The Company Work Group

Thomas P. Rohlen

Although their labels have been diverse, few writers comparing Japan with other industrial societies have failed to give a central place to the strong tendency toward group involvement one finds in Japanese institutional life. This group emphasis is generally understood to predate the modern period, and thus it is a major element in the discussion of such issues as cultural continuity in modern Japan and the reinforcement of modern organizational patterns by traditional values. The consensus about the importance of group involvement is not, however, enriched by a very extensive understanding of just what Japanese groups are like, how they vary, and what factors are causing them to be refashioned. Obviously the term group is a vague category encompassing numerous quite distinct entities. This paper aims to bring into sharper focus one crucial subcategory—the company work group. The discussion is based primarily on material from a one-year fieldwork study (1968–1969) of a bank of three thousand employees and a three-month study (1972) of a plastics manufacturer employing six hundred people in two factories and a head office.^[1]

[1] Support from a NIMH predoctoral fellowship for the earlier study and from an additional NIMH grant for the more recent fieldwork is gratefully acknowledged. A more detailed account of the bank's work groups and the general organizational context of white-collar work is contained in Thomas P. Rohlen, For Harmony and Strength: Japanese White-collar Organization in Anthropological Perspective (Berkeley and Los Angeles: University of California Press, 1974). This paper relies more heavily on material from white-collar groups. For further detail on the character of blue-collar work relations, see Robert J. Cole, Japanese Blue Collar: The Changing Tradition (Berkeley and Los Angeles: University of California Press, 1971); and Arthur M. Whitehall and Takezawa Shinichi, The Other Worker: A Comparative Study of Industrial Relations in the United States and Japan (Honolulu: East-West Center Press, 1968).

The General Character of Work Groups

A key variable in most employees' response to their company is the quality of their face-to-face relations in office or factory groups (usually called ka ). Informants frequently cite teamwork and related qualities prominent in company ideologies in their descriptions of good relations,^[2] and those dissatisfied point to interpersonal problems in work groups as the source of their unhappiness. Management takes a similar view of the importance of work-group relations. Regarding white-collar work in particular, personnel departments show a preoccupation with improving office morale as the solution for problems as different as emotional instability and declining productivity. As further evidence of managerial concern, personnel departments, which are charged with improving work-group morale, typically have far greater status and authority in the Japanese company system than in American firms. In sum, management regards good relations at the small-group level as crucial to maintaining high levels of energy and attention. This perspective did not originate with modern theories of personnel management (although the thinking of the Mayo School has been influential), but rather it echoes ideals and psychological predispositions characteristic of premodern Japan.^[3] The large organizational framework places various constraints on this pattern, and yet any attempt to characterize Japanese companies without recognizing that their dynamic is generated, fundamentally and quite intentionally, in small-group contexts would obscure a most significant quality.

General Structure .—No one works totally outside a group context. The president and his council of executives have considerable independence, yet ideally they decide matters through frequent joint meetings. All employees, even those with the most isolated tasks, have a place in some work group. Almost all white-collar employees and many blue-collar workers are assigned first to a group and then to a specific job. Rank, skill, and actual task are correlated, but their final sorting out accommodates the requirements of the group first. Nakane notes that Japanese companies hire potential, not skills,^[4] and to a lesser degree this observation holds for the way employees enter company work groups. On-the-job training often brings individual potential into line with group requirements. Regular employees are sorted by a rank system, which gives a pyramidlike

[2] Japanese managers are inclined to enunciate rather elaborate statements of company goals, ethics, and other ideals. See Rohlen, For Harmony and Strength , for a detailed account of the bank's ideology.

[3] See, for example, Kawashima Takeyoshi, Nihon shakai no kazokuteki kosei^[*] [The familistic construction of Japanese society] (Tokyo: Nihon Hyoronsha^[*] , 1950); Kawashima Takeyoshi, Ideorogii to shite no kazoku seido [The family system as ideology] (Tokyo: Iwanami shoten, 1957); and John C. Pelzel, "Japanese Kinship: A Comparison," in Maurice Freedman, ed., Family and Kinship in Chinese Society (Palo Alto: Stanford University Press, 1970).

[4] Chie Nakane, Japanese Society (Berkeley and Los Angeles: University of California Press, 1970).

structure to the work groups. In manufacturing firms, clerical and engineering workers tend to be organized into one rank system, while blue-collar workers are organized into a separate rank system. Even more than for white-collar workers, promotion up the blue-collar ranks is a direct function of years of service.^[5]

The result is a hierarchy in most work groups that closely corresponds to age. Exceptions occur if (1) older women are present, since women are rarely promoted above the lowest few ranks,^[6] (2) older people with short company service are present, or (3) men who have been promoted either very slowly or very fast are present. In the blue-collar situations, the first two factors were the primary causes of age-rank discrepancies, whereas the third possibility was most influential in white-collar situations. In both companies there is also a small percentage of persons (such as temporary workers and custodians) who are not regular employees, and thus of lower status. They are typically older and unobtrusive to a fault. The group's informal social activities seldom include them.

Size of groups varies considerably. The smallest branch in the bank and the smallest factory section both have less than six people, while the largest units number over forty. Whenever size passes something like twenty-five, subdivisions become the loci of face-to-face interaction and interpersonal relations. This can be a source of confusion, and furthermore some larger units lose the valued atmosphere of small-group involvement. The bank's managers recognized the special problems of its large branches in this regard.

The ideal office, from the point of view of smooth relations, has ten to twenty people and an ascending distribution of ages within its hierarchy. Groups of ideal size and symmetry are not likely, given the varying work requirements and the numerous ways age-rank disjunctures may arise. In the bank, the youngest section chief (kacho^[*] ) is thirty-eight, but many are in their late forties and early fifties. In the manufacturing company, section leaders are as young as thirty-three, and before age fifty they tend to be promoted to the position of department chief (bucho^[*] ) without any change in their actual leadership responsibilities. Factory supervisors (kantoku ), the highest blue-collar rank, are all over fifty. Whenever the leader of a work group is old, younger men are usually crucial in subordinate positions to

[5] At least this is true of the promotion system of the plastics manufacturer studied.

[6] The bank employs a large number of women (one-third of the total personnel), and offices almost invariably have a ratio of two men for each woman. The women are all unmarried and almost always young. They must leave employment upon marriage. Factory work groups, by comparison, have unmarried and married, old and young women, and there is a much greater variation in the proportion of women in each group, since work requirements vary greatly from unit to unit. Recently, the bank promoted three women above the lowest rank. This "first ever" move was made to boost morale among working women, but it did not result in women supervising men. The manufacturer, thirsting for inexpensive but steady workers in its rural plant, has taken to hiring local farm women on a half-day basis. They are mostly middle-aged and without previous factory experience.

keep close relations with the very young, lead the group's recreational activities, and generally serve as a bridge between the generations. This bridge is missing in some cases, and there the generation gap, so prominent in Japanese public culture, emerges full blown. Older men, furthermore, reportedly resent working under a younger man, but few such problems actually occur in either company. It seems that because the younger leaders are sensitive to the problem, they give less cause for resentment.

Group Competition .—Competition among the branches of the bank and the sales teams in the manufacturing company is regular and highly conscious, but it is carefully avoided among individuals within groups. Each branch of the bank is assigned a quota for new deposits, and it is well understood that the standard of success is relative to the record of other offices. Branches that do well are given public notice at awards ceremonies. Other classes of recognition (for effort, cooperation with other branches, enthusiasm, and office efficiency) are also distributed to the most successful branches, and awards given to deputy-group leaders (dairi ) almost invariably reflect office-group performance. Only the lowest ranking employees appear to receive awards without particular consideration for the record of their offices. Although branch chiefs seldom receive individual rewards, any honors to their office are honors to them and strong indications that they are looked on with favor by those who decide promotions.

Neither company gives a commission or other direct-pay incentives. There are cases in the bank of individuals with good collection records actually giving credit for their deposits to others whose records are down for the month. Similarly, sales in the manufacturing company are handled totally on a team basis. The result is a situation in which (1) group leadership is emphasized, (2) leaders find themselves in competition with one another, and (3) leaders must rely on their followers for support and further promotion.

Morale is, thus, not simply a matter of the group, but also a question of loyalty to the group leader. Although the production side of the manufacturing firm does not lend itself to intergroup competition, the sense that motivation is in proportion to the group's enthusiasm for its leader still exists.

Morale .—Several general patterns of organization characterize the normal activities of office groups. These are manifested in the adjustments of the group to two basic requirements—work output and group maintenance.

The two patterns are a formal hierarchy and an informal circle. The arrangement of desks within any office reveals much of interest in this regard. Typically, the chief's desk is at the back-center of the office. His deputies' desks are on either side and a bit forward. From there, the desks of the rest of the staff are gathered in small infacing clusters that form a haphazard ring. Whether they become an actual circle is less relevant

than the fact that the desks are arranged to face one another. The exact pattern is determined by the space available and the preference of the chief, but never will one see a section arranged so that people are looking at the backs of others in the same group. Even with this group orientation, the positions of importance are easily determined at a glance. The chief, at a slightly larger desk, is clearly the focus of the arrangement and his location permits him to take in all that transpires within his section in a glance.

The office at work generally follows the lines of hierarchy that link employees to deputies and deputies to chiefs, but there are qualities that distinguish it from an impersonal kind of office hierarchy. First, the general group relationship is expressed by the inward directioning of desks. Second, the character of vertical ties within the office is direct and close, for no barriers are set up between the leader and his followers.^[7] There is a striking absence of any sense of individual isolation or of mechanical relationship in this pattern. The office is arranged to maximize the group sense, and interferences such as noise are not considered serious enough to justify separation or seclusion. Furthermore, authority does not require the creation of impersonal distance.

In factory situations, one finds the foreman moving about discussing problems, helping with a piece of work, talking to outsiders and instructing the inexperienced. With the exceptions of lunch and evening quitting times, break-taking and consultation are not formally regulated but determined rather by the flow of work. The intermediate ranks in most factory work groups share some supervisory responsibilities, but they also have regular tasks that limit their movement. General get-togethers for the communication of messages occur at a brief daily meeting before work starts, and more extended discussions of common problems take place after work, without overtime pay. The same is true for office meetings.

An informal circle is the common arrangement for relaxation and fellowship. Group morale and commitment and the intensification of individual friendships are sponsored through activities arranged to reduce the sense of rank, age, and even sex differences. The circle form stresses the fact of common membership. If the essence of the working order is hierarchy, the essence of the group principle is the circle. It is not difficult to perceive the interplay of the two in most office activities, for seldom is the pattern strictly linear or circular. Each activity and context calls forth slightly different arrangements, and the group finds it quite normal to shift from one to another. One characteristic of vital work groups, in fact, is their frequent but orderly shifts from one arrangement to another,

[7] Most offices have special side rooms for meetings with visitors, but only men of directorial rank in the bank and the president of the manufacturing company have private offices. The very high cost of office space in Japan could be cited as a reason for the minimal development of the individual office pattern, but this is far too simple an explanation given the emphasis on familiarity in work-group relations and the availability of side-room space.

permitting realization of the spectrum of possible official and personal relationships within the group.

It is, of course, imperative that hierarchy regulate the daily affairs of the office. The shifts to informal, group-centered activities require effort and time (and are most vulnerable to the debilitating effects of complex organization and patterns of modern living). The general expectation that group relations will be satisfying can produce a breakdown or loss of motive force in the working order unless discussion meetings, parties, and other maintenance gatherings are forthcoming. It is part of the leader's job to be sensitive to this general need and to meet it energetically.

The work of group maintenance, above all, requires time. Just working together in a hierarchy is not sufficient. For the group to coalesce and realize the potential for cooperation and team motivation, informal social activities after work and on weekends are needed. This is why administration and office leaders plan, conduct, and partially pay for a heavy schedule of group social affairs. These activities are ostensibly voluntary but virtually all members participate—some out of group loyalty, some out of interest, and a few with a sense of being coerced.

The emotional momentum and sense of group solidarity are largely cultivated during such social activities.^[8] The annual calendar of office events typically includes two overnight trips (ian ryoko^[*] ) at company expense, monthly Saturday afternoon recreation at company expense, and an average of six office parties held to celebrate one thing or another. In the manufacturing firm economic stringency has caused the company recently to stop paying for work-group trips, but most groups have continued to go at least once a year at private expense. Also for financial reasons that company has no monthly Saturday recreation program, believing that company-sponsored sports teams and summer camp-outs are the more effective means of satisfying its young people. Several basic differences between the two types of companies are indicated here. First, company prosperity and size are correlated with greater expenditure on such recreation and, second, the branch banking system offers no alternative to office groups as units for the organization of leisure, whereas a factory or main office readily accommodates sports teams, youth groups, and other alternatives. We should also note that hikes, after-work drinking expeditions, mahjong and bowling, involving some but not all people in an office (age and sex being the primary sources of separation), are frequent activities in both companies.

The drinking party is a particular and important part of many work-group social relations, although the blue-collar workers report its decline. The reasons for its influence can be appreciated from an analysis

[8] Such activities are common to most Japanese work groups. A teachers' group party is illustrated in John Singleton, Nichu^[*] : A Japanese School (New York: Holt, Rinehart and Winston, 1967).

of its dynamics and the expression of emotional participation involved. The first stage is marked with the serving of food, toasts by leaders, and much reciprocal pouring of drinks within the group. To keep up with the generosity of neighbors, one is constantly obliged to drain a glass or cup only to have it refilled by another waiting patiently with bottle in hand. The purpose is primarily to get everyone loosened up and happily inebriated, but the exchange of offerings and acceptances provides opportunities for the subtle expression of feelings of affection, admiration, and even repentance among persons who in normal circumstances are circumspect with one another. The gesture of filling another's glass is a convenient bridge across strained relationships. No one is forgotten. No one need pour his own drink. Support and attention for all revolves within the group, and a swaying, relaxed fellowship is created. This first stage prepares the atmosphere for the rounds of solo singing performances that often follow.

It is no easy thing to stand before a group and sing. Trembling hands, shaking voices, and nervous faces reveal the stress many suffer at the moment they are selected. The group responds sympathetically, however, with clapping, encouragement, and sometimes by actually joining in the song. When the ordeal is over, it is normal for the performer to experience a rush of relief and a feeling of gratitude to the others for their help. From that point he is deep into the group emotionally, for he has revealed his humanity and been accepted by the others; what occurs is much like a confession-forgiveness sequence. Each also acts as part of the group in helping other soloists through the ordeal and receiving them into the fellowship. With the solos over, restraint decreases noticeably, and the drinking has brought the men, at least, to a plateau of drunkenness. The process of uniting the group has ended, and there is no further obligation for full participation. Those in the mood are left to continue the party.

I estimate that in a week the bank office groups spend fifty-six hours working together and four to six hours socializing. Drinking parties do not occur more than every month or two, but they do represent pinnacles of emotional involvement that are not soon forgotten. The point that the work group does considerably more than simply share an office and a set of tasks is evident. This fact colors all that transpires within the hierarchical framework.

Decision-Making .—At the office level, discussions are another activity that serve maintenance functions. The manner of arriving at conclusions, however, varies greatly. The chief is empowered with the authority to make all decisions, and if he wishes he may maintain a strict order in which he alone decides. On the other hand, since the entire group is involved in the work of the office, the chief is free to submit decisions to discussion and, again if he wishes, to the rule (a rather flexible one) of group consensus. He may also encourage open discussion but decide the issue himself. There is no fixed procedure; among the crucial determinants

are the nature of the issue, the type of group, and the personal style of a given leader. Group processes require considerable time, and to push for an early consensus or to fail to involve everyone can often lead to resentment and opposition. It is difficult to hurry discussion, and one simple rule influencing office decision-making is: the more urgent a decision, the more likely it will be made according to hierarchy. People favor group discussions, however, and have an expectation of participating in considerations of group affairs. This includes the joint consideration at the office level of new company-wide policies and directives. Opening an issue to general office discussion may take time, and may even reveal an embarrassing gap in opinions, yet the sense of unity and the identification with the solution that usually results make the practice worthwhile from a leader's point of view.

The subtle question of the delegation of decision-making power arises in this context. A chief, anxious to utilize the positive effects of group discussions without relinquishing his own control over major decisions, may hold numerous meetings but submit mainly issues of minor significance, retaining his authority over crucial matters. He may also choose to discuss his own decisions with the group. The alternative of introducing a question for the group's consideration only after being certain that his opinion will previal is also available. The conclusions may be foregone, but the fact that discussions have been held is very important since a chief is expected to share his opinions, ask for advice, and permit dissenting voices to emerge. His trust and respect for the others and his acknowledgment of the group's importance are symbolized by the discussion process.

These general considerations only begin to accommodate the range of variation. Some offices are characterized by a steady stream of meetings, others may have no more than one a month. What are the differences involved? In the factory, production units do not gather frequently for discussions. They are not free to leave their tasks, and unless the issue is serious, they do not feel inclined to stay around after work just to take up minor matters. For the most part the unit leader consults with his men individually, and sometimes there are huddles of three or four men over some problem. Routine and minor adjustment characterize the situation, and there is little for group consideration except when new production goals or the Zero Defect program are discussed. At the opposite extreme are design engineer teams, sales teams, and groups empowered to formulate new programs. Meetings are frequent, open, and animated with leaders taking rather nonassertive roles, but always formulating the conclusions. Offices characterized by a high degree of routine, such as accounting, meet less frequently.

The degree to which central authority delegates decision-making power to groups is another crucial matter. Characteristic of the management of both companies is the practice of assigning specific goals and expecting the

work groups to develop the means to achieve them. Obviously, groups with established routines will need many meetings only when the goals assigned create a crisis because the routines must be overhauled or new ones established. In the factory it is notable that blue-collar groups are strongly inclined to regard the established procedures of production as their domain. The resident engineers do not have control over regular production, and supervisors (men at the top of the blue-collar ladder) are given the time and latitude to solve problems in conjunction with their subordinates that would quickly draw intervention from higher (outside) authority in American plants. This is most true of units using techniques and equipment of long standing, since experience has led to mastery. In the case of new forms of production, however, engineers are much in evidence. While such considerations help explain the degree of variation in small-group autonomy, the general rule that goals come from above, implementation from groups below, remains characteristic.

I should add that stagnation and demoralization can result from a lack of direction and encouragement from managers (usually the department chiefs) intermediate in the hierarchy between the work group and the general goal-setting top level. What the group-level leader needs is a guarantee that his group's actions will be accepted, supported, and brought into coordination with the total organization. Forceful middle managers are those who form close personal relations with group leaders below them and encourage them to dynamic action.

Handling Deviance .—Turning back to the internal processes typical of small groups, we should consider briefly the way the group responds to deviance. Actions at variance with the expected and required behavior within the group seldom if ever take the form of open confrontation with authority or a refusal to perform required work. Numerous factors, including the ethic of cooperation and the general restraint of expression in Japan, inhibit direct confrontations in the office. Deviance and opposition take more private and qualified yet discernible forms, and the response of the rest of the group to even mild nonconformity is worth considering.

Deviation is most often expressed by withdrawal from participation in the group and with its leader. The individual who is unhappy or in strong disagreement, but who cannot express these feelings, is likely to set himself at odds with group norms in areas of behavior where matters of personal preference and the standards of the group intersect. Some typical examples would be avoiding office social activities, excessive make-up, drinking too much, joining a leftist youth group, and remaining silent during group discussions. These kinds of actions disturb the sense of unity, but cannot be labeled insubordination and summarily dealt with. They are conventional symbols of dissatisfaction and isolation from the group. It is true that some people quit in protest against a group or its leader, but to do so without a

prospective job can be a form of career suicide, especially for white-collar men. Such eventualities are avoided by restraint and a sensitivity to the nuance of individual participation.

Conformity and full participation are, in one sense, offerings individuals make to the common existence. They should not be forced from them. The individual right to disagree and criticize through acts of nonparticipation is recognized, and the group must attend sympathetically to such indications of a problem. Often acts of "resistance" (teiko^[*] ) to the group or its leader are understood as analogous to the rebellious behavior of children toward their families, particularly when hurt feelings rather than principles are at the heart of the matter. The individual's natural state of existence is within some group, it is assumed, and resistance is easily interpreted as a sign of unhappiness and an indirect expression of personal need, which should be answered with sympathy and special attention that create a sense of belonging. The group's members, especially its leaders, experience considerable irritation to be sure, but these reactions should and usually are repressed. Understanding and tact are required, reminiscent of the general expectation that parents will be flexible and tolerant in the face of their children's recalcitrance. That order and participation begin with individual feelings and cannot be forced through rules and punishments is the crucial assumption.

What about difficult individuals, we might ask? How does the group deal with the contentious, the overly ambitious, the lazy, and other similar types? This question is of interest primarily because personalities are seldom typed this way in small group contexts. Actions that would fit such categories are not absent, but the practice of writing people off into immutable character boxes is avoided, and in its place one finds a faith in the possibilities of group socialization to smooth rough edges. Much of the most effective smoothing occurs as part of senior-junior relationships, and the group's informal social activities offer further opportunities to generate a cooperative spirit in the less well adjusted. Even so, there are people who come to be recognized as permanently difficult, and the solution is to place them in isolated jobs (company librarian, warehouse attendant, truck driver, and so on) that effectively separate them from work groups even though they remain attached to one or another group for formal purposes. Rarely is a troublesome person treated with anything but sympathy and constraint, and this in turn undoubtedly generates constraint and cooperation from most potentially difficult people. The calm surface of mutuality, however, can be accompanied by much hidden disenchantment and resentment.

When most of the group object to their chief, it is likely that complaints will find channels to higher authority, particularly the personnel department. In the bank such a leader is likely to be transferred, but in the much smaller manufacturing firm such shifts are more difficult, and attempts to change the leader's ways is the likely outcome. In either case, quitting or

the threat of it from established employees is a certain spur for the reconsideration of a group's situation by management.

Short of such extremes, it is usually necessary to live with a degree of problems, and the chances of redemption often hinge on the departure of those who do not fit in, or the arrival of a new chief or deputy able to draw the office back together. Even so, efforts to engineer cooperation and full participation, through focusing increased attention and affection on those withholding themselves from group life, are highly predictable. The effectiveness of these methods is not as predictable, however, partly because of the very fact that they are expected. Finally, it is my impression that the white-collar world is less likely to show the tension of work relations than the blue-collar, perhaps because open hostility is even more inappropriate.

Superior-Subordinate Relations

Employees in both companies share a vocabulary of labels for work place relationships, and while their typology is not fully discreet, these labels do mark the basic starting place for the distinctions that constitute the interpersonal world. As such they deserve enumeration and clarification, but we should be forewarned that like most typologies, they only begin to sort out the situation.

As in most organizations the most crucial face-to-face relationship is the one between a superior and subordinate who work together. In Japanese company organization these are regularly employees of different rank who are assigned to the same work unit.

Because of seniority in promotion, it is highly likely that a superior will be older. No man in either company had a woman superior, and only a small percentage of women had another woman as her superior. Two terms, uwayaku and joshi^[*] , refer to superior or boss, and the term buka refers to subordinates. The characters for "above" and "below" are used to write these words, indications of the clarity of the hierarchical order involved. When addressing one another, subordinates and superiors have a different set of terms. One may address his boss using either his family name (plus the suffix san ) or his rank (for example, kacho^[*] ). The superior has the option of using the subordinate's family name suffixed by san or the more intimate kun . (Only once have I ever heard a woman subordinate addressed with kun , however.) While details of address may seem rather mundane, they provide an outline of some of the basic dimensions of the relationship. It shows that hierarchy is a constant in both address and reference and that the option of openly expressing intimacy is given only to the superior. The mutual use of family name (plus san ) offers an egalitarian form of address that bank policy urges, but in practice both official titles for addressing superiors and kun for male subordinates is often preferred. People say there is more flavor and a sense of connectedness to these expressions. While

women may be addressed familiarly by their first names, the widespread use of the intimate kun for male subordinates reflects the centrality of male-male relationships in business organization. Male-female relationships are essentially appendages to a central core of hierarchical male ties.

Much of this will not seem particularly unfamiliar to Americans with experience in complex organizations. Clearly hierarchy, degrees of intimacy, and sex distinctions are part of our own systems of address and reference. Our language, however, provides greater flexibility for the establishment of degrees of intimacy and equality (such as the use of first names or initials), and we may obscure hierarchy altogether by using first names to address superiors. Sense of hierarchy is more constant in Japanese.

One's boss may or may not be personally close. It is his official position alone that makes him a uwayaku or joshi^[*] , and this becomes clear when we contrast the superior-subordinate dyad with another relationship, the senpai-kohai^[*] (senior-junior) relationship that is also hierarchical but centered on interpersonal involvement. While senpai often means any person senior in age, it is common in a company or other specific organizational context to use the word to refer to (and occasionally address) one or a few older people who are especially close and supportive. Ideally, one's superior in the official hierarchy is close and supportive ("like a father or older brother"), but for a variety of reasons, the ideal is generally hard to realize. It is, therefore, necessary to distinguish the official roles of superior and subordinate from the interpersonal senpai-kohai relationship even though both are hierarchical and both can, and ideally should, overlap. The distinction is crucial to any analysis of the nature of authority in face-to-face relations in Japanese organization.

Before discussing authority, however, it will be useful to examine the senior-junior relationship, for it provides an ideal model for its official counterpart.^[9] To begin, the Japanese have a comparatively positive attitude toward hierarchical relations involving explicit age differences. They do not assume that hierarchy precludes the opportunity for intimacy. The term for senior, senpai , is a compound of two characters: the first meaning before or ahead, and the second meaning companion. A senpai is understood to be a person who proceeds or leads with the

[9] There is much more to be said about senpai-kohai relations in Japanese companies. For further information and discussion, see Cole, Japanese Blue Collar , Nakane, Japanese Society , and Rohlen, For Harmony and Strength . It is worth noting that (1) most relations of this sort are significant while the younger man is still unmarried, (2) that it serves important socialization functions, and (3) that it is near universal for young men in their late teens and early twenties but of greatly diminished importance afterwards, except as the basic relationship within the cliques that exist at the managerial level. The great majority of senpai-kohai ties among young men derive from working together, living in dorms together, or playing on a sports team together. The relationship among older men is based more on mutual career interests and/or mutual recognition of ability.

implication that those who follow are his companions in the same pursuit, career, or institution. Kohai^[*] , literally "companion that is behind," expresses the other half of the relationship. The complete image created by the characters is one of "friends," one ahead and the other behind, passing along the same path of endeavor.

While all those who are older in a company are, in a general sense, seniors, the term is used most often to refer to one or a few specific older individuals of the same sex who are particularly close and protective. A senior, however, may have more than a few juniors—one of a number of interesting correspondences between this relationship and that of parent and child. Among the reasons for the division along sex lines is the implied comradeship between senior and junior, a form of comradeship that does not easily cross the boundaries of sex.

The basic characteristics of the relationship are:

1. The senior is older than his junior, has worked longer for the company, and is in a position of relative power and security. This position enables the senior to assist the junior.

2. The senior is beneficially disposed toward the junior, and befriends him.

3. The junior accepts the friendship and assistance of the senior.

4. These acts and related feelings are the basis of the relationship. There is no explicit agreement.

5. Ideally, the junior feels gratitude to the senior for his beneficence, and this feeling is accompanied by a desire on the part of the senior to become a good senpai for those younger.

The similarities between this and parent-child and older-younger sibling relationships are more than coincidental. It appears that family relationships provide the underlying model for the senior-junior pattern. Interviewees occasionally stated that "seniors are like older brothers"—a revealing analogy, particularly if we examine the ideology of the Japanese family. First, interdependency and continuity are central to the family ideal. Second, the emphasis on continuity translates into gratitude toward past generations and obligations to assist future ones. Third, affection and hierarchy are understood as mutually reinforcing rather than contradictory. Last, the dependency of the younger, weaker party is not only accepted, it is the focal point of the relationship. These are basic elements in one major Japanese code of interpersonal relations, which make seemingly dissimilar situations and involvements, such as a family and an office group, in essence quite comparable. Because of this code, company senior-junior relationships receive powerful reinforcement from the patterns of interpersonal relations throughout the society since all repeat the same basic message.

The kind of assistance expected of a senior varies greatly. Not only will he advise, console, and protect his junior, but like a good older brother, he is also likely to be a strict judge of performance and a stern task master. A

good senior is one who fosters the growth and adjustment of his junior with a strong mix of encouragement and criticism, all in the context of a close friendship. Socialization of young men into the company is primarily accomplished within this and the larger work group contexts.

The pattern describes an image of one kind of ideal working relationship, one that is secure, beneficial, reciprocal, and selfless to a degree. The ideal of close involvement between older and younger, experienced and less experienced, more established and less established is often applied to the relationship of superior and subordinate in the official system of organizational roles. Company leaders, for example, are occasionally referred to as senpai in company publications. The paternal and familistic philosophies (onjo^[*] shugi and kazokuteki keiei ) espoused by most companies also suggest that leadership should be as sympathetic, protective, and unselfish as good senpai . While clearly the ideal may go unrealized, it does establish a set of expectations about proper leader and follower conduct.

Returning to the question of authority in the superior-subordinate relationship, we should note that legalism (such things as explicit rights and duties) is not characteristic of official hierarchical relations in Japan. Instead, short of leaving the organization entirely, a subordinate has no recourse for contending directly with his superior. Open argument with him, refusal to obey orders, other forms of "insubordination," and open appeal to higher authority are all possibilities so rare in the Japanese context that they never occurred during my observations. There are no rules that would condone or permit such actions. People with exceptionally strong grievances usually hide them behind a screen of silence until they can leave the organization or the authority of their superior entirely. Thus, the superior has absolute authority, but without legal basis. Like parental authority, it derives from general custom and the compliance of subordinates. In the companies I studied, no union power or set of company rules stand between a superior and his ability to exercise authority. Nor are there any reports of dramatic scenes of inflamed passion or righteous resistance to an unfair boss. The only times such overt expressions occur is when union and management adopt recalcitrant and hostile stands toward each other; at this juncture the company is likely to cease existence in organizational terms. Such a rare occurrence underlies the fact that in face-to-face relations Japanese are, by Western standards, unused to and unable to cope with open expressions of resistance to authority.

In another equally real sense, however, the superior finds his authority strongly limited by the general expectation that he will look after the best interests of those who work below him—that he will be their guardian or patron in the senior-junior model.

It would be erroneous, however, to view Japanese superiors either as pseudo-parents or as extraordinary and unlimited tyrants. These two extremes, although possible, are hardly ever realized in actuality. Confu-

sion does arise about the nature of authority in instances of trouble, since the expectations of obedience and paternal regard can become separate and opposing interpretations in times of discord.

The possibilities inherent in the Japanese superior-subordinate situation are of even greater complexity. On the one hand, there is a strong tradition, particularly emphasized in the military heritage, for a very strict form of order and discipline, and it is not unusual to come across scenes in Japanese offices highly reminiscent of the most thoroughgoing military procedure. At other times, however, one may find leaders and their followers gaily drunk together after work, staggering arm-in-arm down some back alley. This is part of a tradition of co-worker camaraderie. Superiors seldom hesitate to ask subordinates to do unpaid overtime work (at least in the white-collar world), yet it is equally common to find them involved in helping a subordinate with a private problem at considerable personal sacrifice. Such seemingly contradictory patterns are actually characteristic of common themes such as group involvement, an emphasis on feeling, and little capacity to separate the people from their roles.^[10] Definitions of what is proper vary with the people involved and the context, not with some contracted or administrative formula. The energetic, warm leader creates for himself great power, whereas the opposite commands no more than a mechanical authority based on his position alone. In Japanese companies, loyalty to an immediate leader is obviously a crucial aspect in the flux of motivational levels. The more important group motivation is to accomplish the work at hand, the more important leadership becomes.

It is ironic, then, that companies (on years of service considerations) regularly place many men in group leadership positions who lack the capacity to motivate their followers in the highly personal manner required. In such cases, the ideal model, with its heavy demand for personal involvement, actually serves to generate disappointment and resentment among subordinates. Companies have a particular dilemma in this regard since they consciously seek the benefits of group-ordered work, yet are bound by other factors to retain a promotion system based on considerations other than leadership capacity.^[11] High group morale is hard to produce and impossible to guarantee in large organizations, and

[10] Pelzel asserts this view when he writes, "I argue that no attention is in fact paid to what may be called the social personality of the individual member. It is the emotional aspects of the personality that are instead given outlet in the procedures of the Japanese small group, and a wealth of standard behavior patterns can only be interpreted in this light." See Pelzel, "Japanese Kinship," p. 247. Nakane's emphasis on the emotional quality of group relations is also a closely allied perspective. See Nakane, Japanese Society . In Japanese the central concept and value of wa (translated as "harmony" but better grasped as teamwork, high morale, and general group vitality) emphasizes this same quality, particularly as it is coincidentally the group's internal ambition and the means to accomplish its external goals.

[11] I have in mind the factors of seniority and promotion based on technical skill in particular.

this fact has encouraged Japanese companies to establish supplementary efforts of a different kind. Increased emphasis on ability in promotions, on character-building programs, and on penalties for poor performance are all aimed at the individual rather than the group.^[12] These attempts, although minor, are indications of management impatience with the inadequacies of the group-motivation approach.

Modern Obstacles to the Work-Group Ideal

The effects of large-scale organization and modern life styles on the effort to establish satisfactory work-group relations can be pointed out, but to measure their impact requires historical and comparative frameworks beyond the scope of this paper. Among the most significant organizational constraints to note are the absence of any work-group-level control over recruitment and promotion.^[13] Transfers may be initiated by problems in office or factory groups, but no group leader has much power over who will be assigned to his unit. The personnel department in the manufacturing firm does on occasion attempt to place factory workers with an eye to special connections (such as kinship, regional affiliation, and personal obligation), but this practice is very rare for white-collar work groups and, in fact, requests by section leaders for individuals are generally discouraged. Since promotion is presently governed by seniority considerations, some are given leadership positions for which they are unsuited. The fact that the rank of section leader is a meaningful career goal, a reward for service, and a mark of status complicates the problem of establishing greater coherence between the work group and the promotion system.

Personnel turnover is another source of work-group problems. Transfers are greatest in the bank with its many branches, less frequent for office workers in the manufacturing firm, and almost negligible (except on a voluntary basis) for blue-collar workers.^[14] Although the men in the bank

[12] On promotions and penalties, see M. Y. Yoshino, Japan's Managerial System (Cambridge, Mass.: MIT Press, 1968), pp. 225–253; and Rohlen, For Harmony and Strength . For a discussion of a company character-building program, see Thomas P. Rohlen, "Spiritual Education in a Japanese Bank," American Anthropologist 35.5 (1973): 1542–1562.

[13] The account of the role of foreman labor bosses in early factories offered in Mikio Sumiya, Social Impact of Industrialization in Japan (Tokyo: Japanese Commission for UNESCO, 1963), describes them as having the power to recruit for and promote within their own work group. The control of these functions by management came after World War I. It would appear that one legacy of this early work-group autonomy is the continuing view that someone at the "foreman" level is more the workers' than the management's man. The plant engineer's general reluctance to interfere with production groups might also derive partly from these early circumstances.

[14] The plastics firm has recently opened a new factory in a rural prefecture, and the lack of private housing suitable for city families coupled with the long commuting time from suitable housing has made the move to a rural area rather unpopular, despite the much improved facilities, lovely scenery, and lack of pollution. Bachelor men at the new plant, housed in dormitories, complain of the absence of city amusements. All of this is instructive in the light of the Tanaka government's plan to move industry to the hinterlands as a solution to the devastating urban pollution and congestion.

retain a long association with their company, their participation in any particular office group is interrupted by periodic transfer. We can estimate an annual turnover in any office of 15 percent, and the bank's personnel department offers the additional information that the average length of time spent in an office is 3.75 years. Transfers involving women seldom occur and only when it is possible for them to commute from home to a different office. In the manufacturing firm, transfer rates in general are much lower, with only salesmen being shifted at approximately the same rate as men in the bank.

Thus continuity, so characteristic of a man's relationship with his company, is not characteristic of office work groups, an interesting point regarding the nature of Japanese organization.

Office groups are also seriously affected by the high turnover rate of women who leave to marry. One in four (in both companies) leave each year for this reason. For male blue-collar groups, quitting and mid-career entrance (a few each year among three hundred factory workers) are the major sources of change. In one factory the quitting rate for young people in their first five years of work was almost 50 percent and in the other 30 percent. We must conclude that the small work group in modern companies does not resemble the traditional family, the village, or the small commercial enterprise in this respect. The ideals and procedures may be the same, but the depth of acquaintanceship and involvement are certainly affected by the greater turnover. The intense interest in creating good group relations shown by the bank is partly explained by its high rate of transfer, just as the manufacturing company's emphasis on youth-oriented leisure activities is best explained as a response to the high turnover rate among its new, young workers.

The size of office and factory groups is not determined by considerations of group dynamics, but by the work load and other external factors. In consequence, many groups are too big to be organized on a small-group model. Attempts at subgroup arrangements are seldom satisfactory since the lines of cooperation in work cannot be easily aligned with boundaries established by informal involvement.

Private living patterns also have considerable impact on work-group relations. Dormitories, for example, draw residence and work together, whereas dispersed private housing tends to reduce the time co-workers interact, particularly if commuting time is great.^[15] The worker's degree of involvement with social ties outside the company (such as family, neighborhood, and voluntary associations), as well as the strength of his

[15] For descriptions of company dormitories, see Rohlen, For Harmony and Strength , and Cole, Japanese Blue Collar . For descriptions of the living patterns of married white-collar workers, see Ezra F. Vogel, Japan's New Middle Class (Berkeley and Los Angeles: University of California Press, 1965); Rohlen, For Harmony and Strength ; and Christie W. Kiefer, "Personality and Social Change in a Japanese Danchi " (Ph.D. dissertation, University of California, Berkeley, 1970).

desire for private as opposed to company-oriented leisure, weigh in the final balance of after-work time allocation, and this in turn has much to do with the capacity of work groups to hold lengthy discussions, enjoy parties and otherwise maintain close personal involvement.

Both companies, while not intending to increase family-type company housing, still emphasize dormitories for unmarried employees and have recently built new ones. Dormitory life is a major source of friendship for many young people (only for men in the bank, but for both men and women in the factories). Wherever dormitory social life is dormant, there is the complaint that modern salaries have provided luxury which in turn has caused young people to separate socially. Individual rooms, private stereos and TVs, the independence of a private automobile, and other similar factors all appear to have reduced the inclination or opportunity for communal forms of leisure. Dating, clearly on the increase, also has this effect. Where dormitory life is active, on the other hand, young people prefer it to socializing with older co-workers. Activities popular with the young such as sports, camping, hiking, bowling, and weekend travel are all forms of relaxation in which older people have difficulty participating. There is, in short, a trend toward more age-segmented after-work leisure, and while this is often encouraged by management because it satisfies the young, it also points to the widening gap between the ages in work-group situations. The bank has gone further than simply recognizing the trend. It has recently created formal associations for its unmarried men and women.

We need to remind ourselves that although Japanese company people are remarkably company oriented by contemporary Western standards, modern living is eroding this involvement at what managers see (using the criteria of the Japanese work-group tradition and the present ideal) as an alarming rate. In the two companies studied, for example, it was no longer shocking to hear of a group leader so much absorbed in his own private affairs that his after-hours socializing with co-workers had become minimal. For workers, marriage and family, while by no means Western in character, have increased in significance, particularly with the growth of family-oriented leisure. Furthermore, people with very long commuting times, new homes or an unusual interest in outside affairs or hobbies—matters that are regarded as modern trends—find it hard to give time to work-group relations. Watching television, the most mundane of these, has for some become a nighttime preoccupation.

The separation of young and old is another common topic everywhere, and both ends of the spectrum comment on how hard it is to understand and talk frankly with the other. There is no better way of measuring the solidarity of a work group than to ask about the relationship among the generations. Only where the group is well knit and the leader personally involved does the age gap seem inconsequential. Such cases are certainly in the minority, particularly among factory workers, but they do illustrate the fact (and hope) that when successful the work group has considerable

potential to integrate young and old regardless of their very different experiences, attitudes, and life styles.

Perhaps the most debilitating of all on work-group relations is the general decrease in identification with and enthusiasm for company and work, something that apparently derives from both the scale of organization and the nature of contemporary public culture. While there are employees of all ages thoroughly involved in their work, there are many who mention as reasons for uninvolvement such things as the distance between themselves and top management, the lack of a sense of common purpose, the attractiveness of private forms of leisure, and their sense of a conflict of interest in company-employee relations.^[16] Alienation may not be nearly as strong in Japan as in America, but it is present and does affect the morale of work groups. There is no easy way to gauge its general increase, but the increased company-to-company movement of young blue-collar workers is a strong signal. On the other hand, many of the young people most expressive of a modern sense of alienation are likely to adopt a different, more accepting stand with age. Finally, it is well to remember that any general organizational arrangement can tolerate a good deal of dissatisfaction, particularly if it is rarely expressed openly as in the Japanese case.^[17]

Work-Group Dynamics

My case material indicates, then, that solidarity in work groups is something to be achieved and that, compared to the ideal, groups are likely to be viewed as imperfect. Basic understandings and values create a

[16] There are literally hundreds of opinion surveys that document the general shift in attitudes (particularly among the young) away from conscious preference for company and work-oriented lives. Kiefer reports that while overtly denying any special company loyalty, the behavior of white-collar workers in the apartment building (danchi ) he studied actually expressed deep involvement with their companies and work groups. See Kiefer, "Personality and Social Change," p. 194. Undoubtedly, at the individual level the number of potentially contradictory layers of opinion, feeling, and unconscious assumption are great enough to require a very cautious use of any single expression (for example, to a questionnaire, a TAT, a Rorschach or an interviewer's question) as representative of the respondents' actual or predictable behavior.

[17] Japan's recent economic contraction (late 1973 on) will greatly affect the trends discussed in the preceding paragraphs. Those companies that suffer the greatest setbacks will undoubtedly witness considerable strife and alienation as reorganization of personnel policies, lay-offs and the like occur, however the long-term effects of an economic retrenchment are likely to reinforce the importance of (1) an individual's tie to a particular company, and (2) the perception that company and employee interests are fundamentally allied. This will come through the general contraction of available jobs, the greater relative success of some companies, and the decrease in voluntary labor mobility, particularly among younger workers. All of these shifts will serve to arrest and perhaps even reverse some of the trends toward greater individual autonomy just discussed, depending, of course, on the duration of the economic problems. Furthermore, a decrease in youthful alienation and mobility will be accompanied by less company emphasis on youth-oriented activities and an increase in worker concern with good personal relations with their seniors and superiors. In other words, jobs and good companies will replace workers (especially young ones) as the scarce commodity.

dimension of concern that not only points to goals, but also highlights existing inadequacies. Acknowledging the various qualities of small groups to be matters of degree compared with a set of ideals, creates the basis for examining the dynamics of the situation in terms of shifting conditions within the group. When the sense of solidarity is down, for example, hierarchy is less agreeable and more apparent. The reverse is also true. Solidarity itself is essentially a subjective quality or mood that arises primarily from the emotional aspects of the group, particularly the relationship between the group leader and his subordinates. When things go well, when the leader is followed out of affection as well as because of his formal position, the internal structure of the group appears quite useful and proper. The three general factors of positive emotion toward the leader, solidarity, and structure are interrelated in essentially the following manner: (a) positive emotion and working solidarity are mutually beneficial, (b) structure as a means of ordering work may also contribute positively, but (c) when structure comes to stand for separation and difference, then it is viewed negatively. To prevent the ever-present hierarchy from becoming onerous the group has recourse to activities of the circle kind. The leader's style and energy are also crucial, particularly as he must be able to create a mood of respect and fellowship.

The most poignant source of stress in this situation is the requirements of work itself. Pelzel is perfectly justified in underlining the importance of the task orientation of Japanese groups.^[18] Without a common task, office and factory groups would have nothing to justify their existence, and their enjoyment of one another informally would seem frivolous. Yet, it is the demands for greater production, the problems of sharing the work, and the responsibility and fatigue of strenuous effort that most tax the emotional ties and the sense of mutuality within office and factory groups. This pressure falls particularly on the chief, for he faces the demands and criticisms of top management.

It is interesting to note the response of employees in the bank to the characterization of their organizational life in terms emphasizing their close personal relations and sense of unity. These, they say, are qualities a foreigner (using a cross-cultural perspective) might choose to emphasize, but such a description ignores the daily problems and inadequacies of Japanese organization. When the problems are discussed, however, comments from the same people imply that models of close relations and solidarity do exist. Outstanding groups are the ones with good working relations, and they provide standards for comparison. Cross-cultural perspectives, we may conclude, are significant precisely because they highlight a particular conceptual framework and a set of ideal models, and therefore provide a perspective of reality that will deal adequately with the thought-action dynamics in real problem-filled situations.

[18] Pelzel, "Japanese Kinship."

The question of authority is a central issue in any cross-cultural comparison of bureaucratic organizations. Recently Crozier, citing cultural differences as significant, has suggested how impersonal authority problems in French bureaucracy are closely related to various themes and patterns of French culture.^[19] Face-to-face dependency and warm primary group relations in work are difficult for the French, and the result is a high degree of impersonality. Crozier's general approach, one that views the response to authority as varying with such things as dependency, is of special interest in the light of the fact that Japanese work groups seem so different from the French.

The acceptance of dependency (with limits) is a definite part of successful work relationships in Japan, and there is much evidence that dependency is very often regarded as a positive aspect of social relations.^[20] But dependency does not automatically lead to the "acceptance of most arbitrary discretion," another of Crozier's preconditions for bureaucratic success.^[21] We have seen that work-group discussions and situations of informal intimacy are regular aspects of the program designed to maintain participation in the group and acceptance of its leader's prerogatives. Without such efforts to maintain a sense of unity, the group ideal of mutual interest would soon disappear. For the office leader, the ability to be arbitrary one moment depends greatly on his readiness to be responsive to the group the moment before and the moment after. This shifting of emphasis, which serves to preserve the acceptance of leadership and trust mentioned by Crozier, is supported by official company policies and is therefore institutionalized and expected to a large degree.

Although Japanese procedures indicate much the same authority/compliance complex discussed by Bernard (1938) and those after him, the actual practices of American and Japanese work groups are quite different.^[22] For the Japanese, the procedures of discussion and participation are institutionalized, office groups are far more sensitive to the process of inclusion, and their leaders are far more inherently equipped to manage

[19] Michel Crozier, The Bureaucratic Phenomenon (Chicago: University of Chicago Press, 1964).

[20] For discussions of dependency from a variety of perspectives, see Doi Takeo, Amae no kozo^[*] [The structure of dependency] (Tokyo: Kobundo^[*] , 1970); William Caudill, "Around-the-Clock Patient Care in Japanese Psychiatric Hospitals: The Role of the Tsukisoi," American Sociological Review 26.2 (1961): 204–214; William Caudill and David W. Plath, "Who Sleeps by Whom? Parent-Child Involvement in Urban Japanese Families," Psychiatry 29.4 (1966): 344–366; William Caudill and Helen Weinstein, "Maternal Care and Infant Behavior in Japan and America," Psychiatry 32.1 (1969): 12–43; Kimura Bin, Hito to hito no aida [Between person and person] (Tokyo: Kobundo, 1972); and Ezra F. Vogel and Suzanne H. Vogel, "Family Security, Personal Immaturity, and Emotional Health in a Japanese Family," Marriage and Family Living 23 (1961): 161–166.

[21] Crozier uses dependency in a less psychological and more organizational sense, making it necessary to discuss the acceptance of authority as the intersection of psychological predispositions and the basic arrangement of small-group interaction. See Crozier, The Bureaucratic Phenomenon .

[22] Chester I. Bernard, The Functions of the Executive (Cambridge, Mass.: Harvard University Press, 1938).

this form of direction.^[23] In fact, the terms for authority (ken'i and kenryoku ) are not used in everyday descriptions of group dynamics. Acceptance (nattoku ), participation (sanka ), resistance, and opposition (hantai ) are the key dimensions, and neither impersonal rules nor formal role definitions are of much significance in adjusting behavior from the negative to the positive sides of these dimensions. Instead, the leader's virtue, his concern for others, and the general esprit within the group are the most effective means to gain individual acceptance and participation. Involvement and trust, once established, do indeed permit considerable "arbitrary discretion." Pelzel attributes essentially the same qualities to Japanese house-hold groups. He writes:

The tolerance for authoritarianism is perhaps higher in the Japanese than in many Western or Chinese situations, but in Japan no head can expect well-motivated action on any decision that has not been deliberately accepted by members who have the right by interest or competence to be heard.^[24]

In Weber's terms, Japanese work groups might appear to be highly traditional, but in Crozier's they are remarkable examples of the reinforcement of bureaucratic requirements by the psychological predispositions and small-group patterns of the parent culture.

Consequences of Work Groups for the Larger Organization

The general nature of work groups is acknowledged by the treatment they receive from above. We have already noted the common practices of encouraging competition between groups and of judging section chiefs according to their group's performance with the result that rivalry is often strong among men of that rank. A team form of motivation is heavily relied on, even in cases where the lack of comparability makes intergroup competition impossible. This complex has the following general correspondences with the larger organizational setup:

1. The role of group leader has special significance: its status is high; it is crucial to most operations, particularly nonroutine ones; and careers are either made or destroyed primarily at this level.

2. Individual forms of reward, especially immediate ones, do not receive

[23] Our everyday understanding of authority assumes that (1) close personal involvement undermines the latitude to make hard decisions about personnel, and (2) familiarity drains respect and status from leadership positions. In Japanese companies the first is a problem, but the second is not. Since close personal involvement is virtually built into the work-group hierarchy, special means are necessary to maintain discipline. All crucial personnel decisions are made in the personnel department far removed from the work-group situation. Furthermore, group leaders often ask intermediaries (usually the senpai of troublesome or recalcitrant group members) to privately caution others. In other words, the power of the most intimate ties, usually those between senpai-kohai^[*] , are brought to bear on cases we would label disciplinary.

[24] Pelzel, "Japanese Kinship," p. 246.

strong emphasis by American standards in the salary and promotion systems.

3. It is difficult to organize activities between work groups without direction from a higher authority.

Work groups, because they are recognized as independent social entities, are allowed considerable autonomy in the Japanese company system.^[25] Experts are not utilized to solve operational problems for specific groups; this would be regarded as outside interference and would undermine morale and leadership. The bank does have an inspection team, which visits each branch once every few years, and engineers are called in to solve technical problems in the factory, but basically the work group has full latitude to handle its own daily affairs. If it does not do this well, a new leader is likely. Even central administration policies may not have a binding force on the conduct of groups. In the bank, for example, most offices still were ignoring the 5:30 P.M. quitting-time policy one year after it was made the official rule for Wednesdays. On such matters, work-group compliance is encouraged, but not forced. If the issue has great overall concern or less consequence to the internal conduct of the group, however, central authority can be thorough and swift. Management watches the performance of work groups with a keen eye, but it is reluctant to interfere unless quite necessary. Finally, the sense of domain and the expectation of discussion and consensus in small groups means that much of any effort to change work methods must be directed at gaining the voluntary participation of each group. The manufacturing firm, for example, instituted a "Zero Defect" program only after each factory section had reviewed and endorsed it. Similarly, the bank management frequently calls for branch-level discussion of company problems. New programs are often initiated at a ceremony attended by representatives of each office group.

Neither company has a written set of rules defining the decision-making procedures for any level. Obviously, the hierarchy of ranks and the establishment of central authority serve these purposes to a large degree; but even at the highest levels, group discussion and consensus are acknowledged, although their significance varies with the situation and the arrangement of power at the top.

Above the president in the formal organization of companies is the board of directors, a group that contains the company's top executives. They meet often and are ostensibly the small group that runs the company.^[26] Other meetings to discuss, to decide, or to recommend

[25] There are fascinating parallels between this autonomy and the autonomy from direct governmental control of villages in the Tokugawa period. See Kurt Steiner, Local Government in Japan (Palo Alto: Stanford University Press, 1965, pp. 9–18.

[26] The board of directors may or may not be the locus of power. In "one-man companies" the group is usually a rubber stamp for the policies of the president or chairman, and in many companies including those divided by factions, decision-making can be located inseveral places due to the impossibility of open discussion and consensus in a central group. In such a situation, the board of directors is a place for formal approval of plans already laid by separate elements within management.

characterize the intermediate levels of organization above the work group. Branch chiefs, section chiefs, and managing directors' groups meet frequently, usually on a weekly basis. These middle-level meetings rarely decide issues, but they are important for the purposes of creating general acceptance of policies made above, for recommending new policies, and for improving lateral coordination and cooperation. While I do not possess sufficient material to satisfactorily discuss the variable procedures actually involved in decision-making above the office-group level, it is evident that even at the top, many of the group procedures and many of the same alternatives in decision-making patterns that characterize low-level work groups characterize the interplay of leadership and group discussion within the realm of management.

A final way the large organization and the work group are related is through the influence of the ideals of small-group conduct on the manner of administering the entire company. The image of a company comprised of people dedicated to the same goals, personally involved, and united by interest and common spirit underlies company ideology. Executives, rather than consciously concocting such an image, tend to regard it as an assumed goal of any company and any group of employees, and numerous events, policies, and statements of regret about contrary tendencies all point to the underlying model of the company as a small group.

The fascination here is the obvious fact that any organization over a certain small size will find it exceptionally difficult to preserve a sense of personal, emotional connectedness that extends beyond the work group and other face-to-face relationships. It is precisely on this point that Nakane Chie's general characterization of company organization is weak, I think, for while she is correct about the way Japanese small groups should work, she does not clarify sufficiently the basic differences between the small group and the large organization. Nakane's approach does have the virtue of describing the nature of informal cliques and personal ties often in the background of high-level influence, but companies are far more rationalized and impersonal than her account implies. It is as if her primary focus never leaves the small group and thus factions, cliques, academic circles, and other small entities within larger organizational spheres dominate her analysis.

The total company situation is more complicated. While the nature of work groups is best approached from the question of their fundamental conceptualization, it is less likely that an ideology emphasizing small-group values will have the same significance for large organizations. Company leaders may be devoted to such values, but they are not in a position to unite the personnel of a company in the same immediate way

group leaders are, and the best they can accomplish is to make a success of the directors' and other high-level small groups. The effect of an ideology of small-group togetherness on management-policy decisions is impossible to gauge, but certainly one major result is the frequent expression of regret that the small-group ideals must be overlooked in the face of larger organizational realities. Companies have been able to elaborate a wide manner of activities of symbolic participation (such as ceremonies, gatherings of representatives, company-wide outings, and civic and charitable programs), but none of this will be sufficient to secure the sense of connectedness of individuals to the whole, a crucial matter to Japanese companies.^[27] This, it seems, can be accomplished in one or several of the following ways: through the value of a company's public image (size and status), through the opportunities for rapid advancement (growth), and through the satisfactions of company work (above all social relations). Work groups occupy a central position in the last consideration for they are the primary environment in which the individual and the organization are interrelated, with all that means for such matters as authority, motivation, and individual satisfaction.

[27] See Rohlen, For Harmony and Strength .