Organizational structure refers to social influence patterns and social rules which help determine crewmembers' typical reactions to one another. Organizational structure is external to the individual, and at a high level of abstraction. Structural factors constrain individual behavior. They encourage dependability in performance, and, according to classical theory, make it possible for one person to substitute for another without a major disruption in the functioning of the whole (Weber, 1947; Graen, 1976; Perrow, 1979).
Structure is reflected in the distribution of authority and tasks. Structure defines organizational positions and specifies their relationships to one another. Associated with each position is a social role; that is, a set of expectations that define certain behaviors as essential, certain behaviors as admissible, and certain behaviors as unacceptable.
 Power Structures
Power structures refer- to the social influence patterns that regulate individual conduct and have the potential of coordinating activities within and between groups. The multilevel hierarchical or pyramidal form of power structure is the most common form in large governmental and industrial organizations today. This is evident in the military or paramilitary command structures which have typified past space missions and which may carry forth to future missions as a matter of convenience, tradition, or choice. A number- of common assumptions support the perpetuation of this form: (1) there must be a strong advocate of the sponsoring agency's interests aboard; (2) only a single individual's decision could be made fast enough to cope with certain situations; (3) crewmembers would prefer this type of arrangement because they are used to functioning in hierarchies; and (4) the average crewmember will easily understand a form of organization that approximates one that he or she has frequently encountered on Earth. These assumptions, however, are open to question. Not all potential crewmembers will be used to functioning around the clock in formal hierarchical structures; autocratic decisions, although speedy, may be resisted by the group; and a form of organization that is readily understood does not necessarily mean that it will be readily accepted. Thus, after reviewing some of the implications of the multilevel hierarchical control model for future space missions, we will consider two alternatives: the community democracy model, and the labor relations model (Whyte, 1 967).
Multilevel hierarchical control model- Hierarchical control structures require that individuals holding positions at one hierarchical level be able to direct or influence the activities of individuals at lower hierarchical levels. Social power- refers to the maximum influence that person A is capable of exerting on person B. Multilevel hierarchical control structures may thus be conceptualized in terms of the differential distribution of social power such that per-sons at each level have greater power than do persons at successively lower levels.
Bases of social power - According to French and Raven (1960), there are several bases of social power. Each of these bases may be eroded under conditions of extended-duration spaceflight.
Legitimate power arises from social norms and internalized values which dictate that person A is entitled to influence person B  and that person B is obligated to accept influence attempts by person A if these attempts fall within a prescribed range of conditions. As usually conceived, this form of power is based on person A's and person B's mutual acceptance of the organizational norms and values. On a space mission, legitimate power may be maximized through (1) a careful specification of authority relations; (2) the creation of a strong value system within the sponsoring agency; and (3) the use of selection and training procedures which ensure that the individual astronauts subscribe to the organization's value system. Prolonged seperation from Earth, however, may undermine the expectations and prescriptions that provide a basis for legitimate power. As Haythorn (1970, p. 164) notes:
Uncovering the means for mitigating loss of legitimate power under conditions of extended- duration spaceflight is of primary importance. One type of research which could contribute to this area would focus on techniques for maintaining crew allegiance to the Earth-based authority and the formally designated Ieaders. For example, social rituals, tangible reminders of allegiance, and symbols of rank might accomplish this. Other research might examine the feasibility and consequences of establishing Iegitimacy through means other than appointment by the Earth-based authority. For example, if legitimacy were controlled by the crew, legitimate social power would flow from proximal rather than distal sources, a change which could circumvent some of the problems identified by Haythorn (1970). Indeed, some findings suggest that leaders with internally conferred legitimacy have an advantage over Ieaders with externally confer-red Iegitimacy. Comparisons of elected and appointed leaders suggest that the former are perceived as more influential and more qualified and are Iess likely to promptly lose influence following failure (Hollander and Julian, 1970, 1978). Thus, it will no doubt prove useful to select future spaceflight managers from among the  ranks of former crews, perhaps even through the nomination and/or vote of the crew itself. A third possibility for future research involves systems under which legitimacy flows from both internal and external sources.
Reward power and coercive power refer to person A's abilities to affect person B's outcome or level of satisfaction. Such power is based upon the control of rewards (such as pay increases, recognition, and advancement through the hierarchy) and punishments (such as fines, censure, and termination). Although on-board managers can be invested with reward and coercive power, these forms of social power may also be undermined as Earth becomes remote. For example, an augmented or docked paycheck may lose significance in a microsociety where regular currency is not used, and it is difficult to "fire" someone who cannot be replaced and who has no place to retreat The problem is thus one of finding incentives and disincentives that are effective under conditions of isolation, confinement, and risk, and ensuring that these remain under the control of legitimate occupants of leadership roles.
Expert power flows from competence. Person A is said to have expert power over person B to the extent that person B perceives person A as an exclusive possessor of superior knowledge or skill. Leonov and Lebedev (1975) believe that astronauts must have total confidence in their leaders" expertise. To help achieve this, astronauts can be chosen for important positions on the basis of their expertise, or invested with expert power by means of special training programs. In addition, steps can be taken to help experts maintain and enhance their skills under conditions of extended-duration spaceflight. For example, computer simulations of systems dysfunctions can allow experts to exercise their problem-solving skills, and telecommunicated courses and teleconferenced or on-board seminars can contribute to professional development.
Although expert power is considered fairly robust ( Katz and Kahn, 1978), it, too, may be eroded under conditions of extended-duration spaceflight. First, a certain amount of cross- training may be necessary to prepare back-up personnel. Since expert power implies an exclusive access to knowledge, cross-training will serve to dilute it. Second, McCarthy (1979) hypothesizes that leaders risk loss of expert power whenever they make foolish statements, even when such statements regard matters that fall outside of their acknowledged area of expertise. Presumably, the risk of a display of ignorance is increased in isolation and confinement, conditions which  may promote protracted superior-subordinate discussions on farranging topics. Self- imposed isolation on the Ieader's part may be too terrible a price for the leader to pay. Thus, a major problem is one of finding ways for leaders to maintain credibility without remaining highly isolated from the rest of the crew.
Additionally, there is referent power, or power that is based on liking or identification with another person. A prerequisite for referent power is that the leader have "membership character" in the group; that is, he or she must be perceived by the other astronauts as "one of us." Crew leaders can be chosen on the basis of their appeal, and training procedures may increase their attractiveness. However, as missions progress, leaders may have to take unpopular actions that would reduce their overall level of attractiveness. Furthermore, in accordance with the reinforcement model of interpersonal attraction (Byrne, 1971), it is hypothesized that the negative affect generated by isolation, confinement, and risk may become associated with the leader, again to the detriment of the leader's referent power. Once again, then, there is a potential loss of social power under conditions of extended-duration spaceflight. In this case we must look for ways in which leaders can inspire continued affection from the crew under conditions of isolation and confinement.
In summary, legitimate power, reward and coercive power, expert power, and referent power make it possible for individuals at one level of the organization to coordinate and direct the activities of individuals at a lower level of the organization. Our research task is to understand the causes and the rates of erosion of each form of power under conditions of extended-duration spaceflight, and to identify those techniques which could reverse or retard such erosion.
Centralization of authority- Authority is centralized to the extent that decisionmaking power is concentrated within the hands of one or a limited number of individuals. In a highly centralized authority structure, managers retain decisionmaking power and expect subordinates to operate within relatively narrow latitudes. In a highly decentralized authority structure, managers delegate power downward to lower-level managers and expect subordinates to demonstrate initiative and independence.
The optimal degree of centralization of authority depends upon a complex array of interrelated factors (Wexley and Yukl, 1977). For the most part, centralization is desirable to the extent that (1) there were good communications with subunits; (2) higher-level managers  have access to staff specialists or other information sources which are not readily available to lower-level managers; (3) higher--level managers have the same knowledge of local conditions as do lower-level manager-s; and (4) decision speed is only a minor consideration.
As missions become more remote from Earth, involve more people, and last longer, decreased centralization may gain the advantage Under tomorrow's spaceflight conditions, communication with Earth may become erratic or delayed; the crew is likely to contain an increasing number of staff specialists; and the crew commander's knowledge of local conditions will surpass that of Earth-based personnel located hundreds of thousands or- millions of miles away. In addition to promising efficiency under tomorrow's spaceflight conditions, decentralization promises two long-range advantages (Perrow, 1979). First, it will free the time of higher-level leaders on Earth to deal with long-range issues. Second, it will help train or prepare lower-level leaders aboard the spacecraft for higher-level responsibilities. The alternative to decentralization, maintaining a high degree of central control, requires a solution to communication problems and ensurance that common or superordinate goals will be maintained by both flight and ground personnel.
Hierarchical positions and leadership tasks - To some extent, organizational charts and rule books mate it possible for selected individuals to lead. But structural factors themselves cannot ensure an optimal level of organizational functioning (Katz and Kahn, 1978). Such factors must be supplemented by the exercise of judgment and the application of human-relations skills. By themselves, organizational charts and rule books do not make adequate allowance for (1) the juncture between the organization and the (external) environment; (2) changing environmental conditions; (3) internal dynamics and development; and (4) the needs of participants to be treated as unique individuals.
On missions involving few people, there may be only one formal leader. On missions involving scores or hundreds of people, there may be many leaders ranging from low-level supervisors to the person in overall command. The extended discussion of small-group leadership presented in our group dynamics chapter is certainly pertinent to leadership within the organizational setting. But, suggest Katz and Kahn (1978), the demands placed upon Ieaders change as a function of hierarchical level. Low-level supervisors, middle management, and top- level Ieaders require different intellectual and human-relations skilIs.
 Lower-level Ieaders work within the framework of preexisting policies and procedures. The intellectual requirements are technical skills (for example, those regarding equipment use or organizational procedures) and the human-relations requirements (concern with equity and fairness when rewarding and sanctioning subordinates).
Intermediate-level Ieaders supplement and piece out preexisting policies and procedures. The intellectual requirements are an understanding of the subsystem and a two-way perspective which encompasses superiors and subordinates. The human-relations requirement is an ability to integrate primary and secondary relationships; that is, to deal with people as unique personalities and as role occupants.
Higher-level leaders create policies and procedures. The intellectual requirement is an understanding of the system as a whole and its relationship to interfacing systems. The human- relations requirement is charisma; that is, a magical aura or halo which is derived from followers' emotional needs, and from bold and imaginative acts of Ieadership.
Competent leadership at the top Ievel will become increasingly important as missions encompass ever-increasing spans of time. The authority structures, roles, norms, and other organizational variables which are set prior to mission departure may or may not fully withstand the passage of time. Although dramatic events could force reorganization shortly after a mission's departure, and space settlements could preserve traditions for generations, we anticipate that the status quo will be preserved on relatively brief missions, but that organizational change will be likely on missions which involve extended periods of time. Future top-level leaders would thus do well to master the techniques of organizational development; that is, planned organizational change (French and Bell, 1978).
Isolation, confinement, and risk are expected to increase the demands on leaders at all levels. As mission planners and managers undertake Ieadership selection and training, consideration might be given to the intellectual and human-relations requirements of each leadership level. For example, with respect to selection, one assessment device might examine candidates' tendencies to adopt individual, subsystem, or systems perspectives. Such a test might consist of case histories that can be interpreted at different levels.
Another research issue involves the effects of physical remoteness on the performance of top-, inter-mediate-, and low-level Ieaders.  Leadership level and communications media may have interactive effects on leadership effectiveness. For example, a physically remote low-level leader, who communicates with subordinates via teletype, may find it difficult to convey a high level of social support to subordinates, and may hence find it difficult to gain their cooperation. Alternatively, a physically remote high-level leader who is rarely or never actually seen by subordinates may find it relatively easy to preserve charisma.
Community democracy and labor relations models- According to Whyte (1967), certain organizational models, such as the multilevel hierarchical or bureaucratic model, are so common that it becomes difficult or impossible to envision alternatives. Yet no single model will prove optimal for all purposes. The bureaucracy has many strengths (Perrow, 1979), but it also has many weaknesses. For example, in many multilevel hierarchies, downward communication is impaired because leaders at each level feel compelled to pass messages through subordinates rather than directly to those for whom the message is intended. Upward communication is impaired, not only because such messages also pass through many levels, but because they may be distorted to achieve compatibility with subordinates' perceptions of their superiors' desires. Within bureaucracies, decisions may be seen as imposed from above, and hence fail to be endorsed or implemented by the organization's rank and file. In space there will be another problem: each crewmember will be so expensive to maintain that extremely careful consideration will have to precede the addition of yet another bureaucratic level.
Perhaps the best-known alternative to the bureaucratic model is the community democracy model (Whyte, 1967). According to this model, each member of an organization participates in the decisionmaking process. Participation may be organized along one-person- one-vote or discussion-to-consensus lines. Under this model, authority is deemphasized and negative sanctions may be applied to people who fail to make appropriate inputs. This model rarely prevails in truly large social units, but is closely approximated in the New England town meeting (Whyte, 1967).
The community democracy model is consistent with American democratic ideology, and the decisions that result, at least when reached through consensus, are likely to be accepted and implemented by the membership. However, the community democracy model is difficult to apply when the group is large and when it is necessary to reach prompt decisions. Not every member of a large  group can make informed and useful inputs; not every member of a large group cares to be consulted on each and every issue.
The community democracy model has particular disadvantages when it is bogus or only partially applied. Sometimes, for example, the group is allowed to reach minor decisions but not allowed to vote on "hot" or consequential issues. Sometimes, the rank and file is openly channeled toward a certain decision, or discovers later that its efforts have been negated or reversed by a higher authority. Such misapplication of the community democracy model wastes time and energy, invites disillusionment and cynicism, and undermines morale.
An alternative to the bureaucratic and community democracy models is the labor relations model (Whyte, 1967) which openly combines both authority and participation. One group, the managers, has the power to make and enforce decisions. The other group, the rank and file, has the right to suggest changes, influence decisions, and protest decisions that have already been made. The managers and the rank and file take each other seriously, and bargaining and negotiation surround the decisionmaking process. On the one hand, this model acknowledges conflicts of interest and, according to Whyte, makes it possible for such conflicts to be dealt with in realistic ways. For example, in the labor relations model the rejection of the rank and file's demands invites protest, but not disillusionment or cynicism. On the other hand, there is always the risk that the conflicts inherent in this form of organization will exceed acceptable limits.
Each model has some advantages and disadvantages. The point is that no one model is necessarily the best for all future missions, and that rather than relying on the best-known models, planners and managers would do well to carefully evaluate both familiar and novel alternatives. The crew size and the nature of the crew (e.g., military or civilian) are among the many variables that need to be taken into account.
Kanas and Fedderson (1971) observe that four types of task or work roles would seem to be important on missions of any appreciable size. These roles are here labeled flight- operations roles, scientific-investigative roles, environmental-support roles, and personnel- support roles. To this list should be added a fifth type of role, the production role, which will become important on certain types of missions. On small missions, we have seen "doubling up,'  when a given individual performs more than one work role. On future missions, expanded crew size may give rise to the option of having many crew members able to perform each type of role.
Flight-operations roles involve command, navigation, flight engineering, systems monitoring, and telecommunications. Historically the fir-et to develop, such roles remain essential to any space mission. Thus far, commanders have been drawn from among flight operators, but this might not always be the case (Helmreich et al.,1980). That is, during these early years of spaceflight, all other considerations are insignificant when compared with the problem of getting there. In the future, "getting there" may become routine and top leaders may require skills which are useful to attain other mission objectives. Thus, the top leader of an orbiting work station may be a professional manager rather than a pilot.
Scientific-investigative roles involve research tasks. People who perform such roles are expected to generate new data that have relevance beyond the immediate flight. Although perhaps not essential for any one mission, in the aggregate scientific-investigative roles help justify space exploration.
Environmental-support roles involve the management of supplies and the maintenance of facilities. Environmental-support roles are essential for a mission, and on small missions they are easily combined with flight-operations functions. On truly large missions, people who perform such functions might range from the equivalent of a quartermaster to crewmembers who perform very routine facilities-maintenance tasks.
Personnel-support roles include promoting physical and mental health, attending to the psychological needs of crewmembers and attempting to maintain the morale of the crow as a whole. Astronauts who fill such roles will serve what Katz and Kahn (1978) refer to as the maintenance function. On small missions, personnel-support requirements may be met by a physician who is well trained in interpersonal relations (Kubis, 1972). Large missions may involve an elaborate personnel-support system that includes physicians, psychologists, and personnel administrators.
Production roles will be prominent on missions that pursue explicit industrial or economic goals. Included in this category would be roles r-elated to the manufacture of products in space and the exploitation of energy or other resources. On certain future missions,  the proportion of crewmembers who per-form production functions may be larger than the proportion of people who perform all other functions combined.
Most writers assume that crewmembers should perform well-planned and well-learned work roles (Berry, 1973; Helmreich, et al., 1979a; Helmreich et al., 1980; Kubis, 1972; Leonov and Lebedev, 1975; Natani and Shurley, 1974; W. Smith, 1966). Expressing the views of many authors, Berry (1973a, p. 1142) notes:
Among the continuing concerns of mission planners and managers are (1) defining articulated task roles which collectively satisfy the mission's behavioral requirements, and (2) ensuring that each crewmember learns and performs his or her role. Although the careful specification of roles is often a necessary and desirable goal, there are reasons to hypothesize that this goal should be pursued with some moderation. Succumbing to the temptation to prescribe roles to the smallest detail could have at Ieast three adverse consequences. First, the sharp definition of roles can discourage functional as well as dysfunctional out-of-role behaviors. (For example, Leonov and Lebedev (1975) report that two Antarctic explorers refused to help put out a fire because it "wasn't part of their job.") Second, stringent behavioral prescriptions for individuals can generate psychological reactance; that is, threaten the person's sense of freedom and autonomy and provoke grumbling and defiance (Brehm, 1966) Third, the minimum Ievel of performance that is prescribed tends to define the maximum Ievel of performance that is obtained (Katz and Kahn,1978). Further definition is required to identify the optimal degree of role specification over a range of space missions.
Prescribed and emergent roles- Roles may be imposed by higher authorities (prescribed roles) or they may arise Iess formally in the course of social interaction (emergent roles). Emergent roles typically complement and supplement prescribed roles, but occasionally the two will conflict. Four conditions encourage emergent  roles (Katz and Kahn, 1978); the first two of these conditions will be particularly pronounced in space.
First, prescribed roles may not make adequate allowance for people to satisfy their personal needs. For example, well-being requires some behavioral variety. Rigid role prescriptions limit behavioral options. Conditions of isolation and confinement can intensify this problem, since, under any circumstances, crewmembers will have a limited opportunity to enjoy social variety.
Second, prescribed roles are devised on the basis of the anticipated and known. They are likely to prove inadequate in light of the unanticipated or unknown. For example, a polar expedition described by W. Smith (1966) was intended to have highly defined work roles, with each person assigned specific duties such as driving or taking geological measurements. However, an icy blast through the SnoCat's floorboard made sustained driving unbearable, and the geological measurements were extremely difficult for one person to obtain. This led to a breakdown of preassigned duties. Men took turns driving, and offered the geologist assistance. Because spacecrews, like other adventurers, are likely to encounter the unanticipated or unknown, emergent roles become likely.
Third, roles are prescribed on the basis of conditions that are in effect at a given point in time. A change in conditions may make certain roles obsolete, force the rapid development of new roles, or require one person to perform another's functions. For example, the death or disablement of a crewmember could force another crewmember to assume the deceased crewmember's duties. To cover such contingencies, some degree of redundancy must be associated with key roles.
Finally, prescribed roles cannot fully cover even routine situations (Perrow, 1979) For example, crewmembers may discover that they have to respond to gradations and distinctions which the official plan ignores. To illustrate, a flight-operations officer and a personnel officer may enjoy the same official status or rank. Yet the latter may enjoy greater informal influence if he or she is responsible for performance evaluation or for the assignment of duties.
Defining space crew roles, then, requires a sensitivity to socioemotional as well as to task requirements; the inclusion of options and alternatives to cover unanticipated, unknown, and changing conditions; and an acceptance of the fact that not every contingency can  be fully covered. Useful in this regard would be procedures for identifying those situations for which predefined roles are necessary, those situations for which predefined roles are desirable, and those situations which need not be covered by predefined roles. Another important research topic is identifying the conditions under which emergent roles are likely to complement and supplement (rather than undermine) prescribed roles. We hypothesize that such incompatibilities are likely to be minimized when (1) selection and training yield a crew with a high degree of commitment to official structures, prescribed roles, and overall organizational goals; (2) prescribed roles are flexible in the sense that they can be adjusted in response to changing conditions; and (3) prescribed roles provide latitude for discretionary activities and the development of voluntary socioemotional relationships.
Role overload- Role overload exists when a position carries a combination of tasks that are sufficiently burdensome that they cannot be competently performed without undue stress. Attempts to adjust to overload, such as omitting certain tasks or back-logging tasks, result in inefficiency and a lowering of performance standards; in addition, overload increases wear and tear on the individual performing that role (J. G. Miller, 1960, 1978). Thus, care must be taken to ensure that individual crewmembers are not overloaded.
Combinations of roles, in particular, may prove especially burdensome for individual crewmembers because the extent of the duties associated with one role may not be fully taken into account when the other role is formulated. Managers may underplay the burdens of each role in an attempt to induce crewmembers to accept the combination. Thus, crewmembers who accept multiple roles may not be fully aware of the demands that each role will place on their energy and time. If aware of potential overburdening, crewmembers may nonetheless accept multiple roles because of a high need to achieve, a "can do" self-image, real or imagined long- term career considerations, or a lack of assertiveness to decline. Managers may be sorely tempted to assign heavy combinations of duties to each crewmember, because weighted against the addition of new crewmembers are engineering considerations and immense costs.
Overload conditions appeared in the course of Spacelab Mission Development Test III (SMD III). This project was a ground-based simulation of a series of life sciences experiments conducted by Ames Research Center and Johnson Space Center-. Personnel from Ames Research Center planned and prepared experiments which  were then delivered to Johnson Space Center, where they were carried out in the context of a 7-day spaceflight simulation (Helmreich et al.,1979a).
For many of the people involved, the addition of heavy SMD III requirements to normal responsibilities resulted in an onerous workload. First, commitment to SMD III made it difficult to fulfill administrative and other routine obligations at the two Centers. Second, principal investigators often felt frustrated, for their commitment to SMD III made it difficult to carry on with their usual activities. Third, administrative duties made it particularly difficult to complete high-quality scientific research; that is, the combination of managerial and scientific-investigative roles proved particularly burdensome (Helmreich et al.,1979a).
Certain procedures may help minimize the problems associated with role overload. For example, Helmreich et al. (1979a) suggest that some of the difficulties that managers and workers encountered in the course of SMD III could have been reduced by (1) ensuring that both had a clear understanding of the extent of their multiple obligations; (2) ensuring that both set appropriate priorities among multiple obligations; (3) improving communication among occupants of adjacent roles; and (4) minimizing people's exposure to irrelevant communications. Despite such progress, important topics for future research include more definitive specification of the conditions that lead to, and ameliorate, role overload during the course of a space mission.
Role-related conflicts- A certain amount of conflict may be expected among occupants of different spacecrew roles. One potential form of conflict, that of role conflict, is associated with expectations and demands centering around the roles themselves. Such conflicts occur when the person's expectations are inconsistent with the expectations of people in related roles, the person is subjected to inconsistent demands from people in different interlocking roles, and the person is assigned two or more roles which prescribe incompatible behaviors.
Role conflict causes frustration, vacillation, and deteriorating performance (Wexley and Yukl, 1977). Although it is doubtful that role conflict can ever be eliminated, careful analysis and planning followed by a high degree of training should help reduce the potential for- role conflict. Direct tuition, observational learning, and role-playing are among the techniques that are available for ensuring that  crewmembers understand the behaviors expected of them and maintain appropriate expectations concerning each other's behavior.
Another potential source of conflict is the different perspectives of the people who are likely to occupy the different roles. Systematic differences in the personal backgrounds and hence the values and interests of different role occupants may generate on-board frictions (Doll and Gunderson, 1971; Natani and Shurley, 1974). Although the conflicts usually involve minor issues, such as musical selections (Doll and Gunderson, 1971), they may be blown out of proportion as the term of isolation and confinement continues. Perhaps the most serious incident recorded involved sailors on an Antarctic research vessel who threw overboard the scientists' 2-yr collection of biological specimens. The sailors did this to make more room for beverages in the ship's freezer (Bluth, 1981). It is important to learn more about how people with noncomplementary personalities and interests gravitate toward interfacing roles. Remedies may be sought in personnel selection or in role redefinition.
The fact that some roles are more important than others also has the potential for generating conflicts. For example, some of the military test pilots who initiated the conquest of space have expressed nonsupportive attitudes toward the "hyphenated astronauts" who joined to play scientific and other nonpiloting roles (Cunningham and Herskowitz, 1977; Wolfe, 1979). In the future, environmental support roles may be assigned low status by other crewmembers, and line officers who are expected to risk their lives in extravehicular activities might view staff officers as second-class citizens. We might expect crewmembers to see colleagues who have the less important jobs as not fully carrying their own weight, and, over time, the colleagues may lose their self-esteem. Certainly, one task is to identify and establish those conditions under which each crewmember recognizes the other crewmembers' importance. Perhaps the critical variable here is perceived fairness; that is, a conviction that everyone is carrying his or her weight and receives rewards that are appropriate, given the level of his or her contributions (Adams, 1965; Walster, Berscheid, and Walster, 1973; Leventhal, 1976) Planners and managers need to know more about perceived fairness and equity under conditions of isolation, confinement, and risk.
Yet another threat is that people within a role category will form factions or cliques and mark boundaries that discourage constructive interaction with members of other cliques or with  nonaffiliated individuals. Factions may show a certain amount of prejudice toward one another, or attempt to reach special-interest goals that are incompatible with mission goals. For example, scientific-investigative personnel might argue in favor of a dangerous, but curiosity-satisfying, change in course. Thus, another research task is to find means to cope with the dangers of factionalism. Techniques are needed to detect and combat prejudice, promote social interaction and communication, and encourage far-ranging friendship networks.
Allowing crewmembers to (sequentially) perform many different roles may discourage the formation of cliques along work lines and offer other benefits as well. Shurley et al. (1977) suggest that role rotation and personnel exchange can (1) al low crewmembers to gain other crewmembers' perspectives; (2) foster tolerance for behaviors that are perceived as idiosyncratic, but which are actually due to structural or role variables; (3) encourage mutual problem-solving; and (4) help meet some of the needs frustrated by the abandonment of everyday roles. To this roster of benefits we add that role rotation and personnel exchange should (5) promote a breakdown of invidious status distinctions; (6) underscore the importance of each role for the integrity of the overall mission; and (7) provide welcome variability in a relatively unchanging environment.
Role rotation, of course, is not a panacea. Many roles require a high degree of knowledge and skill. It is unrealistic to expect individuals to be competent in several highly technical areas. Moreover, unless carefully controlled, role rotation could contribute to role overload or to role conflict. For example, the expectations which persist from yesterday's role may make it difficult to perform today's role.
In summary, many of the most important research and planning issues surrounding tomorrow's space missions have to do with defining and articulating crewmembers' roles. Although it is tempting to specify roles in great detail, an extremely high degree of specification can yield dysfunctional consequences. Emergent roles will accompany prescribed roles; an important research topic is identifying the conditions under which the former will complement and supplement, rather than conflict with, the latter. Further research and planning issues include preventing role overload, minimizing role conflict, minimizing conflict among the interests and needs of people who are drawn to certain types of roles, fostering the perception that  different roles have appropriate balances of obligations and rewards, and preventing the rise of role-related factions or cliques.
Social norms or mutually shared expectations about attitudes and behaviors are another structural determinant of individual conduct and interpersonal coordination (Hackman, 1976). In addition to prescribing appropriate ways of behaving, norms prescribe means for dealing with individuals whose methods or levels of performance deviate to some degree from those considered acceptable by the group. Normative structure refers to the entire set of norms that is in use by a group or an organization. Normative structures do not provide a framework for all potential behaviors; only those that the group considers relevant to its functions and purposes (Hackman, 1976). In most settings, people who work together do not also live and play together, with the result that the norms tend to be limited to job-related activities. In spaceflight and other isolated and confined settings, the normative structure is likely to encompass not only those activities that are directly related to the job, but to living and recreational activities as well.
On American spaceflights, the crew's normative structure will in part be derived from our society's norms, in part derived from NASA's norms, and in part derived in the course of interaction within the crew. In some cases the attitudes and behaviors encouraged by society or by NASA may be preempted by norms that emerge in the course of interaction within the group. Typically, such emergent norms support lower levels of accomplishment than those which are officially prescribed, and support ways of proceeding which ignore or perhaps violate the official rules. From the outsider's (or mission control's) perspective, emergent norms are based less on standards of excellence than on personal convenience, and less on organizational objectives than on peer-group compatibility. From the "insider's" (or crewmember's) perspective, emergent norms reflect first-hand experience regarding what actually can be accomplished in a given situation, and a realistic concern for getting along with one's peers As in the case of prescribed and emergent roles, official and unofficial norms can complement, supplement, or conflict with one another.
A group that maintains an elaborate and clearly defined normative structure is expected to function smoothly. But norms that are too rigid or too strictly enforced can have dysfunctional  consequences. As noted at several points in our discussions of group and organizational variables, social structures require some flexibility to encourage innovative behaviors, allow social variety, and minimize the dangers of ostracism.
It is important to gain a better picture of spacecrew norms and their likely rationales. One concern is identifying dysfunctional norms that impose excessive restrictions on those behaviors that do not threaten social stability or detract from attaining mission objectives. Another concern is identifying conduct areas that appear to require social constraints but which the existing normative structure leaves uncovered. A third concern is identifying and dealing with conflicting norms that may be maintained by the different groups that participate in a mission.
In summary, spacecrews, like other groups, will develop shared expectations regarding appropriate attitudes and behaviors, and means to ensure that these expectations are confirmed. Because astronauts rest and play as well as work together, crew norms are likely to extend into many different areas of human activity. Some of the norms that emerge from within a spacecrew may conflict with the norms that are maintained by the mission sponsor. We see some value in mapping and monitoring spacecrew norms and devising means for changing normative structures when necessary.