attitudes rather than by concentrating on the most obvious features of the work environment and how these influenced leisure activities. Meissner described

WORK AND NONWORK

the work place along three dimensions: the degree of technical constraint rather than personal discretion in work procedures (i.e., job autonomy), whether a worker had an instrumental orientation to the job or an expressive involvement in work, and the amount of social interaction permitted on the job. Leisure activities were treated in two ways. The major criterion of leisure activity was participation in voluntary organizations. Other leisure activities were divided into three main categories: activities that were considered to be organized and purposeful such as active sports or community activities, social activities such as visiting friends or drinking with companions, and expressive activities including fishing, watching television, and religious activities. This categorization was carried out along a presumed discretion continuum, with purposeful activity requiring high discretion and expressive activities low discretion. Meissner's (1971) results suggested that experience with work of little discretionary potential carried over into reduced participation in voluntary social organizations. Similarly, opportunities for social interaction on the job carried over into increased participation. Meissner (1971) concluded:
In my view, the findings . . . indicate that, when questions are put which are designed to find out what people actually do with their time, and if they have control over the space, time and function of their work, the job has a long arm indeed, (p. 260)

Nevertheless, there are a number of points in Meissner's (1971) study that call for cautious evaluation. Although Meissner concentrated on defining the actual nature of the work rather than worker attitudes about the job, he must have assumed that all workers who had similar jobs had identical reactions to the job to predict how work affects leisure. The role of individual work values in determining people's reactions to their work is still unclear (Stone, Mowday, & Porter, 1977; White & Ruh, 1973), but it is certainly desirable to measure individual reactions to or perceptions of the technical features of work. Another possible explanation is that Meissner's results demonstrate

not so much the effects of work on leisure as the fact that people are consistent in the sort of work and leisure they choose (Mansfield, 1972; Taylor, Kelso, Cox, Alloway, & Matthews, 1979). Furthermore, Meissner categorized leisure activities simply on the basis of his own judgment about their discretionary value. Leisure activities were classified into only three categories, and one of these categories included fishing, religion, and television. The rationale behind this classification is neither obvious nor compelling. Finally, although the findings are generally consistent, the strength of relationships is difficult to assess. The study reported leisure differences only in terms of average hours difference between participation rates in various types of activities or differences in the number of meetings or clubs attended. Yet, as Mansfield and Evans (197S) pointed out, Meissner's study is as sophisticated as or more sophisticated than most to date and is therefore instructive. Unfortunately, Mansfield and Evans's (197S) own study cannot lay any great claim to sophistication. They asked a sample of bank management and clerical personnel to describe the sort of satisfactions they got from their work and leisure, using five of Porter's (1961) need satisfaction dimensions: use of personal abilities, sense of accomplishment, prestige, interpersonal relations, and security. They concluded from their findings that groups of workers who experienced deprivation at work aspired to seek compensatory rewards outside work. Again, it is difficult to draw any firm conclusions from the results. First, as Mansfield and Evans (1975) pointed out, they showed only that employees aspired to compensate in their leisure, not that they actually did so. The results may tell us more about the implicit, compensatory theory of work and leisure held by employees than what actually occurred. Furthermore, their sample is suspect. They reported that the majority of their clerical respondents sought personal rewards mainly at work, which is inconsistent with a previous finding by Parker (1965), who found in a representative sample of British bank employees that only 14% of bank

64

BORIS KABANOFF

clerks considered work to be their central life interest. A recent study (Rousseau, 1978) followed Meissner in examining the effect of actual job attributes on nonwork, Rousseau had employees from two different organizations describe their work using a short form of Hackman and Oldham's (197S) Job Diagnostic Survey (JDS). The same employees then described their nonwork life using dimensions also derived from the JDS. Rousseau found that the work attributes and nonwork attributes were positively correlated. She suggested that this finding indicated a spillover rather than a compensatory relationship between work and nonwork. However, as Rousseau pointed out, a serious problem in the design of the study was the problem of correlated error, since work and nonwork measures were similar and were derived from the same questionnaire. A further problem was the ambiguity of the nonwork measures. Intuitively, it is difficult for respondents to meaningfully rate their nonwork on items such as "Projects that I start usually get finished; I feel I have a lot of control over my life." It is plausible that the nonwork measures simply tapped the degree of optimism respondents felt about life in general, a factor that might indeed be associated with their level of job satisfaction. Apart from this methodological problem, the study included only two fairly small organizations, Alienation Hypotheses A research stream related to the one being examined has investigated alienation as a possible mediating variable between work and nonwork activities. This research has not been concerned explicitly with work's effects on leisure; however, it is clearly relevant in view of its emphasis on the nonwork effects of work life. Commencing with Marx, the alienation literature includes a large number of contemporary theorists of whom Fromm (1968), Mills (1956), and Blauner (1964) are but a few. In the past few decades alienation has emerged as a preeminent explanatory tool in both sociology and psychology (Geyer, 1974;

Kanungo, 1979; Seeman, 1975; Brown, Note 1). Seeman (1971) identified three main "principles of generalization" that have been used to predict the effects of alienating work on nonwork: (a) the principle of frustration-aggression which states that alienated work builds frustrations that find release through ethnic hostility, punitive family relations, and so on; (b) the principle of social learning which states that alienated work teaches persons to behave in a passive, dependent, uninvolved fashion, hence its relationship to low political and social involvement; and (c) the principle of substitution which states that since alienated work fails to provide desired intrinsic satisfaction, people will seek to substitute nonwork, often shallow satisfactions such as status striving or conspicuous consumption. The latter two bear an obvious similarity to the traditional work/leisure hypotheses identified by Wilensky (1960). Yet, Seeman (1971) pointed out that several studies he carried out (Seeman, 1967; Seeman, Bishop, & Grigsby, 1971) did not provide any support for any of these hypotheses. In a Swedish study (Seeman, 1967) it was found that alienating work, defined by the degree to which work was uninvolving, did not correlate with ethnic hostility, status striving, voting behavior, or level of political knowledge. A similar result was obtained with an American sample (Seeman et al., 1971). Hagedorn and Labovitz (1968) also found that work alienation, which they defined as social isolation on the job, did not predict the level of participation in community organizations. However, Seeman (1971) did identify a theme in the alienation concept that seemed to have an explanatory and predictive usefulness not possessed by the alienation as low-work-involvement idea. This form of alienation he called powerlessness and identified it closely with what Rotter (1966) has termed an external locus of control (see, also, Fischer, 1976). In both of the studies discussed previously, it was found that a powerlessness index was related to measures of ethnic hostility and level of political knowledge. Seeman argued that the powerlessness concept may be useful for discriminating a specific sort of alienation rather than dealing

WORK AND NONWORK

65

with a "totality of alienations" that presumed to include a generalized malaise of low work involvement, powerlessness in politics, dislocation from common values, generalized distrust, and the like. A more recent study by Bacon (197S) examined the validity of what he termed three radical theories about the effects of alienating work on leisure; that alienated work generates false consumption obsessions and/or leads to passive uninvolved leisure pursuits (i.e., generalization) or results in violent, compensatory leisure behavior. His alienation scale consisted of four items that asked respondents whether their work was monotonous or uninteresting, constraining or oppressing, trivial or uninvolving, and demanding or frustrating. Bacon concluded from his results that there was no evidence for any of the three traditional hypotheses. Instead, he proposed that in most prosperous sectors of contemporary industrial society, a sharply segmented relationship has developed between the spheres of work and leisure. To a considerable extent the things that people choose to do in their free time are unrelated to the nature of their occupational experiences, (p. 179)

work, with simple work corresponding roughly to the notion of uninvolving work. That is, a sense of powerlessness may be a necessary intermediate step for the relationship to exist between uninvolving or alienating work and the types of alienated, nonwork behaviors discussed previously. Thus Bacon's finding is hardly surprising in view of Seeman's earlier results and does not necessarily invalidate the traditional hypotheses. The findings discussed previously also clearly indicate a need to explore the powerlessness concept as a possible mediating variable between work and nonwork behaviors and attitudes. Segmentalist Hypothesis Thus far the review seems to suggest that the belief of a number of writers (e.g., Emery, 1977; Gardell, 1976; Meissner, 1971) that work has a profound and generally deleterious effect on leisure is precisely that—a belief without substantial empirical support. Dubin (1956, 1958, 1973) is the main proponent of a school of thought which argues that this result is not the fault of imperfect methodology but that it reflects the axiom he offers: that social experience is inevitably segmented for most individuals in industrial society, with each social segment lived out more or less independently of the rest. That is, the worlds of work and leisure are essentially psychologically separate. Dubin's position revolves round five basic points: 1. the previously mentioned axiom of socially segmented experience; 2. the assumption that people's social participation may be necessary in one or more sectors of their social experience but may not be important to them; 3. the logical conclusion that adequate social behavior will occur in sectors of social experience that are mandatory but not valued by the individual;
3 For another review of this study, see Kelly (1976). * Despite this, a recent study (Kanungo, 1979) equates a lack of work involvement with alienation, which indicates that the controversy is not yet resolved.

This suggestion is explored in greater detail later. However, there is at least one crucial weakness in Bacon's (197S) study that questions the validity of his conclusions.3 Seeman (1971) appears to have convincingly shown that measures of alienation similar to Bacon's do not predict a number of nonwork attitudes and orientations. Bacon's study seems to confirm this. However, Seeman has further argued that perhaps work involvement is not the appropriate concept; rather, powerlessness and its effect on nonwork variables should be examined. Certainly Seeman has not accepted the segmentalist model as a result of his findings but rather has questioned the validity of some current concepts of work alienation.4 Several other studies (Shepard & Panko, 1974; Tudor, 1972) suggest that such a rejection of the traditional alienation model is premature. These studies have shown that there is an association, admittedly small, between people's sense of powerlessness and the simplicity of their

66

BORIS KABANOFF

4. the second proposition that in situations of necessary but unimportant social participation, the most direct and obvious features of the situation become bases for the individual's attachment to that situation; and 5. finally, that primary social relations take place only in situations in which the social experience is valued by the individual (Dubin, 19S6, p. 132). Dubin (1956) developed a Central Life Interest (CLI) questionnaire that measured expressed preference for a given locale or situation (job vs. nonjob) in carrying out an activity for four different areas of activity. These four areas were the formal aspects of membership and behavior in organizations, the technological aspects of the environment, the informal group life experiences, and general everyday experiences. In summary, he found 1. that three quarters of his sample of blue-collar workers could not be labeled as job oriented; 2. that the work place was not congenial to the development of preferred, informal human relationships; 3. that the work place was perceived, mainly, as the most important formal organization in workers' lives; and 4. that the work place was the focal point in our society for experience with technical environments. These results seem reasonable enough, but Dubin's interpretation of these findings is controversial. Dubin (1956) concluded that his results demonstrated that informal relationships on the job are unimportant in terms of both general social experience and as sources of attachment to the organization. Further, he proposed that the characteristics of industrial work that are alleged to be disturbing to the individual (monotony, repetitiveness, specialization,) are the very features that may be functional for society by etching for the individual some awareness of the division of labor and its resultant interdependence. Both these conclusions are startling, yet as Friedlander (1966) pointed out, Dubin's study does not seem to have fazed industrial sociologists or psychologists to any great ex-

tent. This relative lack of impact may be attributable to a number of causes. What Dubin measured was preference for work or nonwork as the locale for a given activity, thus explicitly requiring workers to differentiate between the two and hence emphasizing segmentation. His either/or approach also assumed that people can generally be categorized into two types—work centered or nonwork centered. This ignored those who might have a balanced orientation between work and nonwork (which, after all, is a socially desirable sort of response, Odaka, Note 2 ) , those who, in a sense, are alienated from both spheres, and excluded the possibility that people may give a variety of weightings to the relative importance of work and leisure. On this point, it is interesting to note that in more recent studies (Dubin & Champoux, 1974; Dubin & Champoux, Note 3) scoring of the CLI has been amended to allow a person to be classified as neither clearly work nor clearly nonwork centered. A study that has been interpreted as supporting the segmentalist position (London, Crandall, & Seals, 1977) also seems to have assumed the existence of a universal work/ leisure pattern. London, Crandall, and Seals measured the level of both work and nonwork satisfaction of a large American sample. They found that there was no correlation between work satisfaction and leisure satisfaction and concluded that this supported the segmentalist hypothesis. However, a number of alternative interpretations of this result are possible. It may be that work and leisure attitudes are not linearly related, but there may be some curvilinear relationship. For example, only persons who are highly satisfied, or perhaps highly dissatisfied may extend (Parker & Smith, 1976) their on-thejob feelings into the nonwork sphere. Or there may be persons who compensate for unsatisfying work, whereas others generalize unsatisfying work so that again there will be no overall linear relationship between work and leisure. Clearly, the simple absence of correlation between work and leisure attitudes is not strong evidence for segmentation. A more recent study (Orpen, 1978) disagrees with the London, Crandall, and Seals

WORK AND NONWORK

67

(1977) result. Using a cross-lagged correlation approach, Orpen compared the work and nonwork satisfaction of 73 front line managers over a 12-month period. He found that work and nonwork satisfaction were significantly related and also that work satisfaction had a larger influence on nonwork satisfaction than vice versa. Orpen's results are not in direct disagreement with Dubin's model, since it seems reasonable that many of the front line managers studied had work as their central life interest. However, his results do indicate that the segmentation model may not be appropriate for all members of the work force, as London, Crandall, and Seals imply. Several other studies have also disagreed with the Dubin hypothesis, at least in its most stringent form. Friedlander (1966) found that some job factors, such as amount of social interaction, were considered more important satisfiers of people's expressed life values than a number of off-the-job experiences. Harry (1971) proposed that social attitudes are not the only area in which work attitudes might influence nonwork attitudes. Harry concluded that although occupational type was found to be lacking in effect on the purely social content of leisure attitudes, there appeared to be some transfer of cultural content. By cultural content, Harry meant largely specific skill attributes or general technological attitudes. This suggests that Dubin's original suggestion of a fairly rigid segmentation is incorrect, although it must be pointed out that Dubin was concerned mainly with the transfer of social attitudes. A final feature of Dubin's position is that his findings could be interpreted as identifying a compensatory sort of process: his sample value off-the-job social relationships because these are restricted on the job. In later research (Dubin, Champoux, & Porter, 197S; Dubin, Champoux, & Stampfl, Note 4) Dubin has concentrated on examining how organizational commitment, job performance, and personality characteristics are related to an individual's central life interest and has not expanded his proposed segmentation between work and nonwork. Perhaps

for this reason the holists rather than the segmentalists (Parker, 1971) hold center stage in the work/leisure debate (Milton, 1975). In general, it seems that the idea of segmented spheres of experience is not acceptable in its strict form from either a scientific or commonsense point of view, although again it should be pointed out that Dubin was concerned mainly with one aspect of social experience—the formation of preferred informal relationships. Parker (197S) has argued that to attempt to compartmentalize leisure and work ignores some of the most important questions we need to ask, such as what is the relationship between behaviors and attitudes in leisure and in other spheres? What is the role of leisure in societies in which the nature of work is undergoing drastic changes? Overall, the holist assumption seems to be the appropriate one to make until it is comprehensively disproved. This review of the empirical evidence indicates that there is no unequivocal support for any of the three major work/leisure hypotheses. A number of reasons were advanced for this, dealing mainly with conceptual and methodological inadequacies in studies within this field. A central and recurring problem in most of the studies was a lack of consistency in the way that both work and leisure were defined. Work has been described in terms of occupation and by using various task attributes such as autonomy and interaction. In some cases, work has been described simply in terms of the personal needs it is seen as meeting. Definitions of leisure have been even more diverse. Leisure has been described by the frequency of participation in activities, impressionistically in terms of overall leisure quality or perceived psychological meaning to participants, and by classifying activities using broadly taskbased concepts. Definitions Work refers to the set of prescribed tasks that an individual performs while occupying a position in an organization. The organization is generally considered to be a work organization if an agreement is made to supply the individual with monetary rewards in

68

BORIS KABANOFF

return for his or her services to the organization. In general, work is a spatially, temporally, and, to an extent, socially discrete, well-defined role that we have little trouble in identifying. A task-based definition of work is the appropriate and common one. Leisure is also a common concept in our everyday language; however, in contrast to work, there is little precision in the way either the researcher or layman uses it. Dumazedier (1967) noted the essential ambiguity of the term when he described it as having elements of its etymological meaning in the concept of freedom of choice and elements of substantive meaning in its definition as residual time. Friedman (1961) said of leisure that it was a term that was "required to be used with all sorts of reservations" and that it "arouses mistrust" (p. 102), whereas Young and Willmott (1973) likened the attempt to define the concept to trying to grasp a jellyfish. Rapaport and Rapaport (1974) identified a number of different criteria that have been used for defining leisure. Perhaps the most widespread of these is the concept of leisure as free time. Time, for all its cultural and personal plasticity, at least offers a standardized objective basis for consensus. This approach has been used from the earliest "time budget" study by Lundberg, Komarovsky, and Mclnerny (1969) to a recent crossnational study by Szalai (1972), and, as Meyersohn (1969) pointed out, such studies of free time form the backbone of leisure research. Yet even free time concepts have their problems of usage. Allowing for various exclusions in what might be called the "time out" approach, such as overtime work, the journey to work, and essential functions and obligations, one is still left with the problem in deciding what is free time. For example, if a man helps with household chores do we define it as leisure only if it is not the role obligation of a married man? Of if a man spends his free time building a seaside cottage, do we define it as leisure only if he enjoys building and not if he's building it at the insistence of his family or as an investment? This then raises the point that if it is really the psychological

state that makes the difference—the feeling of pleasure, freedom, or relaxation—-then at least for some people dt is impossible to exclude their work from the leisure concept (Gerstl, 1961; Noe, 1971). This subjective approach to the definition of leisure ultimately makes it extremely difficult, if not impossible, to distinguish between work and leisure activities. Whether a particular activity is to be classified as work or leisure depends on the actor's mental state, and the same activity for a given individual may change its classification as often as the individual's feelings change. Because of the difficult measurement problem involved—how does one monitor individual feeling states? —the subjective approach does not seem to be useful. Another approach taken by a number of writers (e.g., Kelly, 1972; Parker, 196S) has defined leisure by its function in relation to work. In essence, these theorists treat leisure in what Wilensky (1961) labeled spillover or compensatory fashion; that is, leisure is either the image or the opposite of work. Recently, there has been growing criticism of this tendency to treat leisure as the subsidiary of work, a tendency that, as Cheek (1971) puts it, makes leisure the "cafeteria concept residual" of work. The implicit devaluation of nonwork as an autonomous activity sphere within this approach has been highlighted by the work of Dubin (1956), discussed previously. A major problem of this approach is that it allows many activities to be included as leisure activities that are not leisure activities in the common language usage of the term leisure. Are sleeping, cleaning one's shoes, washing up after dinner, and buying food to be classified as leisure? There appear to be a number of activities that are essential to basic physical maintenance that everyone is constrained to perform. These activities are performed mainly in a person's residential social setting and are best described as maintenance activities. They may be distinguished from leisure activities because of the element of constraint. Of course, there is an overlap or imprecision involved in the distinction. For example, people may eat to

WORK AND NONWORK

69

live or they may eat to enjoy themselves. However, the activities of the gourmet are not imposed. He or she chooses to spend more time and effort on the activity of eating than is necessary for the satisfaction of his or her basic physical appetite. Another approach to denning leisure has been to think of it as an ideal state that may or may not exist within a person's nonwork time. This Platonistic conception of leisure with its emphasis on leisure as a virtually unreachable or rarely attained state of consciousness (most of us seeing only dim reflections of this ideal state) is espoused by De Grazia (1964) and Pieper (1963) among others. This definition excludes from leisure all that is not chosen in full freedom, all that is shaped by work, and all that is largely a quality of body rather than spirit (Kelly, 1972). Neulinger (1974) described it thus:
To leisure means to be engaged in an activity performed for its own sake, to do something which gives one pleasure and satisfaction, which involves one to the very core of one's being. To leisure means to be oneself, to express one's talents, one's capacities, one's potentials, (p. xi)

This approach shares the problems of the subjective approach by denning leisure in terms of mental states and also shares the problems of the "residual work" approach by not being specific about the nature of the activities to be considered leisure activities. Any objective approach to understanding work and leisure must start from a definition of leisure that is reasonably specific and that allows the researcher to quantify dimensions of leisure activity independently of work activities. The most general approach, then, to a definition of leisure is that it is a set of activities that an individual performs. Leisure is a set of activities that individuals perform outside of their work context and excludes essential maintenance functions. These activities, although including some element of perceived choice, once chosen, involve the actor in performing a fairly clearly prescribed set of tasks or operations. In contrast to the work setting, leisure activities are primarily carried out in the pursuit of personally valued goals or in the expectation of fulfilling individual needs

rather than in return for monetary reward. Nevertheless, leisure activities may be work related if the actor perceives such activities to be personally meaningful, irrespective of any monetary gain directly contingent on these activities. There are six points that are explicit in this definition of leisure activities: (a) their nonwork context, (b) their status as nonmaintenance activities, (c) the element of choice involved, (d) the set of prescribed tasks or operations that comprise the activities, (e) the personal or individualistic source of motivation, and (f) the absence of monetary reward as a primary motivational factor. There is also an important element in the definition that is largely implicit. Since leisure activities are personally chosen and directed toward meeting individual rather than formally prescribed organizational goals, then it might be suggested that a leisure activity can have as many different meanings as there are participants (Kelly, 1972; Neulinger, 1974). To a degree this suggestion is valid, but it is arguable (Meyersohn, 1972; Zuzanek, 1974) that it would be a mistake to extend too far the idea that there is nothing meaningful in leisure activities per se and that because meaning is provided by the user, it can be derived equally from all activities. Since in the present definition leisure activities consist of a series of tasks, in the broad sense, chosen by the actor, these tasks impose some degree of constraint on the kinds of meanings that a person can attribute to his or her leisure activities. It is proposed therefore that it is possible to describe leisure using task dimensions that have previously been shown to have important psychological or behavioral consequences for the actor. Although it cannot be claimed that this sort of description provides an exhaustive or exclusive description of any given leisure activity, it will enable researchers to describe the profile of a person's leisure activities over a limited number of important task dimensions. Research that has attempted to classify leisure activities on the basis of the psychological needs that various activities are seen as gratifying (e.g., Bishop, 1970; London, Crandall, & Fitzgibbons, 1977; Ritchie, 1975) can be seen to have a similar goal.

70

BORIS KABANOFF

The task-based definition provides a strictly behavioral orientation to leisure as representing a set of activities that consist of the actor's task-defined operations. The nature of the leisure tasks the actor has chosen, to an important degree, determines the kinds of actions he or she carries out and the satisfactions he or she can experience. Since both work and leisure can be described by common task attributes, this definition should facilitate the investigation of work/ leisure relationships. Meissner's (1971) study provides perhaps the best example of this approach. However, it is necessary to expand on Meissner's approach in two major respects. First, it is desirable to use a theoretically based and larger variety of task attributes to describe work and leisure. Autonomy, interaction, and job involvement do not constitute an exhaustive list of task attributes (in fact, job involvement is not a real task attribute but more a work value) in terms of commonly used task descriptions (e.g., Hackman & Oldham, 197S; Barrett, Dambrot, & Smith, Note 5). Second, instead of arbitrarily assigning a leisure activity to one of a few crudely defined categories, it is preferable to objectively locate various activities along a continuum from low to high on a particular task attribute. This should allow a greater number of leisure activities to be described and should provide for a wider range of possible leisure patterns. Rousseau's (1978) study, although it represents an advance over Meissner in terms of the first problem, suffers even more than Meissner's from the weakness of its nonwork measures. The task-based approach seems to have a number of substantial advantages for dealing with the problem of how to relate work and leisure. By defining both work and leisure in terms of task attributes, it is possible to specify a limited number of task dimensions for describing work and leisure. By using a specific definition, it should be possible to use objective methods for measuring work and leisure attributes rather than impressionistic, purely qualitative descriptions. Also, work and leisure can be directly compared, since they are described along common dimensions. Finally, the approach appears to have con-

ceptual clarity. By describing work as a set of prescribed tasks and leisure as a set of chosen tasks, the assumed communality or continuity between the two spheres is clearly demonstrated. Furthermore, it is reasonable to assume that by defining a common task quality in both work and leisure, it is possible to explain the processes underlying work/leisure relationships. These four features of specificity of definition, amenability to objective measurement, direct relatability, and conceptual clarity are major requirements for an adequate model of work/leisure relationships, and the application of the taskattributes approach is demonstrated later in this review. However, in the next section a number of contemporary work/nonwork models are assessed in terms of these criteria. Work/Leisure Paradigms It has been said about the field of leisure research that
(1) there does not appear to be an overwhelming amount of solid leisure research in progress; (2) a substantial amount of effort is being devoted to the development of conceptual schema which appear to go untested; (3) descriptive rather than analytic studies are the vogue; and (4) indicating where future research is likely to be most fruitful is a task few contributors are prepared to face. (Johannis & Bull, 1971, pp. 246-247)

These conclusions are equally relevant to the field of work/leisure research, Parker's Tripartite Model Parker (1971) proposed a theory of work/ leisure interaction based on three sorts of relationships between work and nonwork: extension, neutrality, and opposition. He described these three patterns in the following way:
Extension consists of having leisure activities which are often similar in content to one's working activities and of making no sharp distinction between what is considered as work and what is leisure. With the opposition pattern leisure activities are deliberately unlike work and there is a sharp distinction between what is work and what is leisure. Finally, the neutrality pattern consists of having leisure activities which are generally different from

WORK AND NONWORK work but not deliberately so, and of appreciating the difference between work and leisure without always defining the one as the absence of the other, (pp. 101-102)

71

The crucial difference that Parker (1971) identified between these patterns is that extension and opposition denote, respectively, a positive and negative attachment to work, whereas people who show the neutrality pattern neither generalize the work they experience nor actively attempt to compensate for it. These descriptions are obviously similar to the classic generalization and compensation mechanisms. The model makes some interesting proposals, but the research on which the model relies has a number of shortcomings. Parker's (1971) sample consisted of people in business or service occupations, and although a bluecollar group was also included, their data appear to be largely unrepresentative (p. ISO). Parker also assumed that his results indicated an interaction between work experiences and leisure activities rather than reflecting occupational choice; that is, people may choose the same types or different types of work and leisure. Of course, the nature of the causal relation is difficult to unravel. However, some attempts must be made to do this at least by initially controlling for such factors as socioeconomic background, education, age, and so on (Kando & Summers, 1971). Parker's (1Q71) research method included asking respondents their reason for enjoying leisure. They were offered three alternative answers: "Because it satisfies interests that you would like to satisfy in your work," "because it is satisfying in a different way from work," or "because it is completely different from work." Parker concluded that "most respondents chose the second or third alternatives, possibly because the idea behind the first is not clear" (p. 83); one might add, not surprisingly, At this stage it is fair to say that the empirical basis for Parker's model is untested and unproved. In general, the model lacks any precise specification of what the three hypotheses mean in terms of behavior or activities. What does one mean by saying

leisure is "usually different" from work? How does one actually define differences or similarities between these two contexts without measuring common work/leisure attributes? Kelly's Two-Factor Model Kelly (1972) has provided a simple theoretical paradigm of work/leisure relationships. Within the leisure literature, he identified two main dimensions that theorists have used to conceptualize work/nonwork interaction. The first of these dimensions is labeled discretion, which Kelly defines as the extent to which leisure may be either freely chosen or determined by work constraints or the pervasive norms of the society. Freely chosen activity would be defined as being optional. There would be no significant penalty for not engaging in it. (p. 55)

The second dimension is called work relation:
Leisure may be independent of work or dependent on the meaning given it by work. The work relation includes not only the economic reward for tasks performed, but also the preparation, appearance, community relationships, residence and demeanour required or rewarded by the work position, (p. 55)

These two dimensions give a four-celled model of work and leisure. Briefly these cells can be described as follows: Cell 1: The activity is chosen and independent of work; it is an end in itself, is not work related in content or purpose, and is not an escape from work. Cell 2: The leisure activity is chosen and yet related to work, that is, leisure that is fully coordinated with work. Cell 3: This includes activity that is determined by structural or symbolic social factors and yet in form and content is independent of the work relation. Such leisure activity is complimentary to work by being associated with certain occupational expectations or by compensating for working conditions. Cell 4: This nonwork activity is socially determined and related to work in form and

72

BORIS KABANOFF

content. Such activity may be either preparation for work or recuperation from work. There are a number of fairly obvious shortcomings to Kelly's (1972) proposed model. The two dimensions he chose as major tend to make some large assumptions about the nature of leisure and the role it plays. Can such a complex sphere of activity be broken up into such a simplistic typology? Certainly, one would not expect to obtain such a clear picture if people were asked about their leisure activities; so it is necessary, as Kelly pointed out, to "concentrate on behavior . . . placing the activity in a context that includes relations to other participants, to family, and to work associations" (p. 54). This suggestion is an important one, but Kelly's own paradigm does not provide any insight into how this "placing in context" can or should be carried out. A general criticism of Kelly's (1972) model might be that in its quest for simplicity, it has tended to force leisure into a small set of narrow molds on the basis of two convenient but fairly crude dimensions. As Kelly suggested, theoretical types are not always easily distinguishable in field data, so perhaps it is of questionable use to draw up such simplistic typologies when their relation to actual behavior is peripheral and extremely difficult to use in any research setting. The model bears little fruitful relation to problems of dealing with people's behavior; rather, it simply categorizes leisure definitions (Kelly, 1972, p. 59). Neulinger's Three-Layer Model Neulinger (1974) has provided another leisure paradigm. Like Kelly's (1972) model and most other models, it included freedom as the primary determining characteristic of leisure. Unlike Kelly's model, Neulinger's did not include work as an independent variable, for he rejected the idea of treating leisure as the opposite of work. This may be partly true, but certainly leisure can be treated in a variety of ways, one of these being how it interlocks with work. Neulinger identified two further dimensions that he used to categorize leisure and nonleisure behavior: motivation for the activity (intrinsic or extrinsic) and

Table 1 Neulinger's Model
Motivation Goal Cells

Intrinsic Intrinsic and Perceived freedom Freedom (leisure) Final 1-2 Instrumental (pure leisure) Final 3-4 Instrumental (leisure-work) Final Instrumental 5-6 (leisure-job)

extrinsic Extrinsic

Intrinsic Intrinsic and extrinsic Extrinsic

Perceived freedom Constraint (nonleisure) Final 7-8 (pure work) Instrumental Final 9-10 Instrumental (work-leisure) Final Instrumental
11-12 (work-job)

Note. (From "The Psychology of Leisure" by J. Neulinger, Springfield, 111.: Charles C Thomas, 1974. Copyright 1974 by Charles C Thomas. Adapted by permission.)

goal of the activity (final or instrumental). Table 1 illustrates how these dimensions form a leisure paradigm. The nature of goal dimension is considered to be a secondary variable within the model, so that the paradigm consists of six behavior descriptive cells. For example, Cells 1 and 2 are described as "the purest form of leisure: an activity freely engaged in and done for its own sake; truly leisure in the classical sense" (Neulinger, 1974, p. 19). On the other hand, Cells 9 and 10 are "perhaps the broadest category of them all, behavior engaged in under constraint, but having both intrinsic and extrinsic rewards" (Neulinger, 1974, p. 20). Not surprisingly, there are no examples of this latter type of behavior. Like Kelly, Neulinger offers no empirical evidence as to the viability of his model. Once goal is treated as secondary, this model becomes similar to Kelly's, with two dimensions—freedom and motivation, motivation in Kelly's model being related to work and in Neulinger's model being an intrinsic-extrinsic distinction. Although Neulinger's model was drawn as a behavioral continuum, it clearly was treated by the au-

WORK AND NONWORK

73

thor as a successive series of cells with all the consequent problems identified in Kelly's model. There is a major, overriding criticism applicable to all three models discussed thus far, especially the latter two. As Kelly (1972) pointed out and as Neulinger (1974) would agree, both of the paradigms are an abstraction, and we are not likely to find any reallife situations that fit neatly into any category. All three of the models in some degree are directed not toward actual behavior but toward creating ideal type categories on an a priori conception of what nonwork and, for that matter, work behavior should be like. It is not the a priori specification of relationships that is the problem but rather that the models foster the reification of ideal concepts of leisure without connection to empirical evidence. None of the models offers any explanation of process—of how different motivations distinguish different cells, of how these variables change across cells, or of how perceptions, motivations, or goals interact across cells. None of these models seems to incorporate the fundamental point stressed earlier, that to study relationships between work and leisure, both spheres should be describable on some appropriate, common dimensions. To what extent can the adoption of this approach overcome some of the problems described previously? An Alternative Strategy A recent study (Kabanoff & O'Brien, in press) examined the work and leisure activities of a representative sample of the population of Adelaide, Australia. People described their jobs on five task attributes that were derived from an analysis of formal job structure (O'Brien, in press-a). These attributes were influence or autonomy, variety, skill utilization, pressure, and interaction. Respondents also indicated which leisure activities they regularly participated in. These activities were reliably rated by a group of independent raters on the same five task attributes. Each person's leisure profile was then described by his or her average score on the five attributes over all their nominated activities. A variety of other demographic,

attitudinal, and personality data were also obtained from respondents. The study had two major aims. First, to examine the validity of the three dominant work/leisure models, namely, compensation, generalization, and segmentation. The second aim was to examine which variables were associated with persons who had different work/leisure patterns. Only a small relationship was found between people's work and leisure attributes on two of the attributes— variety and skill utilization. People with greater variety and skill utilization at work also tended to have higher levels of these attributes in their leisure, which is consistent with the generalization hypothesis. However, the major conclusion to be drawn from the results was that work attributes on the whole have no strong or consistent relationship to leisure attributes. This result did not necessarily support the segmented work/leisure hypothesis. An alternative explanation of the results was that there is a variety of patterns of work/leisure behavior rather than a single pattern or no relationship. On the basis of previous research, four distinct work/leisure patterns or groupings were identified: (a) passive generalization, which reflects low levels of both the work and leisure attribute; (b) supplemental compensation (Kando & Summers, 1971), which reflects a low level of an attribute in work but a high level in leisure; (c) active generalization, which reflects high levels of an attribute in both work and leisure; and (d) reactive compensation (Kando & Summers, 1971), which reflects a high level of an attribute at work and a low level in leisure. The characteristics that differentiated among persons with these patterns were then examined by discriminant analysis. Those showing the passive generalization pattern were predominantly males with low education, low income, and low intrinsic work motivation but high extrinsic work motivation. This group seemed to represent the kind of respondents studied by both Kornhauser (196S) and Meissner (1971). Those comprising the supplemental compensation group tended on the other hand to be older females who were internally controlled, had a low income, and had a low extrinsic work motiva-

74

BORIS KABANOFF

tion. This group seemed to be representative of those who can be described as having nonwork, perhaps in the shape of family commitments, as their central life interest. However, included in this category were persons who conceivably demonstrated a true compensatory pattern. These persons were younger, worked shorter hours, had a relatively high intrinsic work motivation, and had leisure activities that were high on pressure and skill. Persons in the active generalization category had a high income, an intrinsic rather than an extrinsic work motivation, and were better educated. This group seems to represent an extension pattern as described by Parker ( 1 9 7 1 ) , in which a need for self-actualization is perhaps not fully met at work and thus extended into leisure. Those showing reactive compensation tended to be predominantly male, job centered, in terms of work's economic outcomes rather than its intrinsic features, and externally controlled. Since economic outcomes seemed to predominate for this group, it is conceivable that they represented the conspicuous consumers identified by a number of authors (Gorz, 196S; Mills, 1QS6). Thus, there are a number of different work/ leisure patterns, some of which conform to traditional compensation/generalization hypotheses and others of which do not. Many, perhaps most previous studies have made the error of overgeneralization, like the three blind men who tried to describe an elephant from the one part of the elephant they could touch—the trunk, the leg, and the tusk. Conclusions Future research, rather than concentrating on the development of ideal typologies or on proving the existence of a single, universal work/leisure pattern could be more usefully directed to examining how people actually balance needs, aspirations, and satisfactions across different life spheres. Rapaport and Rapaport (1974) put it this way: We need to examine "how individuals achieve a balance in their patterns of social participation among the different sectors of work, family and other interests" (p. 2 2 8 ) . To put it in less sociological terms, it should be our aim to describe

different work/leisure/family patterns, to discover the factors that determine these patterns, and to relate these patterns to other significant life outcomes such as general life satisfaction, work and leisure satisfaction, physical and mental health, and so on. This recommendation is consistent with the views of a number of authors (e.g., Seashore, 197S; Spink, 1975), who argue the necessity of expanding areas of psychology that have been concerned with studying behavior in one sphere, such as work, into multi-life-sphere studies. Furthermore, the task-attributes approach advocated and illustrated here seems to have the potential for providing an acceptable paradigm for the conduct of such multilife-sphere studies. O'Brien (in press-b) found that the leisure attributes described here significantly predicted people's level of satisfaction with their retirement, which further indicates the usefulness of the approach. This alternative approach to the work/leisure problem recognizes the essentially dynamic nature of experience rather than pursuing the static goal of classifying Platonic leisure concepts. It does not imply that leisure be treated as the inevitable subsidiary of work, as do traditional compensation/generalization notions, but rather that under certain circumstances, understanding a person's nonwork experiences can contribute to an understanding of his or her work behavior. For example, some research has investigated how different community features affect people's reactions to various aspects of their work (Hulin, 1969), thus the implied process is not unidirectional, although it can be traced in different directions depending on research aims. In future we should abandon oversimplified, unidirectional models of work and leisure that offer little or no account of the processes that underly the interactions among different life spheres. The time for creating ad hoc typologies is past, and the future must see a concern with deriving empirical evidence dealing with rather than ignoring the undoubtedly complex interchange between the individual's work and nonwork life. Reference Notes
1. Brown, W. Transforming the miraculous into the natural (Working Paper Series No. 26). Unpub-

WORK AND NONWORK lished manuscript, Flinders University of South Australia, Bedford Park, South Australia, National Institute of Labour Studies, 1978. Odaka, K. Work and leisure: As viewed by Japanese industrial workers. Paper presented at the Sixth World Congress of Sociology, fivian, France, September 1966. Dubin, R., & Champoux, J. E. Worker's central life interests and personality characteristics (Office of Naval Research Project Tech. Rep, No. 20). Irvine: University of California, 1973. Dubin, R., Champoux, J. E., & Stampfl, J. Job satisfaction as related to central life interest (Office of Naval Research Project Tech. Rep. No. 17). Irvine: University of California, 1973. Barrett, G. V., Dambrot, F. H., & Smith, G. R. The relationship between individual attributes and job design: Review and annotated bibliography (Tech. Rep. No. 6). Akron, Ohio: University of Akron, Department of Psychology, 1975.

75

2.

3.

4.

5.

References
Bacon, A. W. Leisure and the alienated worker: A critical reassessment of three radical theories of work and leisure. Journal of Leisure Research, 1975, 7, 179-190. Bishop, D. W. Stability of the factor structure of leisure behavior: Analyses of four communities. Journal of Leisure Research, 1970, 2, 160-170. Blauner, R. Alienation and freedom: The factory worker and his industry. Chicago: University of Chicago Press, 1964. Caldwell, G. T. The future of leisure. Journal of the Institute of Personnel Management, 1974, 12, 2030. Campbell, A. Subjective measures of well-being. American Psychologist, 1976, 31, 117-124. Cheek, N. H., Jr. Toward a sociology of not work. Pacific Sociological Review, 1971, 14, 245-2S8. Coburn, D. Job-worker incongruence: Consequences for health. Journal of Health and Social Behavior, 1975, 2, 198-212. De Grazia, S, Of time, work and leisure. Garden City, N.Y.: Doubleday, 1964. Dubin, R. Industrial workers' worlds: A study in the central life interests of industrial workers. Social Problems, 1956, 4, 131-142. Dubin, R. The world of work: Industrial society and human relations. Englewood Cliffs, N.J.: PrenticeHall, 1958. Dubin, R. Work and non-work: Institutional perspectives. In M. D. Dunnette (Ed.), Work and non-work in the year 2001. Monterey, Calif.: Brooks/Cole, 1973. Dubin, R., & Champoux, J. E. Workers' central life interests and job performance. Sociology of Work and Occupations, 1974, 1, 313-326. Dubin, R., Champoux, J. E., & Porter, L. W. Central life interests and organizational commitment of blue-collar and clerical workers. Administrative Sci-

ence Quarterly, 1975, 20, 411-421. Dumazedier, J. Toward a society of leisure. New York: Free Press, 1967. Dunnette, M. D. (Ed.). Work and non-work in the year 2001. Monterey, Calif.: Brooks/Cole, 1973. Ellul, J. The technological society. New York: Vintage Books, 1964. Emery, F. E. Futures we are in. The Hague, Netherlands: Nijhoff, 1977. Emery, F. E., Emery, M., Caldwell, G. T., & Crombie, A. Futures we're in. Canberra, Australian Capital Territory: Australian National University, Centre for Continuing Education, 1974. Engels, F. The condition of the working class in England in 1844 (F. K. Wischnewetsky, trans.). London: Allen & Unwin, 1892. Entwistle, H. Education, work and leisure. London: Routledge & Kegan Paul, 1970, Fischer, C. S. Alienation: Trying to bridge the chasm. British Journal of Sociology, 1976, 27, 3549. Fourastie, J. Les 40.000 heurs. Paris: Gonthier, 1965. Friedlander, F. Importance of work versus nonwork among socially and occupationally stratified groups. Journal of Applied Psychology, 1966, 50, 437-441. Friedman, G. The anatomy of work: Labor, leisure and the implications of automation. New York: Free Press, 1961. Fromm, E. The revolution of hope: Toward a humanized technology. New York: Bantam Books, 1968. Gardell, B. Relations at work and their influence on non-work activities: An analysis of a sociopolitical problem in affluent societies. Human Relations, 1976, 29, 885-904. Gechman, A. S., & Wiener, Y. Job involvement and satisfaction as related to mental health and personal time devoted to work. Journal of Applied Psychology, 1975, 60, 521-523. GerstI, J. E. Leisure, taste and occupational milieu. Social Problems, 1961, 0, 56-68. Geyer, F. Alienation and general systems theory. Sociologies Neerlandica, 1974, 10, 18-41. Goldthorpe, J. H. The affluent worker. Cambridge, England: Cambridge University Press, 1968. Goodale, J. G., & Aagaard, A. K. Factors relating to varying reactions to the 4-day workweek. Journal of Applied Psychology, 1975, 60, 33-38. Gorz, A. Work and consumption. In P. Anderson & R. Blackburn (Eds.), Towards socialism. London: Fontana, 1965. Hackman, J. R., & Oldham, G. Development of the Job Diagnostic Survey, Journal of Applied Psychology, 1975, 60, 159-170. Hagedorn, R., & Labovitz, S. Participation in community associations by occupation: A test of three theories, American Sociological Review, 1968, 33, 272-283. Harry, J. Work and leisure-situational attitudes. Pacific Sociological Review, 1971, 14, 301-309. Hulin, C. L. Sources of variation in job and life satisfaction: The role of community and jobrelated variables. Journal of Applied Psychology, 1969, S3, 279-291.

76

BORIS KABANOFF Neulinger, J. The psychology of leisure. Springfield, III.: Charles C Thomas, 1974. Noe, F. P. Autonomous spheres of leisure activity for the industrial executive and the blue collarite. Journal of Leisure Research, 1971, 4, 220-249. Nord, W. R., & Costigan, R. Worker adjustment to the four-day week: A longitudinal study. Journal of Applied Psychology, 1973, 58, 60-66. O'Brien, G. E. The centrality of skill-utilization for job design. In K. Duncan, M. Gruenwald, & D. Wallis (Eds.), Changes in working life. New York: Wiley, in press, (a) O'Brien, G. E. Leisure attributes and retirement satisfaction. Journal of Applied Psychology, in press, (b) Orpen, C. Work and nonwork satisfaction: A causalcorrelational analysis. Journal of Applied Psychology, 1978, 63, 530-532. Parker, S. R. Work and non-work in three occupations. Sociological Review, 1965, 13, 65-75. Parker, S. R. The future oj work and leisure. London: MacGibbon & Kee, 1971. Parker, S. R. The sociology of leisure: Progress and problems. British Journal oj Sociology, 1975, 26, 91-101. Parker, S. R., & Smith, M. A. Work and leisure. In R. Dubin (Ed.), Handbook of work, organization and society. Chicago: Rand McNally, 1976, Pieper, J. Leisure, the basis of culture. New York: New American Library, 1963. Porter, L. W. A study of perceived need satisfactions in bottom and middle management. Journal of Applied Psychology, 1961, 45, 1-10. Porter, L. W., Lawler, E. E., & Hackman, J. R. Behavior in organizations. New York: McGrawHill, 1975. Rapaport, R., & Rapaport, R. N. Four themes in the sociology of leisure. British Journal oj Sociology, 1974, 25, 215-229. Riesman, D. The lonely crowd. New York: Doubleday, 1953. Ritchie, J. R. B. On the derivation of leisure activity types—A perceptual mapping approach. Journal of Leisure Research, 1975, 7, 128-140. Robinson, J. P. Social change as measured by time budgets. Journal of Leisure Research, 1969, 1, 7577. Rotter, J. B. Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 1966, 80(1, Whole No. 609). Rousseau, D. M. Relationship of work to nonwork. Journal of Applied Psychology, 1978, 63, 513-517. Seashore, S. E. Defining and measuring the quality of working life. In L. E. Davis & A. B. Cherns (Eds.), The quality oj working life: Vol. 1. Problems, prospects and state of the art. New York: Free Press, 1975. Seashore, S. E., & Taber, T. D. Job satisfaction indicators and their correlates. American Behavioral Scientist, 1975, IS, 333-368. Seeman, M. On the personal consequences of alienation in work. American Sociological Review, 1967, 32, 273-285. Seeman, M. The urban alienations: Some dubious

Hulin, C. L., & Blood, M. R. Job enlargement, individual differences and worker responses. Psychological Bulletin, 1968, 69, 41-55. Johannis, T. B., Jr., & Bull, C. Foreword. Pacific Sociological Review, 1971, 14, 243-244, Kahn, H., & Weiner, A. The year 2000: A framework for speculation on the next thirty-three years. London: Macmillan, 1967. Kabanoff, B., & O'Brien, G. E. Work and leisure: A task attributes analysis. Journal of Applied Psychology, in press. Kando, T., & Summers, W. The impact of work on leisure. Pacific Sociological Review, 1971, 14, 310327. Kanungo, R. N. The concepts of alienation and involvement revisited. Psychological Bulletin, 1979, 86, 119-138. Kasl, S. V. Mental health and work environment: An examination of the evidence. Journal oj Occupational Medicine, 1973, 15, S09-S18. Kelly, J. R. Work and leisure: A simplified paradigm. Journal oj Leisure Research, 1972, 4, 50-62. Kelly, J. R. A critique of "Leisure and the alienated worker: A critical reassessment of three radical theories," by Bacon. Journal of Leisure Research, 1976, 8, 129-131. Kornhauser, A. W. Menial health oj the industrial worker. New York: Wiley, 1965. Linder, S. B. The harried leisure class. New York: Columbia University Press, 1970. London, M., Crandall, R., & Fitzgibbons, D. The psychological structure of leisure: Activities, needs, people. Journal of Leisure Research, 1977, 9, 252263. London, M., Crandall, R., & Seals, G. W. The contribution of job and leisure satisfaction to the quality of life. Journal of Applied Psychology, 1977, 62, 328-334. Lundberg, G. A., Komarovsky, M., & Mclnerny, M. A. Leisure: A suburban study. New York: Agathon, 1969. Mansfield, R. Need satisfaction and need importance in and out of work. Studies in Personnel Psychology, 1972, 14, 21-27. Mansfield, R., & Evans, M. G. Work and non-work in two occupational groups. Industrial Relations, 1975, 6, 48-54. Marcus, S. Engels, Manchester and the working class. New York: Vintage Books, 1974. Marcuse, H. One dimensional man. London: Routledge & Kegan Paul, 1964. Meissner, M. The long arm of the job: A study of work and leisure. Industrial Relations, 1971, 10, 239-260. Meyersohn, R. The sociology of leisure in the United States: Introduction and bibliography, 1945-1965. Journal oj Leisure Research, 1969, 1, 53-68. Meyersohn, R. Leisure. In A. Campbell & P. Converse (Eds.), The human meaning oj social change. New York: Russel Sage Foundation, 1972. Mills, C. W. White cottar: The American middle classes. New York: Oxford University Press, 1956. Milton, P. Work and leisure: What is the link? Work and People, 1975, Spring, 27-31.

WORK AND NONWORK theses from Marx to Marcuse. Journal of Personality and Social Psychology, 1971, 19, 135-143. Sceman, M. Alienation studies. In A. Inkeles (Ed.), Annual review of sociology (Vol. 1). Palo Alto, Calif.: Annual Reviews, 1975. Seeman, M., Bishop, J. M., & Grigsby, J. E. Community and control in a metropolitan setting. In P. Orleans & W. R. Ellis (Eds.), Race, change and urban society. Beverly Hills, Calif.: Sage, 1971. Shepard, M. J., & Panko, T. R. Alienation and social referents. Sociology and Social Research, 1974, 59, 55-60. Spink, P. Some comments on the quality of working life. Journal of Occupational Psychology, 1975, 48, 179-184. Steele, J. L., & Poor, R. Work and leisure: The reactions of people at 4-day firms. In R. Poor (Ed.), 4 days, 40 hours. Cambridge, Mass.: Burk & Poor, 1Q70 . Stone, E. F., Mowday, R. T., & Porter, C. W. Higher order need strengths as moderators of the job scope-job satisfaction relationship. Journal of Applied Psychology, 1977, 62, 466-471. Szalai, A. The use of time—Daily activities of urban and suburban populations in twelve countries. The Hague, The Netherlands: Mouton, 1972.

77

Taylor, K. F., Kelso, G. I., Cox, G. N., Alloway, W. J., & Matthews, J. P. Applying Holland's vocational categories to leisure activities. Journal of Occupational Psychology, 1979, 52, 199-207. Tocqueville, A. de. Democracy in America. New York: Vintage Books, 1954. Tudor, B. A specification of relationships between job complexity and powerlessness. American Sociological Review, 1972, 37, 596-604. White, J. K., & Ruh, R. A. Effects of personal values on the relationship between participation and job attitudes. Administrative Science Quarterly, 1973, IB, 506-514. Wilensky, H. L. Work, careers and social integration. International Social Science Journal, 1960, 12, 543-560. Wilensky, H. L. Orderly careers and social participation. American Sociological Review, 1961, 24, 522-529. Young, M., & Willmott, P. The symmetrical family. London: Routledge & Kegan Paul, 1973. Zuzanek, J. Society of leisure or the harried leisure class? Leisure trends in industrial societies. Journal of Leisure Research, 1974, 6, 293-304. Received April S, 1979 •