2
The Scholarly Context of the Study

As an ethnography of an age-homogeneous community in an East Asian culture, this study of Fuji-no-Sato addresses three principal bodies of literature in the West: Asian studies, cultural gerontology, and social change. In the field of Asian studies, it takes its place among many other community ethnographies, each of which reveals a facet of the evolving contemporary culture of Japan. As contributors to this literature, we have sought to clarify some things about the role of the elderly in Japanese society and about the changing meaning of concepts like "community" and "welfare" in the Japanese cultural idiom.

Much has been written about whether Japan serves as a counterexample to Western societies in which age is disparaged (Kiefer, 1990; Palmore, 1975; Palmore and Maeda, 1985; Plath, 1972). The question is a subtle one, and the example of Fuji-no-Sato is instructive on many counts. For one thing, it demonstrates that there are important subgroups of the aged in Japanese society. The residents are upper-middle class in education and income. Perhaps partly for this reason, many of them are either childless or unusually distant from their children and seem to identify more with their achieved occupations and statuses (teacher, executive, public official, minister) than with their more traditional ties (home, neighborhood, friends, extended kin). Our study underlines the relative lack of scholarship on the social implications of such social distinctions among the Japanese aged.

This study also draws attention to the fact that simple comparisons cannot be made between Western and Japanese indicators of status and prestige, at least as applied to the elderly. Public and private assistance programs in advanced societies develop out of native ideas about the rights and obligations of the interest groups involved—government (both local and national), neighborhood, the aged, families, industry. These groups themselves are defined

by custom. What is a "neighborhood," for example, who is a "member," and under what circumstances does a member have "rights" to services shared by other members? Simply showing the distribution of services by age group bypasses such questions and often misses the point. It is the strength of ethnography that it brings such problems of definition into focus.

Much has been written about the group centeredness of Japanese life (DeVos, 1973b; Nakane, 1972). It is now clear that "neighborhood" is no more (or not much more) synonymous with "community" in Japan than it is in any other society where people are as diverse, mobile, and career oriented as the Japanese. Fuji-no-Sato is an example of this fact. Still, the memory (real or imagined) of a less individualistic and alienated society seems to be relatively fresh in the collective Japanese mind. Although some of the residents of Fuji-no-Sato seemed comfortable enough with their privacy and anonymity, many clearly felt that something was missing in neighborhood relations. Their image of community may not have been a realistic one, but it was different from the image held by typical Western elders. The study of attitudes toward community in Fuji-no-Sato provides a close-up of the image that word conjures in the minds of the residents and the planners. As such, it helps clarify what "group centeredness" means to the Japanese themselves, at the same time noting that such images are not uniform or static. The dynamic study of the strain toward community formation should help us understand the unfolding of new definitions and expectations made possible by the untested and unstructured society created in places like Fuji-no-Sato.

Using words and concepts that have become familiar in the social science literature on a particular culture, like Japan, although unavoidable, often creates the unwelcome impression that the writer has simply absorbed cliches about his subject rather than search for fresh ideas. But cliches often express important realities succinctly, and sometimes their use in a new context can even create fresh insights. Most of our use of culture-specific words familiar to the Japan scholar (amae , filial piety, etc.) are taken directly from the speech of the residents themselves and therefore convey something about their attempt to express their own reality (cf. Glaser and Strauss, 1969).

An Age-Homogeneous Community

In contributing to cultural gerontology (as we have chosen to call the ethnological study of the aged), we have an agenda different from the usual. We find a growing number of ethnographies, from a variety of countries and settings, of age-homogeneous living arrangements for old people. This literature has grown out of the mid-twentieth-century Western preoccupation with the aged as a "problem" population; as such, it tends to focus on how society provides, or fails to provide, its various over-sixty populations with a safe, comfortable, and agreeable life. These are appropriate questions, given Western values, and we shall examine Fuji-no-Sato as an example of how the answers are affected by culture.

In this vein, Fuji-no-Sato appears particularly relevant to two important issues that, along with the issue of the relative status and prestige of the elderly, provide the thematic structure to much of cultural gerontology: (1) How can society meet equitably the special economic and health care needs of an increasingly frail and post-productive population? Here, equity is usually understood to include not only equality of access to services like housing and medical help, but also the preservation of values like autonomy and self-esteem that make life worth living. (2) Are the elderly "isolated" from meaningful participation in society, or are they "integrated"? If the latter, how is this integration achieved?

Fuji-no-Sato clearly represents, among other things, an attempt by a Japanese institution to address the first question. We could scarcely tell the story of the community without saying something about the successes and failures of that attempt, and there is much the West can learn from this example. The architectural strategies of the community are a case in point, illustrating some fairly successful solutions to cost and quality problems (for example, the provision of outdoor space for gardening and the use of integrated telephone and safety systems), and some rather unsuccessful ones (the plumbing design and the first attempt to provide nursing home care).

Still, a detailed analysis of the cost and quality issues is beyond the scope of this study. Some attention is paid to the place of this

community within the Japanese system of geriatric health and welfare, but it is an enormously complex system, nested within an even more complex political and economic matrix, with many features that require careful explication for a Western audience. We have thus chosen to focus this study on the social description of Fuji-no-Sato at the expense of its economic and political implications, and we have more to say about the question of integration/isolation than the question of technical success or failure.

A Changing Community

All the industrial societies, including Japan, have absorbed vast changes in the living habits and values of their members and will continue to do so. It is obviously a stressful process that often produces political unrest and individual pathology; but it is as much a product of industrial technology as is material abundance or population growth. We mentioned that Fuji-no-Sato is not only an example of a kind of place (an age-homogeneous community in an East Asian culture), but also an example of the process by which new forms of social life are created and take their place in a complex, rapidly changing culture.

We are not interested in the broad question of major historical change—how it happens, its inevitable stages, or its distorting effects on social organization. Rather, we are interested in the micro-process of how a particular solution (the age-homogeneous retirement community) to a socioeconomic problem results in a particular set of interpersonal relations. We have tried to avoid imposing on Fuji-no-Sato our own value judgments about what community relationships ought to look like and to be sensitive to the ideals of the actors themselves—the planners, the administrators, and the residents.

It is not surprising, then, that we have found the most useful concepts to be those that have grown out of the close study of small interpersonal situations: the concepts of symbolic interactionism. In the study that follows, we repeatedly refer to the various actors' definitions of situations , and we seek to show how these definitions (1) differ among the various reference groups involved, (2) are selected from the range of possible meanings offered by the macro-culture, and (3) produce dynamic results (conflict, cooperation, alienation, and social bonding). Understanding larger scale histor-

ical processes depends to some extent on the careful study of such microprocesses.

Social Integration as Seen From the West

One is unlikely to find a coherent perception of "old people" throughout American society, much less throughout the wider Western world. Individual perceptions tend to be colored by personal experience, which varies according to the age composition of one's family and community, the nature and extent of one's contact with older people, and so on. However, the profession of social gerontologist does expose the professional to a more or less coherent body of observation, with an implicit set of views. Growing as it did from the needs of health and social welfare professionals in Western industrial nations (Quadagno, 1982), gerontology tends to view the aged in terms of the problems they pose for public policy in such nations.

Many of the caretaking functions earlier performed by families and communities have been taken over by public institutions in Western societies. One result has been a growing professional awareness of and sensitivity to the dissatisfactions of the old. Another probably has been a growing sense of loneliness and insecurity on the part of many elderly, who perceive their great dependence on impersonal institutions and unpredictable regulations. Clients' feelings of isolation and insecurity have themselves become a major problem for the geriatric social service provider; and this problem has generated a lively literature on the social integration of the aged—the conditions that inhibit or promote it and its effect on health and well-being.

In such an atmosphere, Western social gerontologists are often surprised to find communities of old people who appear to have a full and rewarding social life somewhat independently of the ministrations of family, social worker, or activities director. Writing about her experience in such a community in the United States, Hochschild (1973: xiv) says, "One reason I have written this book is that these forty-three people were not isolated or lonely. They are part of a community I did not expect to find." Ross (1977:xi), re-

porting on a retirement home in Paris, concludes, "Probably the most surprising finding of my study was that, except for their being old, little about these old people was surprising. Their friendships, fights, and love affairs, their in-jokes and strategies for coping with common problems are only striking because we don't expect old people to go on living like everyone else."

This issue of social integration pervades all descriptive studies in social gerontology, and it has been relatively well treated in theory development. The beginning gerontologist will be familiar with Cumming and Henry's social psychological "disengagement theory" (Cumming and Henry, 1961), which suggests that a certain pattern of social disintegration is natural and often beneficial in old age. The disengagement model provides one of the few examples of an empirically studyable theory in social gerontology and has therefore attracted a great deal of attention, albeit mostly negative (see Maddox, 1964).

We make no attempt to test social psychological models like dis-engagement theory, but we are keenly aware of the bearing of our work on the sociological problem of integration. Among Western sociologists, Irving Rosow has given us the most comprehensive discussion of this problem, and we follow the outline of his argument in exploring the issue in the context of Fuji-no-Sato.

Having described a type of integrated elderly community in his landmark study, Social Integration of the Aged (1967), Rosow turned to the theoretical question of the conditions of social integration in Socialization to Old Age (1974). Taking the viewpoint of the individual rather than the social system, he argued that integration can be understood through the use of three common sociological concepts: social values, social roles, and group membership. Each concept generates a set of independent factors, and together these "provide the ties that bind social norms into institutions, structure social intercourse, place a person in society and order his relation to others. Thereby, they provide the means and substance of integration" (Rosow, 1974:29).

Values, in other words, can be seen as the affective potentials that are embodied in institutional goals and rules and that in turn legitimize institutional arrangements. Roles provide the sets of mutual expectations that give predictability and worth to structured interpersonal relationships, and group memberships help deter-

mine one's role partners and select a certain repertoire of roles from the available cultural inventory.

As individuals learn values and role behaviors and acquire group memberships through socialization, "socialization becomes one major mechanism of integration" (Rosow, 1974:28). Turning his attention to the socialization of the aged in the United States, Rosow concludes that our society does not provide viable roles or norms for its aged members and that no consistent, effective socialization to age-appropriate values, roles, and memberships takes place. Rather, youth-oriented age norms assure that most elderly remain alienated from mainstream American society.

Assuming that socialization to age-appropriate roles is desirable and noting that age-homogeneous communities often seem to provide the possibility of such socialization, Rosow proposes the promotion of semisegregated communities of elderly to deal with the problem. If the aged are to generate new norms and roles that will allow adequate socialization—albeit within a circumscribed community—Rosow notes that two conditions are especially favorable: "first, when the [aged] members are also socially homogeneous on factors other than age , notably social class, race, ethnicity, and marital status. . . in other words, the factors that normally govern voluntary social groupings and spontaneous association. . . . The second condition favorable to such groups is large concentration of similar elderly" (1974:160-161, original emphasis).

The conditions Rosow describes, and the outcome of strong social bonding and high interaction rates, have been found in a variety of studies of age-concentrated living situations, including a small apartment building in California (Hoschschild, 1973), a mobile home park in California (Johnson, 1971), a large retirement community in Missouri (Perkinson, 1980), a neighborhood Jewish "network" in California (Myerhoff, 1978), and a retirement home in France (Ross, 1977). The French ethnography is particularly relevant here, both because it was a study by a cultural anthropologist of social integration outside the United States and because it focused on the internal social organization of the community.

Unlike the typical ethnography, with its innocence of theoretical perspective, Ross's study was designed to test a specific set of hypotheses about internal social integration or "community creation," a set of concepts derived from a review of the literature. As mea-

sures of community formation, she chose (1) physical distinctness, (2) "we" feeling (a "sense of distinctiveness, or shared fate, of things in common, in short, a feeling that 'we' is the right word to describe a collectivity of individuals") (Ross, 1977:3), and (3) social organization (a term she left undefined). She hypothesized that these factors would be strengthened by seven "background factors" describing the residents before they entered the community—homogeneity, lack of alternative, financial investment (amount and irreversibility), material distinctness, social exclusivity, leadership, and size—and six "emergent factors"—participation in communitywide events and decision making, proportions of certain kinds of contact among residents, interdependence, unpaid communal work, threat from outside, and the formation of community-specific symbols. Ross found that her French retirees had indeed forged a rather strong community and that the explanatory model made sense. Later, she surveyed other studies to substantiate the point (Keith, 1980a).

Social Integration From A Japanese Perspective

We are eager to maximize the usefulness of the Fuji-no-Sato example for Western cultural gerontology, but we believe it would be unwise simply to transfer the concepts of Rosow and Ross to the Japanese setting without noting that basic Japanese assumptions about the relationship between individual and society are often at odds with those of Westerners and that the difference renders many of the Western concepts problematic in our context.

All three of Rosow's key terms, "norm," "role," and "group," as they are used in the West, contain assumptions that do not easily fit the Japanese case. Americans use "norms" to refer to social standards that give values to social behaviors, and these values are supposed to be universally accepted, at least within one's reference group. Americans are embarrassed to be reminded that they change values when they change social settings, but Japanese culture has relatively less need for universal norms. Rather, different social situations easily call forth different sets of context-specific norms, and the Japanese have little trouble accepting the value incongruity of situated behaviors.

The term "role" presents similar difficulties. The American drive toward consistency among situated norms derives in part from the American perception that the actor is first and foremost a person , an individual with a personality distinct from his role performances. The Japanese, although able to distinguish between person and role, place a positive value on the total psychological identity of actor and performance. Kiefer has made this point in earlier work on Japanese Americans: "The psychological process of disengagement requires a perception of one's social roles—a perception that is alien to the issei. Behind the Westerner's perception of himself as purposefully engaged in, and thus potentially disengaged from, his intimate social relations lies a long history of philosophic individualism. . . . The issei, however, cannot disengage from his social roles because he is those roles" (Kiefer, 1974:207). DeVos has used the term "role narcissism" to characterize the importance of formal roles for the Japanese: "For some Japanese the occupational role completely takes the place of any meaningful spontaneous social interaction. A tradition of role dedication still can lead to what appears to be an extraordinary capacity for self-sacrifice, such as, in the very recent past, that of soldiers going to certain death. . . . Role dedication can be viewed as the core of the two chief Japanese traditional virtues of loyalty and endurance" (DeVos, 1973:13).

This indicates, among other things, the relative meaninglessness of informal roles for the Japanese. They have some difficulty establishing social relationships that are not governed by formal, institutional roles, whether these be familial, occupational, or communal. Formal roles, where they are operative, tend to take over and dominate Japanese self-perceptions. This is an especially important point in newly created communities like retirement homes, where there may be a scarcity of community-relevant formal roles available to replace the largely irrelevant ones belonging to family and work contexts and to structure and lubricate community social interaction. The informality of social behavior at Fuji-no-Sato, far from providing the comfort that seems to draw American retirees to similar settings, is a source of consternation to many Japanese residents. It is not enough to say that Americans are more skillful at informal interaction; one must realize that Japanese elders lack the automatic perception of themselves that underlies the development of such a skill: a perception of self as an unconnected personality, ne-

gotiating roles . Consequently, as David Reynolds (1976) has shown, when things do not go smoothly in interpersonal relations, Americans tend to seek some action that will correct the situation, but Japanese tend to redefine their role as one in which the situation is acceptable.

These considerations have already clarified some of the differences between Japanese and American definitions of the concept of a "group." In America, almost any perception of shared characteristics—age, for example—can provide the basis for an emergence of group identity or "we" feeling. This is not so in Japan, where social roles tend to fall into hierarchical patterns and where horizontally organized groups, as among age peers, tend to be small, to develop early in life, and to command intense loyalty and careful lifelong preservation. A collection of same-age neighbors is simply not a "group" in any real sense in Japan, even if they share other characteristics like social class or political orientation.

The term "socialization" transfers less awkwardly from American to Japanese culture, but socialization in East Asia has unique dynamics. In America, familial roles tend to be rather loosely defined, and many elderly, unable to draw personal satisfaction from grandparenting and the like, struggle to maintain the extrafamilial roles they occupied when younger. In Japan, by contrast, familial roles tend to be well defined and rather all-encompassing for the elderly, especially women. Although the normlessness and rolelessness of the aged typical of American society seems less serious in Japan as a whole, it is probably more serious for those elderly who lack family roles: the never married, the divorced, and those who have no children or who are estranged from them. For these people, there are few groups that can serve as socializing agents and few meaningful normative behaviors that can be learned. Moreover, as families become more geographically dispersed and less economically interdependent, socialization to the all-important familial roles also becomes more difficult.

As a result, Japanese society may appear on the surface to be undergoing Americanization in many ways (for example, the dwindling rate of three-generation households, the growth of nursing homes and retirement communities), but the meaning of these changes is different in the two cultures. Given the American tendency to emphasize ad hoc relationships among peers in forming

group identities and defining norms, Rosow may be correct about the usefulness of age-homogeneous environments in the social integration of the aged in America. Given the Japanese preference for familial and hierarchical relationships, we doubt that the same principles apply here.

In view of these differences, the study of Fuji-no-Sato assumes great theoretical and practical importance in cultural gerontology. We find ourselves at odds with much of what has been written about social integration in the West, especially with claims that Western models may have universal applicability. Consider one researcher's statement on this point:

Although I did my homework about aging in France, my research focus was not on French old people, but on the possible development of a community by old people in an age-homogeneous residence. There is an atmosphere of Frenchness about the descriptions which follow. . . . However, my central conclusion about patterns of community creation and of individual socialization are intended to be distinct from this ambience, to provide a basis for general hypotheses about community, old people, and communities of old people in other settings. (Ross, 1977:4)

Although we found Ross's method—the close observation and classification of interactions among residents—highly useful, we have had to reject her notion that these interactions can be understood or explained without explicit reference to culture. For example, although it is natural for Westerners to talk about "creating a community," the phrase sounds oxymoronic in Japanese. To the extent that the Japanese individual is closely identified with his social roles, to the extent that most Japanese roles are ascribed, not created or achieved, and to the extent that ascribed roles assume some sort of community to begin with, it is difficult to imagine how one could simply create a Japanese community where none existed before. Indeed, this has been a serious problem in other planned living environments in Japan (Kiefer, 1976).

A Model of Japanese Interpersonal Relations

The significance for our purpose of these East-West differences in social expectations can be clarified through the use of a model. We

have noted the relative importance of formal roles as opposed to informal ones in Japan and the relative importance of familial and other long-standing, intimate relationships as opposed to casual or recent ones. These traits, of course, go together: Close, long-stand-ing relationships tend to be functionally diffuse , to provide partners with a variety of choices, and to meet a variety of their needs. Once we get "close" to someone, we can ask him or her for all sorts of things, and vice versa, in contrast to relative strangers or people we have a limited or distant relationship with. For this reason, we are most comfortable with family and old friends, although there may be many important functions they cannot perform for us, such as granting honor or prestige, providing technical services, paying money, or offering the excitement of new experiences.

The formality of Japanese social relations puts a special premium on these long-standing, intimate relationships because it is relatively difficult to contract new relationships, and once they have been contracted, they are relatively demanding. As a result, it is usually impolite, for example, to introduce strangers to one another unless one has been specifically asked to do so for the purpose of beginning an association that is sure to benefit one or both of them.

The result of such rules is that one's social contacts tend to be divided into sharply defined types, each of which is governed by distinct sets of rules. You want to keep casual acquaintances and strangers at a distance in order to avoid heavy obligations, and you want to keep your intimate relations strong in order to be sure of having your needs met.

Consider Figure 2-1. Each sphere represents a different level of social importance for a hypothetical actor, the most important at the center and the least important at the periphery. The first level contains the most socially and psychologically significant others, such as family members, close friends, and long-term colleagues—in a word, confidantes. Level three represents the world of strangers, people with whom one has at most only fleeting, mutually anonymous contacts, such as people on the same train, waiters, casual customers, or shopkeepers. One need not be overly formal or ritualistic in these relationships. Lying in between, level two contains a broader range of people, from those who are almost strangers to the vast number of people with whom one is regularly but not intimately socially engaged. Here one's identity is known, and one's actions may have long-term repercussions. Behavior is accordingly

Figure 2-1.
Depth Model of Japanese Interpersonal Relationships

regulated by role-specific norms, which in Japan are comparatively inflexible and well defined.

Any complex society can be conceived in terms of a model of "circles of intimacy" like this. But if you think of the boundaries between the circles as being relatively durable, easy to see, and difficult to cross, you can understand the Japanese difference. If the social spheres in Western cultures are thought of as separated by flower beds, the Japanese spheres might be seen as separated by fences.

There are also some culturally specific ideas that are useful in characterizing interaction at each level. In the first level, for example, one can indulge in one's need to be humored, nurtured, or spoiled—the complex made famous by Takeo Doi as amae (Doi, 1973). The prototype of the amae relationship is the parent-child role set, and such a relationship may be so close as to be symbiotic and unique, almost unfettered by normal social expectations. Intimate friendships can be included here if they are of long standing.

Second-level relationships are those wherein one has to wear one's social clothes. Here the key concept is enryo , restraint or holding back, which is chiefly used as "a negative yardstick in measuring the intimacy of human relationships" (Doi, 1973:38). Because

this is the level in which formal role relationships dominate, one must carefully control one's behavior, continuously assessing the quality of the relationship and keeping an appropriate distance. Amae also comes into play at this level, but in a different sense. The parties continuously "read" each other's needs and reactions to assess the changing quality of the relationship. Relationships that are close enough to avoid the need for enryo are ideal, but such closeness is often impossible. Enryo need not be used in either first-level or third-level relations, both of which are more casual, although for different reasons.

Looking at this model from the standpoint of social support, level one relationships are the strongest. People tend to refrain from seeking help from those at level two. Having received help from second-level people, one will feel giri , or psychosocial indebtedness, a need to repay the favor quickly. Third-level relationships are not part of one's support network, and help can be sought there only in extreme emergency.

This model is not static. Although the basic structure of social relations may be stable, the real value of this model is that it can also be used to clarify the process of life span social development. Relationships can develop in either direction; a stranger can move into the second level, then gradually into the first as time goes on. Likewise, even a parent-child relationship can grow cold and distant, losing the meaningfulness of first-level relations. Relationships must be defined mutually; it is essential that the parties share a similar perception of their quality and depth. The Japanese are socialized beginning in childhood and continuing at school and at work to be skillful at this kind of perceptual bargaining.

This perspective is a theoretical backbone of Plath's work on the life cycle: "Growth then becomes in part a property of others, particularly of those who are one's consociates . The term may be an unfamiliar one, but it is apt here. It derives from the work of Alfred Schutz and the phenomenologists. If 'associates' are persons you happen to encounter somewhere, sometime, 'consociates' are people you relate with across time and in some degree of intimacy. They are friends, lovers, kinsmen, colleagues, classmates" (Plath, 1980:8). Consociates, then, are essentially the people in the first (and occasionally second) level of our model; associates are most of those in the second level and a few in the third.

Figure 2-2.
Analytical Model of This Study

Plath's use of the term "convoys" to describe groups of consociates, whose personalities, life circumstances, and relationships evolve slowly over many years, can also be understood with the help of this model. First-level relationships are extremely durable; and when they grow out of second- and third-level ones, their evolution usually takes many years as well (Plath, 1980:224-226).

The Model And Fuji-No-Sato

The model, then, helps us visualize the evolution of social relationships at Fuji-no-Sato. Our purpose in using it is analytical. Given that the residents are mutual strangers until they move to the community, it can be theorized that their relationship will develop, as shown in Figure 2-2, from the outer edge of level two toward, and possibly into, level one. How and how far their relationships move in this direction is the focus of our analysis. Our initial intuition is that, at the beginning, the residents employ outer second-level norms and maintain distance appropriate to that mutually perceived relationship. As they get to know one another, they will adjust the application of norms and the perception of distance while changing the definition of their relationship.

We take this as an "emic" analytical model of social integration at Fuji-no-Sato. We do not try to operationalize the concept of "community" as a dependent variable, but simply assume that Fuji-no-Sato functions as a socially defined context, limiting the kinds of possible relationships among its members. Viewing Fuji-no-Sato as an independent social system, then, and following Rosow's definition, we define social integration as the locating of individuals in the system and the patterning of their relationships with others. Likewise, "socialization" is defined as an important mechanism of integration—the process of development of patterned social interaction. "Norms" refer to social standards against which one's behavior is to be evaluated. "Roles" generally mean constellations of rights and duties—meaning both formally recognized roles, like elected offices in the residents association (which are very limited) and the much more important informal roles of neighboring, friendship, and so on.

There are five interrelated factors that render relatively problematic any attempt to develop "deep" interpersonal relations on the part of the residents at Fuji-no-Sato. (1) Although American relationships seem to the Japanese to be too shallow and ephemeral, it is nonetheless true that when Americans meet each other for the first time, they handle the establishment of relationships better than the Japanese do in 'similar situations. (2) It is difficult in Japanese interpersonal relations to differentiate conceptually feelings from other social values. The Japanese generally expect both emotional and instrumental needs to be met in their relationships with others. It is necessary, then, for acquaintances to keep both perspectives on their developing relationships. Although the Japanese may be less skillful than Americans in building new relationships, they expect and often attain deeper and more sensitive relationships. (3) As Plath pointed out, it takes considerable time for the Japanese to develop solid first-level relationships. The depth of relationships generally correlates with their duration. This is a serious obstacle to the rapid development of deeply meaningful relationships among Japanese. At Fuji-no-Sato, however, this time factor is mitigated by at least two of the "background factors" identified by Ross: the physical distinctness of the community (which intensifies "we" feeling among residents) and its limited size (which concentrates opportunities for interaction). Both factors

tend to accelerate the intimacy process. (4) First-level relationships, with the exception of family, are usually informal outgrowths of second-level relations, which are formal and role-specific. In other words, the Japanese develop new first-level relations only with people in the second level who share some basis for sustained, frequent interaction—an attraction or a natural tendency to be together over long periods. (5) Fuji-no-Sato is very unusual as a Japanese setting because the residents' relationships do not begin with the performance of ascribed roles. The residents lack preexisting structural reasons for regular interaction that would naturally lead to the development of first-level relations.

Limits of The Model

The model we have just presented seems to us to make fair sense out of the actual impression we got from the people at Fuji-no-Sato. Often they seemed to be struggling, against considerable cultural barriers, to find their way to a secure and socially comfortable way of life. However, in the process of explaining their struggle, the model does not give much scope to the positive aspects of their lives. In fact, the residents were able to act cooperatively to solve certain common problems and thereby to enhance their feelings of both individual effectiveness and collective "we-ness" vis-à-vis the management. They controlled tension among themselves to some extent through formal channels of communication and representation; and they were able to sustain a modest level of polite interaction that might, in the long run, lead to deeper and more satisfying relationships for many.

To some extent, the model we have presented helps explain these positive features of social life as well. Level two, the circle of giri , demands attention to the things that keep conflict within bounds and sustain instrumental exchange—politeness, respect, return of value received, anticipation of one another's instrumental needs. But there is certainly more to social life at Fuji-no-Sato (or anywhere else) than these.

When the collective behaviors constrained by rules of prescription and proscription are subtracted from the whole round of communal life, what is left is a considerable "free zone"—the area where humor, whimsy, and appreciation of the eccentric and

unique come into play. We have been saying (and we believe it is true) that this area is not large in Japanese society compared with some others. But it is there, and it is important. In fact, without it no society could deal effectively with the novel and the unexpected, a failing that rarely can be attributed to Japanese society.

The description of Fuji-no-Sato contains many important incidents that are not clarified by our model. The "rulefulness" of human behavior is limited.