part two CONFLICT THEORY AND ANALYTIC SOCIOLOGY Ralf Dahrendorf, Lewis Coser, and Randall Collins, the theorists discussed in this section, can be grouped as "analytic" conflict theorists because they share the belief that a conflict perspective is central to the development of an objective, or "scientific," sociology. However, they are different from critical theorists in three important respects. First, whereas critical theorists see social science as intrinsically a part of political action and deny that fact and value can or should be separated, analytic theorists consider such a separation to be essential. It may be difficult or impossible for analysts, particularly those concerned with human society, to frame their hypotheses independently of their own opinions and concerns, they argue: but those hypotheses have implications that can be observed and tested objectively with empirical measures. It follows that ideas are not automatically distortions of reality just because they are also the products of particular social circumstances or tend to favor the interests of a particular social group. Second, analytic conflict theorists do not analyze all societies as stratified along a single dimension, with a ruling group opposed to the masses. Analytic theorists would agree that some societies are of this type, but they believe that many others are far more complex in the way power and status are distributed; they have interlocking patterns of stratification, which do not line up neatly. This is true, they believe, because there are a number of different sources of power and position in society, and one particular set of institutions, such as that based on property, is not always paramount. Third, the analytic theorists do not contrast the present with a rational or conflict-free ideal for the simple reason that they do not believe in one. On the contrary, they emphasize that conflict and its roots are permanent, conflicts of interest inevitable. In all these respects, modern analytic theorists share Max Weber's approach,just as Marx was the primary influence on the critical approach. Weber believed in the vital importance of objective social science. As we saw earlier, he developed a typology of "class, status, and party" as important influences on people's lives, as opposed to Marx's emphasis on property classes alone. Furthermore, Weber considered that the conflicts these gen- erate are permanent features of human society, and he saw modern society tending not towards a communist Utopia but towards a bureaucratic society inimical to human freedom. This is not to say, however, that either he or the analytic conflict theorists he influenced are indifferent to political action. In general, conflict theory appeals to sociologists who hold strong political views; and a number of modern analytic theorists have been very involved in politics and social policy, often in direct opposition to the ideas of critical and Marxist theory. RALF DAHRENDORF Ralf Dahrendorf (b. 1929) is one of the few living European sociologists who is widely known and respected in both Europe and North America. As a teenager in Nazi Germany, he was sent to a concentration camp for his membership in a high-school group opposing the state, and he has continued to be deeply involved in political affairs. He has been a Free Democratic member of the Baden-Wiirtemburg Landtag (Regional Parliament) and West German "Bundestag" (Parliament); and as a member of the Commission of the European Communities, he has been responsible for external relations and for education, science, and research. The bulk of his academic career has been spent in German universities, and in 1984 he became Professor of Sociology at the University of Constance. However, he has also worked in both Britain and America and was Director of the London School of Economics, one of the most prestigious British institutions of higher education. Dahrendorf's work on conflict reveals two major concerns. The first is with what he himself describes as "theories of society"; 158 that is, with setting out the general principles of social explanation. Here, Dahrendorf stresses the primacy of power and the consequent inevitability of conflict. Like Marx, his second concern is with the determinants of active conflict the ways social institutions systematically generate groups with conflicting interests and the circumstances in which such groups will become organized and active. Power, Conflict, and Social Explanation There is, Dahrendorf argues, an inherent tendency to conflict in society. Those groups with power will pursue their interests, and those without power will pursue theirs: and their interests are necessarily different. Sooner or later, he argues-and in some systems the powerful may be very thoroughly entrenched-the balance between power and opposition shifts, and society changes. Thus, conflict is "the great creative force of human history."159 Power. According to Dahrendorf's theory of society, the distribution of power is the crucial determinant of social structure. His definition of power is Weber's: "the probability that one actor within a social relationship will be in a position to carry out his own will despite resistance, regardless of the basis on which this probability rests."160 In this view, the essence of power is the control of sanctions, which enables those who possess power to give orders and obtain what they want from the powerless. However, people dislike submission. Therefore, Dahrendorf argues, there is inevitably a conflict of interest and an impetus for the powerless to conflict with the powerful, the former in pursuit of power, the latter in defense of it. Power is a "lasting source of friction."161 This essentially coercive view of power, which is co mon to most conflict theorists, is very different from that of functionalism.162 As we have seen, Parsons believes that power is embodied in political institutions that solve the "functional imperative" of goal attainment.163 The ability power bestows to get what one wants at others' expense he regards as a "secondary and derived" phenomenon. Dahrendorfs view is the opposite. Power is necessary if large organizations are to achieve their goals; and at times, such as in a defensive war, the powerful may carry out very clearly the common aims of a group. However, what Parsons considers the secondary aspect of power, Dahrendorf considers primary: the powerful are not granted power by the community to carry out some "common will," but grasp and use that power for their own ends. Dahrendorf does not see the struggle for power as the'sum of social life, however. Weber's (and Dahrendorf's) definition of power includes actors "within a social relationship": in situations where other people's actions matter. But there are also times when people are free to do what they like without other people mattering at all. In his recent lectures on the cu I rrent political situation in the West, Dahrendorf has discussed the factors that endow societies with more or less "liberty" in this sense-with what has been called "negative freedom."164 In America, for example, you are "at liberty" to move from one city to another without anyone's permission. In China you are not; and whether you get the requisite permission and papers largely depends on your own "power" and influence. Norms . Like other conflict theorists, Dahrendorf argues that societal norms do not define or emerge from social consensus. Conflict theory, he argues, perceives, unlike functionalism, that norms "are established and maintained ... by power, and their substance may well be explained in terms of the interests of the powerful."165 This may be seen from the fact that norms are backed by sanctions. Vivid examples of what Dahrendorf means can be found in Soviet Russia, where dissidents risk prison camp or a mental hospital, or in the pre-Civil Rights South, where "uppity" blacks or nonconforming whites could lose their livelihoods or even their lives. In turn, sanctions involve the control and use of power, particularly the power of law and punishment.166 "In the last analysis, established norms are nothing but ruling norms," he suggests. 167 Social Stratification. Dahrendorf clearly distinguishes between two facts: first, that positions and jobs are different and demand different skills, and second, that different jobs are treated as "superior" or "inferior" to one another. There are both "social differentiation of positions ... and social stratification based on reputation and wealth and expressed in a rank order of social status."168 Social stratification is what makes college presidents more generally respected than bus drivers, and lies behind claims that teachers "ought" to be paid more than maintenance men. Stratification, Dahrendorf argues, is caused by norms,: which categorize some things as desirable and others as not. In every group, norms defining how people should behave entail discrimination against those who do not comply. During the Vietnam War, for example, those who supported the war were ostracized on some campuses, those who opposed it on others. Moreover, every society has general norms, which define certain characteristics as good (such as being an aristocrat or having more education than average) and which therefore entail discrimination against those who do not or cannot conform. These norms, Dahrendorf argues, are the basis of social stratification, and they are themselves derived from and upheld by power. Once again, therefore, power is the central concept. This is a very different explanation from that of functionalists, who argue that social stratification derives from society's need to attract talented people into important positions. The two may not be as totally irreconcilable as Dahrendorf implies, however. Dahrendorf does not explain how a group becomes powerful in the first place, but this will surely often depend, at least in part, on its offering skills and a type of social order that people value. Not all successful ruling groups are military invaders! It is precisely the existence of a relationship between power and general social values that functionalism addresses. Similarly, economics' statement that income differentials are a result of the market value of skills l69 has given rise to theories of stratification linking success with the provision of scarce services.170 However, a group that gains a hold on power in this way will almost certainly strive to maintain and take advantage of its position, convince everyone of its legitimacy and importance, and prevent competition from groups with different potential power bases-and Dahrendorf's approach is far better suited than functionalism's to analyzing this process. Conflict Theory and Explanation. Dahrendorf believes that a "conflict" approach, which combines general hypotheses about the nature of power and inevitability of conflict with information about the positions and resources of particular historical groups, is the most fruitful one for sociology to adopt. However, he also argues that "it is clearly not very satisfactory if the roles of individuals have to be seen entirely, or even overwhelmingly, in terms of interests conflicting with those of others."171 Although most conflict theorists believe that the absence of conflict is a result of either effective coercion or "false consciousness," Dahrendorf believes that social relationships can be mutually beneficial. In a recent discussion of the need for greater industrial democracy, for example, he suggests that the sharing of information and decision-making may transform conflicts from "a zero-sum game in which the gains of one side are the losses of the other, into an intricate mixture of co-operative efforts to increase the cake, and conflicts about dividing it Up.172 The Determinants of Conflict: A Theory of Conflict Groups In his best-known work, Class and Class Conflict in Industrial Society, Dahrendorf addresses the question of when inequalities and conflicting interests will actually produce conflicts. His central argument is that social conflicts will take place, systematically, among groups that differ in the authority they enjoy over others. By authority, Dahrendorf (again following Weber) means the sort of power that is attached to a social role or position, that is legitimate in the sense of being defined and delimited by social norms, and that is backed by sanctions up to (and no further than) these limits. A university, for example, has the authority to charge you for your courses, board, and lodging, but not to take all your money. A mugger has the power to do just that-but no authority at all to do so. Dahrendorf's position is that the stable and recurrent patterns of institutional authority systematically give rise to social conflict between those who have some degree of authority and those who have none. In a departure from conventional "economic" usage, Dahrendorf labels these groups classes. He writes that "the term 'class' signifies conflict groups that are generated by the differential distribution of authority in imperatively coordinated associations:"173 that is, organizations in which orders are given and taken. Dahrendorf's theory thus implies that authority is dichotomous: you have it or you do not, and your interests are formed accordingly. Critics have suggested that whether you have more or less authority may be equal- ly important and thus that conflict may form around other groupings. But Dahrendorf affirms, with Marx, that conflict involves only two sides. On the other hand, all "classes" do not engage in active conflicts all the time. Dahrendorf therefore attempts to explain when people will actually mobilize. 174 The Mobilization of Classes. The structural requirements for people to form active "interest groups" 175 are "technical," "political," and "social." Technically, Dahrendorf argues, a group requires a founder and a charter or ideology to become active. Politically, the more liberal the state, the more likely is mobilization for active conflict; the more totalitarian the state, the less likely. Finally, three social factors are important. Group formation is more likely first, if potential members are fairly well-concentrated geographically; second, if they can communicate easily (as modern communications technology makes it easier for them to do); and third, if people who stand in the same relation to authority are recruited in similar ways and come, for example, from the same type of families or educational organizations. The most important psychological requirements are, of course, that individuals identify with the interests associated with their position and that these interests seem important and "real" to them. Dahrendorf disagrees with Marx's argument that people's class positions (in either the "property" or the "authority" sense) determine the whole of their social life and behavior; but he does believe class interests will be more "real" to people who also share a culture. He also suggests that people are less likely to identify with their class's interests and to mobilize, the larger the number of associations to which they belong.176 Finally, the greater people's personal chances of leaving their class-in other words, the greater the degree of "intra-generational mobility"177-the less likely they are to identify actively with it. Dahrendorf s "structural" arguments are fairly convincing, although he pays surprisingly little attention to force. He tends to emphasize that if conflict is not to become explosive, there must be some degree of mobility and freedom to express opposition. But as tyrannies throughout history have shown, a sufficient degree of force can suppress conflict very effectively. His discussion of the "psychological" requirements for class action is less satisfactory. Especially in preindustrial societies, the poor have often shared a culture, led lives confined to the immediate community, had few opportunities for advancement, and yet accepted the existing order of things almost without question. Thus, Dahrendorf fails to explain satisfactorily how attitudes of opposition originate. The Violence and Intensity of Conflict. Dahrendorf also discusses at considerable length what affects the intensity and violence of class conflict when it occurs. Violence he defines as "a matter of the weapons that are chosen," intensity as the "energy expenditure and degree of involvement of conflicting parties."178 Dahrendorf argues that there is one preeminent factor affecting the degree of violence. This is how far conflict is institutionalized, w ith mutually accepted "rules of the game"; for "those who have agreed to carry on their disagreements by means of discussion do not usually engage in physical violence."179 For example, the days of extreme violence in strike-breaking and picketing in America predate the general acceptance of unions. Dahrendorf also identifies three important factors influencing the intensity of a conflict. Of these, the first (which he also considers to be the most importantly) is the degree to which those who are in positions of subjection in one association are in the same position in their other associations. The second and parallel factor is the degree to which authority in an organization is held by people who are also "on top" in other respects: in Dahrendorf's terms, whether positions are "pluralist" or "superimposed." Thus, if the managers of firms are also their owners and if they also use their wealth and position to control politics, one can expect particularly intense industrial conflicts. Dahrendorf's third argument is that the greater the mobility between positions, the less intense the conflict will be. This is true not only when individuals themselves can move, but also when their children are mobile. This is partly because mobility makes it less likely that a class will have a common culture and partly because people are less inclined to attack a class their children may one day join. On the other hand, if there is little or no mobility, the struggle becomes more intense. Dahrendorf's theory allows one to pinpoint potential areas of conflict within any given organization. As we have seen, however, it does not give enough attention to the role of brute force. Nor does it help us much in predicting which organizations (of the many in which people give or receive orders) are more likely to experience open conflict. One simple reason for the latter problem is that Dahrendorf does not discuss how important a given institution is to someone's life. In a society of "superimposed" positions, there is almost certain to be one major source of power and authority from which the others follow: when conflicts in totalitarian countries do break through, they center therefore on the party's power and control of the state. In Western societies, since both the state and one s place of work are very important institutions, they are both far more likely to generate active conflict than an athletic or church social club. Dahrendorf's own discussion does in fact pay most attention to industrial enterprises and the state. Conflict in Industry. During the last hundred years, there have been increasing numbers of 'Joint stock" companies, in which ownership by shareholders is divorced from management control. As we have s6len, modern Marxist sociologists generally argue that the change is not very important, since the firms still represent and are run in the interests of the owners. Other writers, such as Burnham,18' believe that it signifies a farreaching change in social structure and the roots of power. Dahrendorf argues that his approach demonstrates both what has really changed and what has not. He suggests that because nineteenthcentury managers and owners were generally the same people, Marx mistakenly ascribed to property differences a conflict that actually centered on authority, although it was intensified by the "superimposition" of industrial authority, wealth, and political influence.182 Today, he continues, industrial conflict will be less intense both because ownership and control are separate and because of industry's "institutional isolation" 183 (which means that one's position in industry has less to do with tb.e rest of one's life than before). At the same time, the split in authority and the conflicts of interest it generates remain. Arguments that the divide between workers and management has blurred are quite mistaken. The study most directly relevant to Dahrendorf's argument is probably The Affluent Worker. 184 This intensive survey of affluent British industrial workers was largely aimed at determining whether there had been a breakdown of the old social divisions between the "working" and the "middle" classes. The results cast doubt on the degree to which industry is, as Dahrendorf suggests, "institutionally isolated." The workers' family-centered social life, for example, remained quite separate from that of whitecollar families. However, although union organization was stronger in these plants than in many much less wealthy areas, these workers were only rarely committed to the "union movement" as a national "Working-class" force; instead they were extremely involved in union affairs at the level of the workplace. Only 8 percent attended union branch meetings often enough to vote regularly in branch elections, but 83 percent voted regularly for their shop stewards. In other words, they were closely involved in union organization at the level where they were given orders directly and where they grouped together, to face and face down the managers who gave them; but not in questions of union and economic policy or "class" politics at the national level. Conflict and the State. Dahrendorf argues that in the state, as in industry, the crucial lines of conflict are between those who give and those who receive orders. The state is the most powerful association,, in society, and the "ruling class" is, in a sense, the elite group that holds the positions at the top of the state hierarchy. But the ruling class is not composed solely of this group. The bureaucracy, too, belongs to a chain of command, and this command of authority makes it part of the ruling class, even though it does not define the concerns and objectives of a bureaucratic state. Dahrendorf's argument helps explain the enormous stability of such bureaucratic states as Byzantium and the Egypt of the Pharaohs. The larger the authority-bearing class, the larger the group that will Marshall against any threat to it from an organized conflict group of subordinates. His argument also implies that the state and bureaucracy are together a separate institution, not simply a reflection of other social groupings, and that other powerful social groups will necessarily oppose the state's authority and try to restrict its control o.ver them. In Dahrendorf's own recent analyses of British politics,187he argues that, "Clearly there is today a conflict between government and industry,"188 in which the trade unions are the most visible protagonists but which also involves the giant companies. A number of modern commentators echo Dahrendorf when they write, that growing government activity will have consequences for the range and intensity of political conflict. For example, Christopher DeMuth, a faculty member at Harvard's Kennedy School of Government who for- merly worked for Conrail, discussed the behavior of the huge Greyhound bus company in the Wall Street Journal. He described how Greyhound lobbies vigorously against the federal subsidies given to passenger trains, with which the buses must compete; while it attempts with equal vigor to stop the Interstate Commerce Commission from allowing new bus companies to start competing with Greyhound and tries to prevent any general deregula- tion of bus transport. DeMuth argues, "There is no reason to expect Greyhound or any other company to compete in the economic marketplace while abstaining from the political marketplace... "(when government in- volvement) becomes ... sufficiently large, it alters fundamentally the nature of competition in the private part (of the economy)-increasing the relative importance of political as opposed to economic competition.189 Summary Dahrendorf provides an illuminating account of the close and per- manent relationship between power, or authority, and conflict. He presents a concrete theory of conflict group formation, which provides a good starting point for explaining people's objectives and identifying potential confrontations. He describes a number of the important factors that either create mobilized conflict groups and intense conflicts or, correspondingly, tend to reduce social conflict. However, his theory of conflict group mobilization fails to explain what makes people aware of themselves as a group with common interests and common grievances. Neither does he specify precisely which institutions in a society will be the ones in which conflict will actually occur. This makes it difficult to "test" Dahrendorf's basic proposition about conflict formation. Nevertheless his general perspective has proven illuminating enough to suggest that the polarization which Dahren- dorf identifies between those with and without authority is at least one important source of social conflict. LEWIS COSER Like many other conflict theorists, Lewis Coser (b. 1913) combines a distinguished academic career with a strong interest and involvement in social policy and politics. He was born in Berlin, and after the Second World War he taught at the University of Chicago and Brandeis University while ob- taining a Ph.D. from Columbia. Most of his academic career was spent at Brandeis, where he moved from being a lecturer to a full professor; but since 1968 he has been Distinguished Professor of Sociology at the State University of New York at Stony Brook. In 1975 he was president of the American Sociological Association. Coser's writings have always displayed his concern with politics and the links between ideas and the nature of society. He is coauthor of a history of the American Communist Party dedicated to Milolovan Djilas,'90 and he has both edited Dissent and contributed frequently to such other serious nonspecialist journals as Partisan Review and Commentary. Some of his more recent work is about "greedy institutions" that demand total involvement from their members. Coser says that in the face of recent "indis- criminate" condemnations of the "differentiated, segmented and 'alienated' character of modern life,"191 he felt bound to point out the threat to human freedom inherent in total involvement. "I wish it to be clearly understood," he writes, "that I consider it essential that an open society be preserved above all."192 Of the modern theorists discussed in this chapter, Coser is the closest to Simmel. He is the most concerned with the "web of conflict," or the cross-cutting allegiances that can both bind a society together and generate struggles and confrontations. Indeed, Coser's major book on conflict theo- ry, The Functions of Social Conflict, 193 is an exposition and development of Simmel's own rather fragmented insights. Coser emphasizes that conflict, although very important, is only one side of social life and no more "fundamental" than consensus. 194 Coser's contributions to conflict , theory are also distinctive in two other respects. First, he discusses social conflict as a result of factors other than opposing "group interests." Second, he is concerned with the consequences of conflict. As we have seen, Dahrendorf's major concern is with the origins of conflict, which, he argues, then creates social change. Coser has less to say about the institutional roots of conflict, but he distinguishes among its different possible consequences, which include greater social stability as well as change. His discussion of the conditions under which conflict is likely to be divisive or cohesive adds considerably to Dahrendorf's analysis of conflict's characteristics. The Origins of Social Conflict In his discussion of the origins of conflict, Coser pays far more attention than do most conflict theorists to the role played by people's emo- tions. 195 He agrees with Simmel that there are aggressive or hostile "impulses" in people, and he emphasizes that in close and intimate relationships, both love and hatred are present. Close proximity, he points out, means that there are also ample opportunities for resentment to develop; hence, conflict and disagreement are integral parts of people's relationships, not necessarily signs of instability and breakup. At the same time, Coser argues, the forms that hostility and conflict take and their relative frequency in different situations have to be explained in terms of social institutions and social roles. A good example of what Coser means is the differences among countries in how much children fight with their parents. This is the sort of "close" relationship in which some resentments are inevitable. However there are also intercultural variations in the father's authority; whether the children have financial independence; whether the steps by which children assume well- defined adult roles are clearly laid down; and whether other family members provide practical and emotional support outside the nuclear family. Coser's own work addresses the way such "structural" factors interact with people's underlying emotions. Coser defines two basic types of conflict, "realistic" and "nonrealistic." 196 In realistic conflict, people or groups are simply using conflict as the most effective way of getting what they want; if they could get what they wanted without a fight, they would give up the conflict immediately. Coser's inclination is to see most kinds of social conflict as essentially "realistic," or rational--explicable in institutional terms. Thus, in the Los Angeles Watts riot of August 1965, Coser saw something other than the "insensate rage of destruction" described by the official McCone report. The path of the riot, he pointed out, led straight to City Hall; the riot was a way of getting the authorities' attention. Similarly, the Luddite machine smashers of the early nineteenth century were highly selective in their targets; they were, he argues, engaging in "collective bargaining by riot" at a time when more organized union channels were outlawed. 197 Conflicts of this type are essentially the same as those that other conflict theorists analyze in terms of "self-interest." However, Coser argues, there are also "nonrealistic" conflicts, in which the conflict is the end in itself, whether or not this is admitted. Nonrealistic conflict serves as a way of releasing tension or affirming one's identity, and it embodies hostilities that actually derive from other sources. Blaming everything on a scapegoat is a classic example. Often, of course, a conflict contains both realistic and nonrealistic elements. The Consequences and Functions of Conflict In his analysis of the results of social conflict, Coser argues that conflict often leads to change. 198 It can stimulate innovation, 199 for example, or, especially in war, increase centralization.200 For the most part, however, Coser concentrates on conflict's role in maintaining group cohesion. This is, of course, the subject that most concerns functionalists. However, Coser is a "functionalist" only in the sense of sharing this interest. He does not imply that it is necessarily desirable for a group to survive and remain cohesive or that conflict occurs because it may be functional for the group; and he sees cohesion as only one of conflict's possible results. In this context, Coser distinguishes between conflicts that are external and conflicts that are internal to a group. Both types, he argues, can define a group, and establish its identity; and also maintain its stability and increase its cohesion. Extemal Conflict. In his most unqualified statement about the rela tionship between conflict and cohesion, Coser argues that external conflic is essential in establishing a group's identity. In this he is following not onli Simmel but also Marx, who felt that only conflict makes a class self-aware Coser states, with Simmel, that "conflict sets boundaries between groups within a social system by strengthening group consciousness and awareness of separateness, thus establishing the identity of groups within the system."201 However, he also distinguishes expressly between hostile senti ments and actual conflict; 202 and we would suggest that it is hostile senti ments, rather than actual conflict, that are essential to group formation. A religious group or utopian agricultural settlement of the type common in American history may coexist with other parts of society without overt conflict. Coser also argues that external conflict can often strengthen a groul It makes group members conscious of their identity by introducing strong "negative reference group" to which they contrast themselves; also increases their participation.203 Nonetheless, the process is not inevitable. If internal cohesion before the outbreak of conflict is very low, the conflict may simply hasten disintegration. Coser contrasts,,,(be disruptive effects of the Second World War on French society with its unifying effects ).204 in Britain, and a less dramatic but similar example is America during the Vietnam and Second World Wars. However, since Coser offers no way of telling in advance which development is likely, his arguments here are not very useful in explaining actual conflicts. Intemal Conflict . Coser follows Durkheim, Mead, and even Marx in arguing that a group's opposition to and conflict with "deviants" makes apparent to group members what they ought to do.205 In this sense, internal conflict is central to defining a group's identity, which is embodied in norms that define "correct" behavior. Coser also argues that internal conflict can increase a groiip's survival, cohesion, and stability. He again follows Simmel when he argues that internal conflict is a crucial safety valve under "conditions of stress ... preventing group dissolution through the withdrawal of hostile participants." 206 If opposition to one's associates were not possible, people would, in Simmel's words, "feel pushed to take desperate steps ... opposition gives us inner satisfaction, distraction, relief." 207 Indeed this is one of the few occasions when Coser falls into the trap of implying that because som ething is functional it will automatically happen, and that because safety valves are important all societies will provide them.208 Coser's argument here seems to confuse a group's survival or stability with its cohesion. He implies that whatever promotes survival also promotes cohesion. However, if it uses brute force and terror to support imbalances of power, a society can survive and even remain stable in the face of a great deal of internal hostility: witness "Papa Doc's " Haiti or Stalin's Russia. Although conflict may help maintain a society by providing a safety valve, this pattern is hardly a necessary or universal one. On the other hand, Coser is probably correct in emphasizing the close links between internal conflict and group cohesion. In rigid social structures with no channels for expressing hostility, when conflict occurs it is disruptive and violent. The controversial "Sonnenfeldt doctrine" illustrates this theory of conflict at work in America's foreign policy. In 1976, considerable furor resulted from leaks of a statement made by Helmut Sonnenfeldt, then a senior State Department official, in which he argued that the breakdown of Soviet control in Eastern Europe was undesirable because chaos would result when stored-up hostilities erupted. Although official spokesmen attempted to tone down the implications of this principle, they never clearly denied it; nor did the succeeding adminstration explicitly disown it. Finally, Coser argues that internal conflict can be important because "Stability within a loosely structured society ... can be viewed as partly a product of the continuous incidence of various conflicts crisscrossing it."209 When people belong to many different groups, each of which pursues its own interests and is consequently involved in its own conflicts, they are less likely to devote all their ener ies to a single conflict that could break the society apart. Thus, Coser suggests, "one reason for the relative:absence of class struggle' in this country is the fact that the American worker, far from restricting his allegiance to class-conflict groupings and associations, is a member of a number of associations and groupings ... (and) the lines of conflict between all these groups do not converge."210 In this context, Coser admires Max Gluckman's work on the importance of cross-cutting conflicts in African tribes. Gluckman notes that. "all over the world there are societies which have no governmental - institutions ... Yet these societies have well-established and well-known c odes of morals and law.... We know that some of them have existed over long periods with some kind of internal law and order, and have successfully defended themselves against attacks by others."211 These "feuding" societies rely on private vengeance instead of government. The process does not tear them apart, Gluckman argues, because they "are so organized into a series of groups and relationships that people who are friends on one basis are enemies on another. (For example) a man needs help in herding his cattle: therefore he must be friends with neighbors with whom he may well quarrel over other matters.... And through his wife he strikes up alliances with relatives-in-law which are inimical to a wholehearted one-sided attachment to his own brothers and fellow-members of his clan.... A man's blood-kin are not always his neighbours: the ties of kinship and locality conflict.... These allegiances ... create conflicts which inhibit the spread of dispute and fighting.212 Divisive Social Conflict In discussing conflict's role in defining a group and maintaining its cohesion, Coser makes the proviso that internal conflict has this role only if it is not about basic values and principles. This, in turn, depends on the rigidity of a society and, once again, the degree to which its members are interdependent. Coser argues, first, that an internal conflict is more likely to involve basic principles (and so be socially divisive) in a rigid society that allows little "expression of antagonistic claims. 11213We have already noted Coser's belief that some conflict is functional because it serves as a safety valve, without which social hostility would eventually erupt violently. Here, he suggests that when conflict emerges after having been suppressed for a long time, it will also split the group around basic issues and Values. Second, Coser returns to the theme of "cross-cutting" allegiances and argues that "Interdependence checks basic cleavages."214 Interdependence therefore makes divisive internal conflict much less likely and reduces the likelihood that external conflict will find a noncohesive group. As we have already discussed, the basic reason is that "interdependence" means that people with common interests in one respect are opposed in another, so that an overriding and polarizing issue is less likely. However, there is also a related psychological process involved. Conflicts are much more intense, Coser argues, when they involve exclusive groups; this further increases the likelihood that divisive conflicts will occur in societies that have non-overlapping groups. Such conflicts are intense because it is in close relationships that love and hate coexist. "The coexistence of union and opposition in such relations makes for the peculiar sharpness of the conflict," he argues.215The fewer the groups people belong to, the more likely they are to become intensely involved in the ones they have. An intense involvement affects the nature of conflicts both within the group itself and between it and others. Although Coser, a sociologist, makes this point formally, the leaders of sects and the founders of revolutionary parties have always been aware that the more a person's relationships all lie within the group, the more his loyalties and energies will be at its disposal. The mass suicides of Jonestown, when several hundred members of a religious sect drank poison at their leader's behest, almost certainly could not have occurred if the group had been living within a larger and long-settled community, rather than isolated in the jungles of Guyana.216 Coser's arguments about interdependence follow reasoning similar to James Madison's in The Federalist Papers. One of Madison's reasons for urging voters to adopt the American Constitution was that the resulting union of states would be one of considerable size and diversity. This would reduce the likelihood of one "faction" trampling over its fellow citizens and of oppression of a minority by the majority. He argued: The smaller the society, the fewer probably will be the distinct parties and interests composing it; the fewer the distinct parties and interests, the more frequently will a majority be found of the same party; and the smaller the number of individuals composing a majority ... the more easily will they concert and execute their plans of oppression. Extend the sphere, and you take in a greater variety of parties and interests; you make it less probable that a majority of the whole will have a common motive to invade th@ rights of other citizens.217 Finally, although Coser stresses the importance for divisive conflict of a society's rigidity and the degree of its members' interdependence, his arguments about the role of ideas in conflict are also relevant here. Coser suggests that conflicts will be more intense when those involved feel that they are fighting on behalf of the group and not just for themselves when, in consequence, they feel that what they are doing is morally legitimate. They also then feel strengthened by power derived from the collectivity with which they have identified and which they embody.218 This is why Coser argues that intellectuals increase the bitterness and the radicalism of conflicts between labor and management; in Mannheim's words, they transform "conflicts of interest into conflicts ideas"219 and provide just this moral impetus. Summary The major importance of Coser's conflict theory lies in its demonstration that conflict can often be neither socially divisive nor a source of change. This is especially clear in his comparison of societies with or without many independent and overlapping groups, and of the differing natures and consequences of conflict within them. However, Coser's account is not fully satisfactory. Complex and interdependent societies that are not particularly "rigid" may experience conflicts that are highly divisive; and rigid, hierarchical societies may survive for centuries without explosive conflict. The recent history of political strife in Central America and the breakdown of democracy in Chile and Argentina illustrate our first point, the thousand-year history of the Byzantine Empire our second. In addition, the abstract quality that Coser's work shares with Simmel's and the way it ignores the nature and bases of group resources mean that on its own, it explains very little about concrete social situations. Finally, Coser's emphasis on the functions of conflict, although a useful corrective, is onesided. At the very beginning of his work, he described social conflict as a struggle in which opponents attempt "to neutralize, injure or eliminate their rivals."220 Thereafter, one looks in vain for him to pay much attention to such behavior or to recognize the stability of many situations based on force and oppression. RANDALL COLLINS Randall Collins (b. 1941), the last of the conflict theorists we shalt discuss, is also the youngest, and his work exemplifies the growing interest among American sociologists in a conflict perspective. We noted earlier the renewed interest in Marxism and the discovery of the Frankfurt School by younger, left-wing sociologists, who see their sociology as inextricably entwined with their desire for social change. Many of their contemporaries, whose aim in their sociology is scientific explanation and not political action, also now believe that a conflict perspective is the most fruitful way to approach sociological analysis. 221 Collins' work is the most far-reaching theoretical synthesis based on this approach to date. Unlike Marx's or Dahrendorf's, Collins' work is not a conflict theory in the sense of describing when social conflict will actually occur. Indeed, in many of the situations he describes there is no overt conflict at all. Rather, he sets out to show that one can explain a wide range of social phenomena on the basis of a general assumption of conflicting interests and an analysis of the resources and actions available to people in particular social situations. Collins took a B.A. at Harvard, an M.A. at Stanford, and a Ph.D. at the University of California, Berkeley. While still a student working as a research assistant at the Institute of International Studies at Berkeley, he began publishing.222 He taught at the University of California, San Diego, and at the University of Virginia in Charlottesville; he is currently at the University of California, Riverside. Collins' work incorporates all the major elements of conflict theory: an emphasis on people's interests, a view of society as made up of competing groups whose relative resources give their members more or less power over each other, and an interest in ideas as a weapon of social confl'ct,and domination. As he acknowledges, Collins' major debts are to Marx, "the great originator of modern conflict theory,"223 and above all to Weber, 224 whose analytic framework, comparative historical approach, and non-Utopian outlook he adopts. In addition, of all the conflict theorists we have discussed, Collins draws most on Durkheim ' who is commonly seen as the father of functionalism. Collins has little good to say about functionalism, but he believes that Durkheim explains a great deal about the ways in which emotional bonds and loyalties are created among people. The most original aspect of Collins' work is the way he incorporates a theory of how social integration is achieved into a conflict approach. Unlike most conflict theorists, Collins also draws on the work of such theorists as Mead, Schutz, and Goffman, who are enerally associated with such "micro-sociological" perspectives as 9 symbolic interactionism and phenomenology. This is the result of his concern with exactly how individual loyalties and emotional bonds develop. The Nature of Conflict Sociology Collins' basic assumptions are that there are certain "goods," namely wealth, power, and prestige, that people will pursue in all societies,225 and that everyone dislikes being ordered around and will always do his or her best to avoid it.226 That is, he assumes that people have certain basic in- terests wherever they live and that they will act accordingly. From this it follows that there will always be social conflict. This is true even though not everybody is equally greedy, simply because power, in particular, is inherently unequal: if I have a great deal of power, other people must necessarily have less, and they must obey me. "Since power and prestige are inherently scarce commodities, and wealth is often contingent upon them, the ambition of even a small proportion of persons for more than equal shares of these goods sets up an implicit counter-struggle on the part of others to avoid subjection and disesteem," Collins con- cludes.227 Such social conflict can take many forms, but at the very heart of it, he believes, lies direct coercion. Force is something people can always turn to, and some people always fare better at it than others. "Above all else, there is conflict because violent coercion is always a potential resource, and it is a zero-sum sort.11228 As part of his analysis of the determinants of social structure and change, Collins provides a typology of the resources people bring to this struggle. First are material and technical resources, which include not only property, tools, and such skills as literacy but also-very importantly- weapons. Second, he stresses the role that strength and physical attractiveness play in personal relationships. Third, he mentions the sheer numbers and types of people with whom individuals have contact and with whom, therefore, they have the possibility of negotiating for material goods and status.229 Fourth, Collins emphasizes the resources people possess in their "store of cultural devices for invoking emotional solidarity. 11230 By this he means their ability to create and maintain a shared view of how things are and should be, which also sustains the favored position of those promoting that view. We can see what Collins is talking about if we compare the position of a rural Indian "untouchable" with that of, say, a nineteenth-century Hindu maharajah. The untouchable has almost no skills he could market in a city, for he is barred by the caste system from all but certain "unclean"jobs, such as curing hides or cleaning latrines. For him to work with caste Hindus would pollute them. His social contacts are extremely limited, so he knows very little about whether he has any alternatives; he is dirty and undernourished; and his religion, dominated by the priestly caste of the Brahmins, teaches him that his lot is the inevitable result of his conduct in previous lives and must be accepted as such. By contrast, the maharajah has enormous resources and breadth of opportunities and social contacts. Being better fed, he is almost certainly bigger and more attractive physically. Moreover, although he has little or no control over the conduct of the Hindu religion and the influence of its priests, his cosmopolitan contacts may well have weakened the priests'influence over him. Obviously the maharajah is in a position of enormous power compared to the untouchable, and he is in a position to achieve a much wider range of objectives. If you compare two very dissimilar people in America, such as a mother on welfare and a company president, you will find that there are again big differences-but also that they are less extreme. This is because the two societies distribute their resources very differently. There are, Collins suggests, three main areas of life in which people obtain more or fewer resources and are more or less dominant or subjected. Together these create the patterns of "social stratification." They are first, people's occupations, where they can be grouped into different classes; second, the communities where people live, with their different status groups, including age, sex, ethnic, and educational groupings; and third, the political arena, where different parties seek political power.231 In every case, what is crucial for social behavior is the degree to which people are in a position to control others and so obtain wealth, status, and overt deference. Collins does not regard one aspect of stratification as primary: people's overall position is simply the sum of their resources and positions in a number of different areas. In other words, Collins prefers Weber's pluralist model to Marx's unicausal one, arguing that "different orders of stratification do not line up neatly. The major part of Collins' work sets out propositions about the concrete ways in which the distribution and use of resources result in different sorts of social behavior and institutions. He also discusses in detail what creates a shared culture and a "legitimate" social order. Social institutions and the Balance of Resources Some of Collins' most interesting arguments about the relationships among conflicting interests, the distribution of resources, and the nature of social institutions concern modern education, organizational theory, and the state. In each case, we shall find that his method is drawn from Weber: for the most part, he discusses general principles in terms of concrete historical settings and the balance of power within them, much as we de- scribed Weber doing in his Religion of China.233 Stratification by Education . Collins has always been interested in how educational qualifications have been used as a resource in the struggle for power, wealth, and prestige; and he has used a conflict perspective to carry out empirical research on the role of education in people's careers. He treats education as one important basis of status-group differences, a sort of "pseudo-ethnicity" that socializes people into a particular form of culture.234 The educational elite, which shares a given culture, will use it as a criterion for employment in elite positions, he argues, and it will also try to instill respect for its culture in the society as a whole.235 Modern industrial societies have longer periods of formal education, and require far higher educational credentials for jobs than preindustrial ones. The usual explanation given is that more jobs require more skills than in the past; but Collins argues that this is, at best, only a part of the truth. The evidence shows that education "is not associated with employee productivity on the individual level, and job skills are leart,,.i!ed mainly through opportunities to practice them",236 and not at school at all. Dif- ferences among industrial countries are very marked, but not obviously related to technological progress. For example, in the postwar period, American education has, on average, gone on far longer than that of the Japanese or West Germans.237 Moreover, Collins points out, engineers are the one professional group with real technical skills that industry needs. If the "technological" explanation of modern education were correct, one would expect the engineering schools to have the highest prestige on campus, and engineering training to dominate education. Nothing of the sort has occurred. Indeed, America, at the undergraduate level, has "the most massively nonvocational system of education in the modern world.11238 The best way to understand the enormous growth of the "credential system" is to see education as a way of setting up entry require ments for jobs (and so limiting competition); and creating exclusive occupational cultures which are defined as prerequisites for doing a 'ob. Thus, if you do not speak the same language (or jargon) as established workers, having learned it during "professional" training, then this provides evidence against your competence. This process can, quite quickly, create the sort of 44 credential inflation" which we have experienced in recent years. The highly educated set upjob requirements that favor them; people recognize the importance of education as a route to success and acquire ever more of it; employers raise entry requirements yet further to screen out the flood of applicants-and so on. Weber showed that the Chinese literati used education in just this way, and monopolized senior advisory positions on the basis of their "qualification" of a literary Confucian education.239 Collins identifies similar examples at work today. For example, the most "educated" executive group is found not in,the most rapidly developing high-technology companies, but in the " traditionalistic" financial and utilities firms.240 Collins' analysis is similar to the Marxist analysis of Bowles and Gintis 241 in its emphasis on the role of education in creating and preserving social position. However, Collins does not see this as a particularly or specifically capitalist process. Rather, education is always a potentially important resource; though societies will vary in the degree of suc'(:,Oss which educated groups achieve. The analysis also has echoes of Ivan Illich's famous diatribe against formal education, Deschooling Society. Illich argues that schools teach little. Their main purpose is to assign social rank. "School reserves instruction to those whose every step in learning fits previously approved measures of social control," he charges.242 Organizational Theoty Collins' general approach to organizations is to view them as "arenas for struggle" in which superiors try to control their subordinates. He argues that, in consequence, one can best understand the structure of an organization by looking at the sanctions available to those in charge. The three main types of sanction which can be used are c...oercion; material rewards given in return for appropriate activities and behavior; and normative control, which invokes people's loyalty to ideals. Each has distinctive consequences. People dislike and resist coercion, and subordinates in the organizations that rely on coercion resort to "dull compliance and passive resistance."243 Forced labor camps are extremely unproductive, and slaves consistently appear stupid and irresponsible to their masters. Material rewards, Collins suggests, are less alienating, but they produce organizations where there are continual fights over pay and piecerates, where informal groups control the work pace and prevent people from working "too hard," and where people tend to do th ngs only if and insofar as they are paid.244 Normative control is the most desirable for superiors, because if subordinates share superiors' goals, they will be far more motivated to cooperate, obey, and work hard.245 However, it too has costs attached. One of the most effective ways to create normative control is to spread authority, because the more people give orders in an organization's name, the more they identify with it. Superiors can, for example, "co-opt members into responsible positions; a related method is to offer them a chance to be promoted. 246 Both these approaches, however, tend to reduce the superiors' own centralized power. An alternative tactic is to recruit members who have a feeling of solidarity or friendship. Nepotism, the recruitment of family members, is a traditional way of creating organizational loyalty,247 and recruiting people with similar backgrounds also makes friendships and thus loyalty to the organization more likely.248Collins also argues that the more an organization emphasizes rituals, is isolated from the world, and feels itself to be in danger from outside,249 the more it will reinforce identification and loyalty. A good example of how isolation increases people's loyalty to an organization is the landslide majority that white South Africans gave to the. Nationalist Party, the architects of apartheid, in 1978, a time of strong public American pressure and criticism. However, Collins points out, people who like each other tend to form strong informal groups, which may disagree and conflict with their leaders. Organizations that rely on normative control can expect continual conflicts about policy, and factional fights about who is best able to implement the group's goals. "Astute organizational politicians, then, always attempt to mix normative incentives with material rewards and perhaps subtle coercive threats, since the latter can be administered much mor6 routinely and kept under more stable control. 11250 To transform the revolutionary Bolshevik party into the bureaucratic Communist party (which has controlled Russia without challenge for over fifty years), Stalin changed from the 4 t normative control" of shared ideals, with its intense ideological disputes, to control by the material rewards available to party members and the underlying fear of the secret police. The State . In using a conflict perspective to analyze the state,,, Collins again emphasizes that coercion is at the core of social life and social ' conflict. The state, he suggests, is a special sort of organization, because it is "the way in which violence is organized"251 and "Is, above all, the army and the police."252 To analyze political systems, Collins suggests, one must ask how violence is organized and which interest groups control and influence the state's policy. Collins argues that the resources enabling people to win control of the state include systems of belief, property, networks of communication, and military technology.253 Indeed, of all the theorists discussed in this chapter Collins is (after Schumpeter) the most sensitive to the effects of military technology and military organization on political and social life. A society in which a large proportion of the men fight with their own cheap weapons will be far more democratic, he argues, than one in which a sma group with expensive individual equipment, like the medieval knights, monopolizes warfare. Similarly, whether an army is supplied and paid from central stores, like the Roman army, or requisitions locally and takes conquered lands as its reward, is crucial in determining whether there is centralized state or a loose-knit feudal order. Unlike Collins' analysis of military force, his discussion of economic resources as a basis for political power is sketchy. Collins sees coercion and military power as primary, and in discussing the economic underpinnings of political systems, he essentially restates Lenski's rather Marxian typology, in which techniques of production are the major causal factor.254 Collins' discussion tends to neglect the effect of different forms of property and economic organization on political structure. He also ignores the different laws and systems of law in which these forms are rooted, although law is as essential a function of the state as is the use of force. Curiously enough, Weber, the greatest influence on Collins'work gave considerable emphasis to economic organization and its legal basis. For example, he pointed to the historical and political consequences of the ability of European townsmen to own their own labor and property and to govern their towns; and he showed the consequences of the way the Chinese were legally tied to the village-based sibs or clans, while their towns had no rights to self-government. Culture, Ideology, and Legitimation In discussing Collins' analysis of social structure, we have noted the importance he attaches to the creation of legitimacy. This emphasis on the role of ideas is, of course, common to conflict theory; but Collins' distinctive contribution is to provide detailed discussions of both the actual process by which common beliefs are generated and of the general relationships between people's outlook and the way they experience conflict situations. Collins is continually at pains to remind his readers that his subject is ultimately individual people. Writers in the tradition of "social phe- nomenology," such as Mead and Goffman, have, he believes, a great deal to tell us about social interaction, because they focus on individual experi- ences and recognize that these are not something given, fixed, and immu- table.255 Instead, our experiences are in large part a result of our own perceptions and values, so that "men live in self-constructed subjective worlds."256 These worlds determine whether we see a photograph, for example, as harmless, as a way of stealing our souls, or as a transgression of God's commandment against graven images. They also affect whether we experience our societies as good and satisfactory or oppressive. Moreover, Collins argues, the sorts of perceptions we have are themselves subject to regular and identifiable influences. We can identify the sorts of individual experiences that make people view a social order as "real" and legitimate, and so alter the course of social conflict. The experiences Collins identifies are: first, and most important,257 the giving and taking of orders; and second, the sorts of communications people have with others.258 He argues that because of the nature of human psychology, those who gave orders will tend to identify with the ideals of the organization in which they hold power and in whose name they justify their orders; and because of their experiences, they will be self-assured and generally formal in manner. By contrast, the more people receive orders, the more alienated from organizational ideals they are likely to be, the more fatalistic, subservient, concerned with extrinsic rewards, and the more distrustful of others. Those in the middle of the authority hierarchy, who give and receive commands, tend to combine subservience with organizational identification, but they are little concerned with an organization's longterm objectives. Collins' basic premise is that people wish to maximize the degree to which they give rather than receive orders. Hence, he argues, a mid-'level official or bureaucrat "attempts to transform order-taking situations into orders that he passes on to others."259 This explains the proverbial inflexibility of mid-level bureaucrats compared to their superiors, who see rules as a means to more distant ends. On the other hand, Collins' general emphasis on the relationship between one's outlook and whether one gives or receives orders implies a world in which everyone works in large organizations. In fact, if you think about people's occupations, you will find that many people-such as farmers, insurance salesmen, and many housewives-are not much involved with "orders" on a day-to-day basis at all. The types of communication that people have with others are extremely important, Collins argues, because they may reinforce, or (and this is crucial for social control) offset the effects of people's experiences of command. They, above all, are what determine the degree to which people accept as "real" the ideas and norms around which their society is organized. They determine whether people see the social order as legitimate. He points out that human beings, like all animals, have automatic emo- tional reactions to certain gestures, sounds, and signals; and he suggests that social ties are fundamentally based on shared reactions, -'@l,hich differ from those of animals because they involve symbols (such as flags or salutes) rather than genetically programmed sounds and gestures.260 The strength of these reactions depends on two aspects of people's communications with each other: the amount of time people spend together, or their 4 4 mutual surveillance," and the diversity of their contacts. These, Collins argues, correspond to what Durkheim called "social density" in his analyses of how "society" creates loyalty and identification among its members. 261 Social Density Collins argues that the greater the degree of mutual survel 'llance-the more people are in the physical presence of others-the more they accept the culture of the group and expect precise conformity in others.262 Conversely, the less they are around others, the more their attitudes are explicitly individualistic and self-centered.263 This is in large part because being physically together makes it more likely that "automatic, mutually reinforcing nonverbal sequences" will develop. These increase "emotional arousal"; and "The stronger the emotional arousal, the more real and unquestioned the meanings of the symbols people think about during that experience."264 The diversity or cosmopolitanism of people's contacts similarly affects the way they think. Collins'argues that the more different sorts of comrnunication people are involved in, the more likely they are to start thinking in abstract terms and in terms of long-range consequences. On the other hand, the less varied people's contacts, the more likely they are to think only about particular people and particular things: and to see the world outside their own familiar circle as alien and threatening. Thus, limited contact will tend to create shared, local views of reality, and a feeling of identification with familiar local people vis-A-vis the outside world.265 Ritual Collins believes that mutual surveillance and limited communications counteract the effects of being on the receiving end of orders. They create emotional ties that bind a group together and make the way it is organized unquestionably "real." Thus, they strengthen the position of the dominant members of a group. In addition, Collins argues, rituals or stereotyped sequences of gestures and sounds",266 can make people's emotional arousal more intense and so commit them more strongly to certain views of reality. Emotional arousal is also affected, he suggests, by the sheer number of people involved;267 and if we look at the ceremonies to which groups and societies attach great importance, we can see that they generally combine highly stereotyped rituals with large gatherings of people. Durkheim pointed this out in his discussion of aboriginal religious ceremonies and their role in "integrating" society;268 Hitler's Nuremberg rallies, graduation ceremonies, and the use of particular songs and symbols at civil rights or antiwar demonstrations are also good examples. Collins' arguments here complement Schumpeter's and Habermas' discussions of the declining legitimacy of the modern market economy.269 The "rational attitudes" of such a society mean that we tend to find ritual increasingly distasteful: compare a president holding barbeques at the White House with a pharaoh or a medieval king holding court. Using Weber's terminology, Collins observes, "Traditional authority ... uses highly stereotyped gestures and verbal formulas, with the result that the symbols of authority are highly reified and emotionally compelling.... (Under) rational-legal authority ... little attention is paid to the surrounding postures, with the result that symbols are regarded as human enactments with little emotional compulsion."270 The empirical evidence gives Collins' arguments considerable but far from total support. Politics, about which Collins says surprisingly little, provide useful confirmation, because party activists often occupy a position in the party that is quite different from their day-to-day occupations in breadth of communication and power-and as Collins' theory would indicate, this affects their outlook substantially. Parents who are manual workers and who are active in politics and unionism tend to have upwardly mobile children who become managers and professionals.271 The federal Office of Equal Opportunity was founded in the 1960s on just these assumptions; for one of its objectives was to create, by involving poor people in "community action," a new set of attitudes towards the "authorities" and a new ability to deal with them. Again, Melvin Kohn's surveys of attitudes and values showed that middle-class parents ... are more likely to emphasize children's self-direction, and working-class parents to emphasize their conformity to external authority."272 Kohn argued that this can be explained for the most part by differences in whether the fathers' occupations allow self-direction or require them to receive and obey directions from others.273 On the other hand, evidence on working-class union mil ',6ncy and the development of revolutionary ideologies suggests that Collins' analysis is too simple. His account of how "ritual communities" are created would lead one to expect that militancy, revolutionary doctrines, and a commitment to "working-class solidarity" are most likely to develop among isolated, homogeneous groups of workers with their own tight-knit communities. For a number of years, sociologists have argued that this is indeed the case, citing the study by Kerr and Siegel, which argued that "isolated masses" of workers are more militant and strike-prone.274 More recently, however, this view has been challenged by a new research group interested in strikes and political violence, and, in particular, by the work of Shorter and Tilly.275 Their study of France during the period 1830-1968 shows that militancy was a characteristic of the metropolis, not the isolated en- clave-and of the relatively cosmopolitan skilled worker at that. It is similarly true, we might add, that tight-knit, fanatical, revolutionary political parties of the right and left have been born in heterogeneous cities, not in small, cohesive communities. This suggests a serious imbalance in Collins' treatment of ideas. In his emphasis on the way ideas can be used to support the ruling order, he neglects the origins of ideas and beliefs that oppose it. Indeed, his arguments sometimes leave one wondering why all ruling groups have not been able to assure themselves of permanent legitimacy by manipulating ritual! Tribal societies seem almost never to generate ideas of "revolution" (changing the entire order) as opposed to "rebellion" against individual leaders,276 but in Europe and America, such ideas have been a part of social life for many centuries, and they are held with quite as much passion and conviction of their "reality" as the beliefs of a tribal "ritual community." A full account of the origins of beliefs and world views needs to explain this process, too; and indeed, "conflict sociology" in general, needs to take more account of the origin and force of ideas as tools of opposition as well as domination. Summary Collins provides an excellent exposition of the basic assumptions of analytic" conflict theory. His work is also important because it provides a large number of concrete propositions that relate institutional structure to the resources available to different groups. Further, it incorporates the insights of "micro-sociological" perspectives, especially in its account of how social experiences affect people's outlooks and so affect the nature of social behavior, conflict, and change. Not surprisingly, Collins' major weaknesses are the weaknesses of conflict theory as a whole. The most important are, we would suggest, an overemphasis on the "zero-sum" aspects of social interaction; too mechanistic a view of ideas as an offshoot of the existing social structure; and an inadequate account of the nature of the state. It is to these general weaknesses of conflict theory, as well as its major strengths, that we now turn briefly in conclusion. CONCLUSION Conflict theory's major strength lies in the way it relates social and organi- zational structure to group interests and the balance of resources. This analytical framework is often very productive. Moreover, whereas func- tionalism never really identifies a mechanism of change, conflict t.heory does, when it points to shifts in resource distribution and power. Conflict theory insists fruitfully that values and ideas must be related to their social environment and not treated as autonomous. Finally, it avoids "explaining" things simply in terms of their results. By tracing social behavior back to individuals' interests and the purposive way they pursue them, it shows how changes may actually occur. However, conflict theory also has important weaknesses. Its insistence that power is people's main, objective and the primary feature of social relationships is too limited. One can hardly account for the behavior of the Pilgrim Fathers in terms of "self-interest" or the search for power, as conflict theory commonly uses those terms. Moreover, the way in which they define and discuss power leads many conflict theorists to imply that the whole of social life is essentially "zero-sum": that is, that one man's gain is by definition made up from others' equal JOSS.277 In fact, this is not necessarily the case. Suppose that we take a hypothetical self-interested ruler who wishes to get as much money out of his subjects as possible. He can either seize whatever he can find by brute force; or he can set up a well-defined tax system in which people know in advance exactly how much they will oweand that nothing more will be demanded. In the latter case, it is still ultimately the ruler's ability to coerce which ensures that his subjects pay up. However, they will also find it far more worthwhile to work hard, save, invest, and create economic growth than they will if anything they produce is liable to arbitrary confiscation. For that reason, the ruler may well, ultimately, do better for himself by choosing the less arbitrary course. Certainly, his subjects will fare much better if he does.278 Of course, no ruler ever has a completely free hand to choose one alternative or the other. However, societies can and do vary systematically in how far they provide the sort of environment which gives people security, encourages economic growth, or creates some degree of the "positive-sum" in their affairs. Contemporary conflict theory279 tends to ignore such important differences in how state power is exercised, and in how far it provides people with a secure and predictable framework for their actions. Consequently, it also tends to produce an inadequate theory of the state, and to treat laws-the product of the state-oversimpl istically, as though they were merely a reflection of group interest rather than systems with their own dynamic and influence.280 The same oversimplification is apparent in conflict theory's treatment of values and ideas. It is important to analyze the degree to which ideas are rooted in a social order and the ways laws and "ideologies" reflect people's interests; but it is also important to be aware-as functionalism is-that they have a degree of autonomy. Conflict theorists tend to treat ideas as though they were simply a reflection of the interests of the powerful; but narrow self-interest is often not a full explanation of events. Self-interest would have suggested the total extermination of the American Indians. That this did not happen was largely the result of notions of justice and morality which, however compromised, were universal in their application. Similarly, conflict theory tends to emphasize how ideas maintain stability,281 when in fact the ideas in a given society often criticize and undermine the current order. Christianity, for example, produced figures like St. Francis and Luther, whose teachings created massive social upheavals; and the Kremlin suppresses the Russian dissidents brutally because it fears their ideas so much. Finally, although conflict theory specifies a mechanism of change, it does not provide an entirely satisfactory account of it. This is because conflict theory is far better at explaining how a group maintains power than it is at showing how it acquired it in the first place. Collins, for example, argues that educational qualifications are an important source of privilege without saying very much about why they are now a more important resource than they were in the past. Yet groups do not acquire resources and power at random, and we had occasion to suggest earlier in the chapter that the origins of a group's power can often be explained by the services they provided. Educational qualifications may be used to protect and strengthen an elite group's position; but they are a more effective weapon now than in the past because education is also necessary to provide the technical skills on which modern wealth depends. The way social life involves the exchange of such goods and services, as well as values and coercive power, is emphasized by exchange theory.