Metta Spencer

Order, Authority, and the Sociological Paradigm

Center for International Affairs, Harvard University: December 1967

by Metta Wells Spencer
Cambridge, Massachusetts
December 1967


By all accounts, the most difficult problem sociology has addressed has been the problem of order. Given that each man is connected by his nervous system, conflict between his purposes and those of his fellows is manifest1y unavoidable. That there exists any collective organization at all stands as a factual challenge confronting the theorist Social order, resting on the willingness to submit personal wishes to the authoritative patterning of a collectivity, is a phenomenon which, though not ubiquitous, is taken as a remarkable and problematic event, whose existence must be explained.

The solution is the crown jewel of sociology, prized above all theoretical constructs. We all owe allegiance to it and it is part of a thought skill which we learn through practice and association. In fact, it constitutes the basic paradigm in terms of which it becomes possible to frame the more specific research problems. Any challenge to it by a well-developed alternative paradigm would amount to the kind of scientific revolution Thomas Kuhn has discussed with reference to the natural sciences.[[1]] Although it has provided the issue about which the most serious arguments have occurred in all political and social theory, it has remained the dominant position because no other model has been well enough worked out to constitutes a consistent alternative to it. Social psychologists, economists, systems analysts and game theorists have not put their heads together enough to learn how much they share in theoretical assumptions, and in methodology.

Let us outline this paradigm by recalling the major axioms which we take for granted by the time which we encounter first as neophyte sociologists which we take for granted by the time we enter upon actual research.

The first thing we learn is that it is a mistake at the outset to consider the individual as an isolated atom with his own independent urges. In effect, there is no problem of order but only a misunderstanding about the nature of man.

Man is, before he is a man a member of a collectivity a social being a cell of a social organism with an immanent organizing principle of its own purposes in accordance with which the individual's own conduct more or less fits.

(Allusions to the social organism are advanced more cautiously nowadays than are allusions to the social system, but the implications for the paradigm are identical). That the person's coincide with the social organisms internal requirements is not remarkable: ends do not develop individually, but in the context of social life not established prior to his social experience; their nature is not to be explained by idiosyncratic voluntarism, nor by "biological reductionism," but the requisites of the collectivity and it's own internal coherence. Personal ends are the products of socialization experience in an ongoing social unit. They are therefore normatively regulated and norms are taken to be external constraints widely recognized throughout the collectivity and justified by values shared by the entire society. These ultimate values are not individual, but socially shared in the first instance. The value system of the person's group points out for him the authority according to whom he must orient himself.

Let us give our mentors credit: the paradigm is elegant. It is successful in disposing of the problem of order (not in solving it, but in disarming it) and it squares better with our experience than did Hobbes' belief that social order rested on awe of political authority. The doctrine has been useful; nevertheless, it is subject to criticism from all sides. The critiques have mostly come as attacks on Parsons, but on1y because Parsons has done most to codify the assumptions of sociology and to work out the implications of the paradigm.

His detractors only throw darts at a bull. I am not a bullfighter either, but will at least try to plot the strategy he must use who seriously intends to take up the red cape. All the objections to current sociological metaphysics is the problem.

Perhaps by considering the critiques systematically we can learn what sort of alternative is required. Let us, then list the objections.


There are, first, a number of people who cannot live comfortably determinism of the paradigm, f or one of two reasons. Non-sociologists and novices in this medium get agitated in claiming more freedom for their wills than the model will allow. They protest that they choose their ends it is not a witless mechanical process at behaving in according to exterior fixed norms. Some theorists probably do derive comfort from thinking of themselves as a cell in the body politic, but more people react indignantly to the idea.

Peter Berger has recalled for some of us the psychological strain, the painful mind-bending that one goes through in taking on the sociological perspective. [[2]] Nevertheless, arguments protesting personal volition do not carry weight in sociology any more than they do in philosophy. Normative determinism is rejected by perfectly respectable sociologists on other grounds, however: the Davises, who are not touchy about their free will, remark that predictions based on normative determinism are misguided because there is a great difference between saying that norms are supposed to control behavior and that they do control behavior.[[3]] No pedestrian in a busy intersection is likely to imagine that the world's a stage and all the people merely players: there may be norms, but one cannot count on them. Accordingly, the theorist must guard against assuming that the existence of a norm is sufficient explanation of the conduct of individuals.

In line with this point, Homans and Coleman argue that norms are not explanatory but are patterns which are problematical and should themselves be explained.[[4]] Sociology would have more substantial statements to offer the world than it now does if its questions centered on why particular normative arrangements are worked out in specific circumstances. We should look at the "exchange model" which generates propositions for these men: it is the germ of an alternative to the standard-model explanation of social order. It does not proceed from pre-supposed order but attempts to account for particular kinds of group accommodations in terms very reminiscent of utilitarian action-theory. The kind of sociology Homans, Coleman, Blau, Schelling and March do, for example, looks and sounds different from that of, say, Parsons, Bendix, Eisenstadt, Smelser, or Merton -- men whose work rests more on the paradigm. [[5]] This is not because one set of writers is more macro than the other, but because the methodological stance they take differs. The exchange-model sociologist begins with "atomistic" actors, often without any historical discussions of the actors' orientations. Indeed, the actors may have no specified identity at all: an exchange-model analyst may say "let X be one actor and Y be another. They are each trying to do such-and-such, and these are some of the situations that may obtain in their relationship" The actors are hypothetical beings and the reader is invited to check the inferences by substituting any appropriate real actors he knows in the equations, persons, firms, or nations. In contrast, the second group of writers generally analyze actions by reference to the preestablished internalized orientations of the actors, other than in terms of the situational or structural dynamics of their transactions. The relevant orientations of the actors are assumed to be special values they have already learned to share through their membership in some group or other; what group and what values must be specified in order to make the "normatively determined" analysis.

The exchange model and the normative-determinism model thus heuristically single out different sources of variation in behavior. In the former case, such exigencies as those of information, resources, coalition memberships make certain strategies rational--and hence predictable. One looks for situational factors and tries to make formal, abstract statements about similar relationships. That is quite different from the analytic enterprise suggested by the paradigm, which postulates that individual acts are to be explained as normatively-prescribed role-performance in the service of a collectivity to which the actor belongs. For those who take seriously the argument against the normative determinism of the major sociological response to the problem of order, a shopping trip through the work of Coleman will afford suggestive alternatives.


The paradigm has been attacked by Dennis Wrong for permitting us to forget the impulsive side of man's nature.[[6]]

He charges the social expectations guiding our participation in human group life are made to seem so obvious and their adoption so routine that failure is not quite credible. There is substance in Wrong's observation that the problem of order has been disposed of so completely with this paradigm that it is now hard to explain any disorder, If ends arise as aspects of a collective endeavor, where do the impulses of the id come from and why are they so intractable? Socialization is arduous and internalization stressful as we all know privately, if not professionally. Yet our paradigm treats personal values as an "output," with social experience the "input", and our industrious socialization technologists are looking for new breakthroughs in new packages of input variables to yield high-quality products. While the endeavor is worthy, it may be salutary to recall that, not only is the flesh weak and the subconscious willful, but the mind is dissociative. That is to say, experiences that are "fed in" get associated with one another in new and incalculable combinations, so that the inferences drawn and the output behavior produced may differ from what was intended or foreseeable. Karl Deutsch has pointed out the great weakness in the possibility of forming firm predictions in the field of socialization research; he observes that "you can take a young boy, put him in a seminary and train him for the priesthood, and out comes Joseph Stalin, And you can take a little girl and raise her inside the Kremlin walls and out comes Svetlana."

Thus, one may take seriously the fact that the sociological solution to the problem of order is inadequate to treat any unsocialized or unsocializable aspect of man. It the theorist considers it important not to treat this aspect as a residual category, he can indeed find theoretical divisions to turn to G.H. Mead's "I", Freudianism, or developmental psychology -- but he will weaken the capacity of the model to explain social order to the extent that he allows any non-social, individualistic ends to have independent causal status (whether those ends proceed from biological instincts, voluntaristic goals, dissociative logical processes, or what not). At best, individual ends can only be grafted onto the basic thesis, not integrated into the explanatory part of the theory. There is as yet no sociological structure which puts all of these pieces together in a coherent way.


Proceeding along somewhat different lines, Bensman and Gerver have raised another critique which deserves notice. [[7]] They observe that the basic assumption must be made in the paradigm that individual lines of conduct are ordered in terms of a system of values and derivative demands which are shared throughout the collectivity. Ends are attributed to the social system, not to particular people. By starting with a collectivity already ongoing, the paradigm evades the hardest prob1em as to how the participants in a factory. This diversity is not papered-over and made tolerable by some into any joint enterprise. Bensman and Gerver will not allow us to escape the very real diversity of ends held by the various participants in a factory. This diversity is not papered-over and made tolerable by some pervasive and super-ordinate values, but is integral. The workers have certain objectives, the foremen others, the executives others, the stockholders others and the purchasers others. Actions which are illegitimate and antithetical to one end are necessary for other ends. Since the participants whose purposes require such contradictory norms themselves occupy mutually dependent roles, they must and do reach accommodations, keeping very delicate social control over the resulting compromises. Their collaboration cannot be attributed to shared ends, but more accurately to a negotiated "deal" based upon exchange, Some sociologists will not like this interpretation and will insist that certainly the factory personnel must have shared something, even if they did not report so. Perhaps so but it is verbal play to look toward presumed shared ends for an explanation of the social order to be found in the airplane factory they described, No doubt people share many abstract values (such as honesty, fairness, physical well-being, loyalty, and so on) and these values may be "talking points" for establishing agreements about specific standards of obligation and desirability, but the concrete demands certainly do not follow as the night the day from any consensus about abstract guiding principles. People's motives may conflict and the effort to harmonize them may require ingenuity and bargaining, and the "deal" they agree upon may or may not be eufuncticnal in terms of some hypostatized collective end, If these observations are accepted, the paradigm can hardly be regarded as useful,


Functionalism, resting entirely on the assumptions of the paradigm, is analysis which hovers very far over toward the pole of "pure theory" and very remote from app1ied sociology." Amital Etzioni has called attention to the inadequacy of functionalism to allow for any causal entry points into the system process which it describes. A po1icy maker is concerned with relations between variables and with prediction of consequences too, but his model of explanation is not as closed a system as is the traditional paradigm. To be sure, some collectivities do have some shared ends; they also have a set of priorities as to various ends and they have schedules of appropriate means. We should not deny such realities. The trouble is that the paradigm assumes that those ends have been established in advance. Actually, of course, collective ends are products of activity, not vice versa. The raison d'être of the paradigm is to render collective ends unproblematic, whereas the substantively significant empirical work is precisely concerned with problems where ends have yet to be formulated, about policy issues which are imperfectly understood, and not precisely calculable. The paradigm guides us away from analysis of possible alternative courses of development, Such a deterministic and static model does not provide any footing for a theorist who seeks a mechanism to effect social change, or who wishes to distinguish between those variables which can be manipulated and those which cannot. Functionalism suggests that nothing can be changed in the system because it is internally regulated: voluntarism would suggest that with enough pluck, one can change anything. A policy researcher, on the other hand, wants realism and needs an image of the social system in which not all elements are equally inexorable, but where strategic points of entry can be sought in the causal chain so as to effect particular results, It is important that the imagery we employ leave room for and encourage this sort of conceptual operation. Otherwise sociology will remain a world apart from other empirical research: right now it remains closer to philosophy than to the kind of socially relevant research that economics can produce. Revision of the paradigm should proceed in the light of this critique.


One of the highest developments of the paradigm can be seen in the contribution made by sociologists to a general understanding of relations of voluntary obedience, The paradigm in effect defines authority as the mystical glue that holds society together. Weber and Durkheim certainly converged upon some of the postulates of the paradigm in formulating their respective notions of authority, Weber's authority derives from a pervasive sense of legitimacy shared throughout the social group, not arrived at individually. Similarly, Durkheim also located norms outside the actor, acting upon him in the person of the social group. In all essentials, sociologists are still singularly committed to these views.

Still, there are occasions for dissatisfaction with the concept of authority which is basic to the paradigm, though the dissatisfaction is centered more within the fields of political philosophy and jurisprudence than in sociology. Weber's concept of authority is often criticized for it is radically culturally relative and is not the kind of concept other scholars have in mind when they attempt to develop better statements about obligation. His concept is psychological and nominalistic -- X has authority when others think he has authority -- whether the classification be charismatic, tradition, or legal, the justification for it is basically psychological. Ultimately it is not rationally defensible. This conception hardly serves as a. common term linking sociology to philosophy, because it does not allow any footing for making even an analytic distinction (and certainly not empirical criteria) between authoritarian relations and those in which commands are just, To be sure, Weber elaborates legal-rational authority in detail, but the thrust of his argument is that formal rationality does not guarantee substantive rationality or justice, and may import new miseries of its own, All authority rests on some legitimating belief, and thus all authoritative commands are equivalently just. Moreover, to give an order is as much an act of compliance as to accept one, since orders are assumed to derive from higher officials or from principles which are already formulated, The legitimating principles are not themselves arguable, This position is not only Weber's, but is shared widely by other writers on authority: certainly by Barnard and Durkheim.

Carl Friedrich has particularly vigorously criticized the irrationalism of the concept of authority suggested by the paradigm.[[8]] Genuine authority, he insists, involves a capacity for reasoned elaboration. Communications carry authority and are given credence because they come from one who is capable of offering extensive reasonable justification for the statements he makes, Personnel come to hold office because they speak with authority, not usually vice versa. If they lose authority while in office, they may either resort to power or find their position dwindling to that of figurehead. This point offers a possibility for avoiding the dead-end cultural relativism of authority as implied by the paradigm. That is, while the paradigm suggests that authority exists in the shared beliefs and values of the group, Friedrich's conception would allow that authority may exist in the communications of people who can argue reasonably in terms of the hearer's own belief and value system -- however it may originate --and it may not be culturally defined nor even shared by the speaker. As an extreme example, some psychiatrists are able to offer effective (authoritative) guidance by entering the psychotic's world and using his delusions to validate their psychiatric interpretations. Such psychiatrists would be exercising authority in Friedrich's sense, but not in terms of the paradigm which would allow for authority relations only on the basis of shared values, cultural perspectives, beliefs, and common understanding in the group at the outset.

The paradigm thus carries a recommendation as to the preferred time order for analyzing authority: the group is to be taken as prior to the individual, logically as well as temporally. To reverse that time order is to start from individual actors and account for how authority is built up. Such a reversal would allow sociologists to speak the same language as rationalists do, who think rational authority is demonstrated by successive competence in producing logically consistent statements of some substance, possibly (though not necessarily) relevant to a means-end relationship.

An emphasis upon the rationality of authority, moreover, would sustain a new attack upon the cultural relativism to which sociology has been unwillingly committed. If, as the paradigm would have it, authority derives from some already-established value system peculiar to each society, then there is a fundamental discontinuity between the societies of men, a basic disunity among men which cannot be bridged by cross-cultural dialogue. Values cannot be augmented nor advanced in any significant sense. The difficulty of living with this relativism seems to have been the great burden in Weber's intellectual life.

Outside of sociology, however, the radical cultural relativism of the paradigm has never been well received and even now we must take note of such contemporaries as Rawls, whose analysis of justice has revived interest in the possibility of a priori reasoning along ethical lines.[[9]] He argues that justice consists of such principles as "Junior may cut the cake and Betsy may choose the biggest piece" i.e. principles which can be seen a priori to give advantage to neither participant.

To bring Rawls' reasoning to bear on authority we must fit together the elements of several critiques. An authority (in Friedrich's sense) is a person skilled in reasoning along such a a priori principles as Rawls outlines (though, of course, that is not the only component of good judgment). A society conserves and institutionalizes such handy maxims, as "you cut the cake and I'll choose the best piece." These a priori principles are understandable cross-culturally and may be discovered independently as well. Aptitude for such reasoning, while it cannot be reduced to socialization in the sense of learning the rules of one's group, does develop through social experience: only in that way can the subject learn to take the abstract general perspectives (two here: cake cutter and first-piece-chooser), so as to compare them and conclude that neither has the advantage, Ability to elaborate reasons within one's own sphere of competence is thus a socially developed experience. Both Cooley and Mead pointed out the importance of the peer group in developing fair-mindedness, a generalized perspective (taking the "role of the generalized other", Head called the process) and ability to formulate principles which can be assessed abstractly.

Though moral reasoning develops socially, it does not develop as internalization of shared group principles, which is the meaning of "socialization" in terms of the paradigm. Authority, to extend Friedrich's discussion, is the granting of discretionary powers to persons in recognition of their special capability of reasoning abstractly. If, as Rawls leads us to conclude, some abstract principles are verifiable a priori, then at least some principles do not derive their validity from group consensus, hence are not culturally specific. There lies the escape from the radical cultural relativism which sociology has been forced to accept, but the escape mechanism leaves a hole in the paradigm: specifically, the axiom that socialization consists of incorporating group ends.

Socialization, the development of regulation in conduct, is something else altogether, and no one has said so with as much clarity as Piaget.[[10]] His developmental theory about children's moral capability was formulated specifically to counter Durkheim's conception of authority, which is the quintessence of the paradigm, Piaget cites Durkheim's statement that "Society alone stands above individuals. From it therefore emanates all authority, It is society that bestows upon such-and-such human qualities, that sui generis character, that prestige which raises the individuals who possess it above themselves, " Piaget is at pains to dispute this, pointing out that there is no society with a capital S, only social relations, And indeed, the sequence of development in the child proceeds from an initial bond of personal respect for the particular parent and for his rules to a relatively late phase of recognizing the parent as an incarnation of society, or perhaps a generalization and abstraction of rules.

Piaget suggests that constraint and cooperation are two different modes of producing social order -- not one and the same thing, as Durkheim held, Constraint, an analog to the unilateral respect relationship of the child to the parent, comes first and is a more primitive moral development than is cooperation, which comes only as mutual respect relationships are developed. Drawing upon Bovet, Piaget describes how the child develops a rational sense of authority based upon self-government -- not by incorporating the laws which "society" imposes on him by constraint, but by himself taking part in the cultivation of rules. Reason works on the commands, extends them wherever possible, reconciles their contradictions and attempts to construct a consistent and harmonious moral universe. The result is a mind "capable of elaborating reasons"; authority in Friedrich's sense, It is learned best not through relations of unilateral respect and constraint, but through mutuality, cooperation and joint activity. Paralleling Cooley, Piaget holds that the peer group is the context in which more advanced styles of moral reasoning are learned. Thus, irrational authority is fostered best in those situations which impose lesser authority in Durkheim's sense.

Piaget's final chapter in The Moral Judgment of the Child has been widely read, but few people have followed its implicit recommendation to turn the paradigm on its head, thus asking authority, is built up among those who do not share common principles of legitimacy at the outset, This is a problem of strategic interpretation as much as a problem for developmental psychology. The question can be approached in the conceptual framework Neustadt proposed in Presidential Power: by successive demonstrations of effective authority a kind of credit (or credibility) is built up which can be used as a resource.[[11]] Neustadt had in mind the credibility of attributed personal power, but the same assessing and checking process goes on also in establishing a man's credit rating in rational authority. In any case, Neustadt's analysis has not been extended into such parallel kinds of situations, where sociologists might find it useful.

The critiques of the concept of authority suggested by the paradigm deserve, we consider more attention than commonly given within the profession. Sociology is none the better for limiting discourse to positivistic treatments of authority, as so commonly occurs. The paradigm's axioms are metaphysical postulates, and they deserve no more ideological commitment than any other doctrine, when their limitations become apparent,


We will go further and suggest that the paradigm is one very large reification. That is what Piaget meant when he said that there is not Society with a capital "S", but only social relations. Reification seizes upon a term which names an aggregate of individual events ( as "society" is a summary term denoting a number of individual people who have some relationships among themselves) and then uses that term as if it were a substance with causal power ,those individual events ( as "society" makes its members do such and such). The paradigm we have criticized is an invitation to that sort of theoretical magic show. It provides a whole vocabulary of reifications, which in turn give rise to theoretical contradictions from which decent men spend years of their lives trying to extricate themselves.

Let us take one simple of the mischief reification has caused, in this case beginning with the paradigmatic rendering of the term "value". Non-sociologists normally assume that they value things independently: if several of them come to assess the same object

similarly, well and good, but it is not to be expected. Some non-sociologists have even elaborated learning theories to describe how each person has acquired his own distinctive tastes or values. Gordon Allport, for example, reminded us of the "functional autonomy of motives," such that in the course of satisfying a rather primitive urge, certain instrumental activities may be taken up and found pleasant in and of themselves. Those means may next become ends, for which other novel means may have to be employed, and so on, until one has a personal organization of motives, some of them proclivities which are not native in any sense whatever. This need not be an atomistic conception for people are often able to influence the outcome of other people's actions and the resulting social interaction and exchanges produce joint activity and mutual social control, whether conflictful or orderly. In such a model, "values" refer to preferences and priority schemes developed by specific people.

That is not what the term means in the sociological paradigm. Quite the contrary: "values" refers instead to the preferences and priority schemes of that great being "Society," who "alone stands above individuals," as Durkheim said. Here, insofar as individuals have values, they partake of some external values, and to explain this process is the hard lot of the sociologist. A value being the property of an aggregate before it is the property of an individual, the sociologist needs to locate the relevant social aggregate from which the individual acquired his values, Since there may be many different groups in any one territory, a very interesting theoretical problem has been worked up for the people in the field of deviance, namely the argument between the anomie theorists and the cultural diversity schools of thought. Anomie theorists see misbehavior as falling short of ideals and values which are shared throughout the society but which people have differential means to pursue. The opposite position is possible if one locates subsets within the society, such as different ethnic groups, and then argues that each one has its own autonomous values. Misbehavior, by that recording, may not be deviation from society-wide values and ideals but rather behavior which is very much promoted and demanded by the values of the specific "subculture" to which the miscreant belongs.

By new the debate on this subject has become stale and it may not spoil. the game to remind the players that both sides are resting their bases upon reification as justified by the paradigm, They are imagining that society is a large real Venn diagram with several subsets inside. Only then can they ask whether the actor got his values from his set or his subset. If the actor develops his own values independently, the issue has no meaning, Values are properties of individuals, but if a group has a "value" it is only in a derivative sense, as "40 per cent of the first precinct supports our candidate", and it is quite hard to find values which can be said to characterize either a set or a subset as a whole. When we want to know how a person actually did come to develop his preferences, we have to inquire about his specific relations with specific people, who agree with him in varying degrees and who are willing to accept differing ranges of rationales for conduct. One need not look for a reified social system, The "value" of a group (if and when it exists) is a coefficient summarizing the distribution of individual preferences in that group and is not a reality sui generis affecting the values of the members. To speak of a value of a group as explanatory is to indulge in the "fallacy of misplaced concreteness."[[12]]

It will be objected that to do otherwise is a worse sin reductionism. To consider values as initial individual properties rather than group properties is certainly methodological individualism and if that is reductionism, then one is forced to choose between two fallacies and which one to choose becomes a matter of personal taste, subject to the considerations which we have raised here in objecting to the reification paradigm.

Actually. it is not altogether necessary to choose one of the other of these two fallacies and yet be able to refer to the "single mind" of the collectivity consensus. There is a middle road which can treat emergent properties of groups as real and important but without either reification or reductionism, We want to be able to give full weight to the fact that people can and do take public action to stipulate their commitment to a collective decision. This may be for the specific purpose of lending their joint decision a degree of exteriority and constraint, making it a "social fact," as it were. Laws may be promulgated with fanfare so that they become indubitably "facts of life." On the other hand, people may acquiesce to a consensus which is in fact false because each member is ignorant of the extent of dissent actually within the group. In both these cases, the manifest collective consciousness is a variable to be reckoned with, apart from the opinions of individual members. Thomas Scheff has described how this social phenomenon occurs, in which the mutual estimation of one another's beliefs and values (whether correct or not) constitute a fact in the world to be taken into account.[[13]]

His important contribution does not, however, mean that the collective outcome is independent of individuals, He has, in fact, produced an account of emergent group properties which is methodologically individualistic without being reductionistic. That so many others had neglected to do so left a lacuna which the paradigm had been used to cover up.


If we take seriously the critiques of the paradigm which have been developed here, it seems that the basic model of explanation which is standard now will require revision in the future. We do not assert at ail that because the paradigm is deficient; most of sociological research is deficient. Much research is not generated by logical questions set by the paradigm, but rather by puzzling problems in the real world, which is where research should start, However, some research does proceed in terms of the logical system of the paradigm and it is generally the worse for it. We need name no one in this connection, but pass on to recommending a list of considerations to be borne in mind whenever the paradigm is to be invoked. and especially when alternatives to it are being entertained. There is no one consistent alternative covering a wide enough scope to apply throughout sociology and yet to generate specific hypotheses, but there are some approaches, such as exchange theory and symbolic interactionism, which if pieced together may begin to form a paradigm capable of challenging the old one. Even new we can stipulate some of the requirements a new paradigm ought to meet.

  1. 1. It would be methodologically individualistic, so that the purposes of the several actors would have logical priority over the purposes of the collectivity for analytic purposes. This is a little like stipulating that the chicken shall be taken as prior to the egg: obviously no one knows which way I went temporally, but the main reason for establishing the logical priority for analytic purposes is to avoid the multiple difficulties presented by functionalism, which treats the collectivity and its system requisites as logically prior to (and explanatory with regard to) the activity of the individual. Another reason is that methodological individualism seems to allow for more determinate causal statements to be made. One of the Oxford philosophers said that an adequate statement of causal relationship is a recipe: we have included enough off the right information in the formula when we can say that these ingredients in that mix will have that outcome. It is safe to say that sociologists are better at predicting collective action by knowing the purposes of individuals than at predicting individuals' actions by hypothesizing about the purposes of a collectivity. Groups and individuals go on and on endlessly, just as chickens and eggs do, but more determinate propositions result by starting with the individual.
  2. 2. The new schema would not rest the whole case on socialization, the incorporation of external rules into the personality, though at the same time it must be able to give an account of socialization without slighting it and without depicting it as irrational and mechanical. It would admit of the distinction which Piaget rescued from Durkheim -- the distinction between the morality of constraint and the morality of cooperation, Piaget was rightly concerned that the sociological paradigm made conformity the measure of morality and gave no conceptual apparatus for describing the moral heroism of a Socrates.
  3. 3. It would have to treat norms as the product of interaction, not simply the starting point presupposed for action, But in so doing, it must not evade the problem which the paradigm avoided the problem of order. Happily, certain contemporary work is most useful in approaching that subject. The most familiar theories which treat norms as the product of interaction are the social contract theories, including that of Bentham and other utilitarians. Persons would have us believe that the utilitarian school's failure to account for social order has been established as a final fact.[[14]] That is not the case. The most promising work done in political theory today is very much within the utilitarian framework.[[15]] Thomas Schelling's work is particularly relevant to Hobbes.
  4. 4. It would treat values as principles of personal organization and ends as individual motives and purposes. But on the other hand, this must not be accomplished at the price of making every man seem an island and at the price of ruling out terms of reference to coalitions, complex organizations, end other emergent social patterns. What has to be ruled out in the reification of such collectivities.
  5. 5. It would allow for the possibility of explaining collaboration among men who do not share common values. When Crusoe and Friday met, social interaction was able to proceed in some fashion and the two men were able to build a collectivity unexplainable in terms of any previous group to which they had belonged. We must be able to speak of the development of new collectivities and institutions.
  6. 6. It would renew the discussion of authority and link it to the discourse in sister disciplines by a) allowing it to be a dependent variable again and (b) taking seriously its normative statue, thus allowing sociologists to discuss such subjects as the perversion of authority in authoritarian relations. At the same time, the new paradigm must not reduce any of the other properties of authority to a less differentiated set of meanings -- i.e. one must still be able to distinguish analytically between power, authority and influence.
  7. 7. It would maintain the possibility of analysis of interest groups and conflict groups without trying to reduce their diversity to a pseudo-harmony.
  8. 8. It would link academic journals to the discourse of planners and men of affairs by treating social action as primarily rational. It would make possible discussions of policy by using a model which allows analysis of the payoff values of alternative sets of logical possibilities, rather than the present closed deterministic model of a functioning social unit with a pre-set thermostat inside.


The reader may wish to consider how one can "speak sociology" without using the concepts of the paradigm, We shall try to show that it is possible. But we are not in a period of scientific revolution, in Kuhn's sense -- at least not yet. We can mock-up a dummy counter-paradigm and play with it for demonstration purposes, but it is not complete and there is no crisis forcing a choice between the paradigm and the counter-paradigm, Nevertheless, the work of counterposing the two models may be productive.

It is possible to get along without the paradigm, simply by referring to the relations that are familiar in social life. Our counter-paradigm would begin with the simplest social unit-two interacting human beings. Not everything, but a great deal can be understood at that level of analysis.

The individual, in this counter-paradigm, is the author of his own goals, values, ends. Everyone experiences a continuous flow of impulses from the moment he is born, It is how these impulses get organized and channeled that constitutes the theoretical problem. The explanation does not require biological reductionism, nor any other device for finding in the past the seeds of the present and future, Goals are built up along with acts in the very course of living. Goals are not past events but present-time images of future anticipated states of affairs, anticipations elaborated in the course of action. The person considers a great range of future possibilities, and may elect to act purposefully with regard to some of them, while discarding any others because they are unappealing or unfeasible. But, as Dewey suggested, goals are like targets and one does not shoot arrows because the target is there; rather, one sets up targets to shoot at. One moves into a new phase of action by setting up new targets: normally (but not always) targets may also be abandoned. In contemplating probable future outcomes many efforts may be rehearsed, many prospective pleasures savored, then found disappointing and forgotten. Most of all, one drops projects on the basis of the anticipated responses of others. Social control as prospective sanctions penetrates the earliest covert phases of action, One checks one's deed before the thought is half formulated.

Actions may redound upon the actor; by acting one may create conditions which open up or limit future options for action. One may inadvertent4y create undesired circumstances for one's future; one may paint oneself into a corner. Sometimes one makes an effort to bind oneself to a future intentionally: in order to create necessary credibility to make joint activity possible, one may seal contracts, make promises and issue warnings about anticipated sanctions so as to deter undesired actions on the part of other people. On the other hand, one may bind himself unintentionally. This is "commitment." By acting now we affect the payoff values of various logical possibilities open to us in such a way that we can no longer rationally afford to make the choice that we might like to make. Painting oneself into a corner is one example. Becker has cited another: by working temporarily in an undesirable job, one nay accrue considerable money in a retirement fund which can only be collected at the end of one' s career: subsequently a better job may open up but by then the retirement money to be lost is so much that one cannot afford to do what one would prefer.[[16]] When factors such as this occur to large blocs of people, we speak of these effects as "structural."

This phenomenon is matched by the commitment of the kind that balance theory treats: by committing oneself publicly to a certain goal or point of view, retreat is made more difficult and one may experience great difficulty in living up to his expressed position. (An acquaintance of the Nazi, George Lincoln Rockwell, said his tragedy began by having been taken seriously while making absurd statements and thereby condemned to strut publicly in his ridiculous role until he dropped.)

What we are saying is that normally one may set up goals and drop them at will, but not always. This possibility of binding oneself to a stated intention may be either useful or tragic, but it is of utmost importance in explaining the regularities of social life, The legal right to bind oneself in contract is normally an extremely valuable asset, but it is a right which the signatory may curse himself for having used. More commonly, one may find himself for one reason or another unable to drop two incompatible goals and the resulting psychological conflict is familiar to everyone.

The possibility of commitment presents a fruitful concept for addressing the Hobbesian question, and one can see it best employed in Thomas Schelling's book, The Strategy of Conflict.[[17]] Hobbes would have us believe that social life is intrinsically a zero-sum game, such that any gain by one member is matched by a comparable loss by another member. Other more sanguine social theorists suggest that social life is inherently a cooperative enterprise, such that a gain by one member is a gain for the whole social system. Schelling instructs us on this point: most social affairs are better considered as "mixed-interest games" -- that is, there are both grounds for collaboration and grounds for suspicion, The terms on which relationships are to proceed are subjects for strategy. If ail participants perceive that each has a great deal to gain by destroying the other, social order cannot be taken for granted. Each may recognize the potentiality for valuable collaboration, yet have excellent reason for mistrusting the other. It is in situations of this kind (which resemble the Hobbesian state of nature) that promises and normative expressions are not explanatory f or they are not generally adequate to enforce the peace. One wants not promises but guarantees. The effort then, possibly by all partners. is riot to alter the negative intentions but to alter the basis for the negative intentions. Each must not say, but make it contrary to his own interests to harm the other and must make it apparent to the ether that this is se, Only in that way is edibility established and the grounds for mistrust "bound off' so that the collaborative relationship can proceed. Wishing does not make it so; sharing of cultural values does not make it so; one has to find some way of actually altering the payoffs of the different logical possibilities open to the actors in order to remove the temptation to violence. In order to do so, one may make a "side bet", e.g. authorize some third person to .impose a penalty for breaking the agreement. We wish to suggest that such a situation is the simplest transaction in which anything like authority or government can be found. Government is, of course, many other things, but it is first of all an institution which enables private individuals to make 'side bets' of such a kind as to bind off certain possibilities which are attractive enough rationally to jeopardize important collaborative relationships which may demand sacrifices so intense that they cannot be counted upon. Two or more persons may postulate an anticipated joint activity and seek to bind themselves to it in the face of interfering conflicting interests; by authorizing a third person to coerce them when voluntary compliance fails, they create legitimate power and a kind of rational authority. Sometimes no basis can be found for neutralizing the very apparent source of conflict; in that case, social order may indeed fail, but it is the task of government to promote the arrangement of carrots and sticks in such a pattern that collaboration is more to the advantage of various social groups than is hostility.

At this level there is no distinction between public and private authority, for we have spoken only of authority as explicitly conferred by the parties so as to enforce a private contract. We should not suggest that this is the only basis for authority. Probably as important is the use of public authority by third parties as protection against the politics of external costs.[[18]] That is, two actors may find mutual advantage in collaborating in a way which has side-effects injurious to neighbors who are not themselves parties to the transaction, In such cases, the effort will be to create a recognized authority capable of restraining such freedom of harmful contract, In any case, here as well as in other situations, authority is exercised by appealing to ends which the actor is no longer free to abandon.

One may, indeed, appeal to principles which the actor is no longer free to abandon. Clearly not all authority rests upon ad hoc contractual accommodations, but we are not forced on that basis to conclude that the residual instances of authority rest on inherited and internalized social norms, one may not be "out there" in the social world as inescapable and pre-determined features of the environment, but a great deal of principled argumentation goes on nevertheless. As we have seen, people seek to generalize and bring some consistency and predictability into their affairs, This implies an activity of the kind Becker has called "moral entrepreneurship," People seek to justify their requests by reference to abstract principles which will do just enough to suit their purposes and not more. From past experience (and perhaps from a special knowledge of legal enactments) they are familiar with a "moral or legal vocabulary" of principles, maxims, and precedents. However, by invoking an abstract principle. one "votes" for its legitimacy and this act cannot always be undertaken lightly, for abstract rules are two-edged swords arid next time may be invoked against oneself. Intelligence comes into play in the selection of guiding principles: far from its being the case that there is a single value-system with self-consistent norms within a social system, there are as many values at hand as one would care to seek out, With values and norms as with proverbs, and usually three or four rationales available to justify any course of action, and just as many rationales to justify the course of action. The limiting factor is not the decisive quality of the norms, but rather the unwillingness of the participants to commit themselves extensively to any and all conceivable principles. For much the same reason that the Supreme Court refuses to rule upon a. matter until it absolutely must, the ordinary human being learns to be judicious in employing very general principles and calling for moral probity.

One trips too easily over one's own normative innovations. There are, of course, many people who have private reasons for promoting new norms, but usually people invoke only those principles which they consider to be, in fact, quite general and which they believe they can live with when applied to them.

Either by invoking the principle or in some other way, a person expresses commitment to it. One acknowledges its general fairness, It is this which opens the possibility of authority, for one is no longer free to abandon such a principle and one cannot protest against its application. That the fairness of certain principles can be established a priori also makes it possible to establish authority in persons as well, for one can assess how well an individual reasons in a certain domain and one can then give him appropriate discretionary powers and judge how well he uses them.

There are several analytically different relationships through which human beings fit together their acts. First, it must not be taken for granted that they do accomplish this: social order is empirically problematical and thus should never become theoretically a sure thing. When conflict occurs, the first mechanism may be to make new commitments which will create a mutual interest in collaboration as well as a sanctioning system to deter disruptive activity. This arrangement is not automatic, however; it may be costly or technically unfeasible. It is the office of present-day Marxians to remind sociology that conflict in this world is real.

The second relationship which integrates the activities of "atomistic" individuals a bond of sentiment which favors cooperation. This is the mechanism of the traditional concept of socialization, resting upon an identification with the collectivity more than with the self. In criticizing that as an explanatory construct, we ought not suggest that people can never value others as ends in themselves. In the course or interaction, some persons become valued objects, around whom additional activity gets organized. We must insist that this is not to be assumed. People who love one another may coordinate their activities with such glad spontaneity that the question of subordination never arises. At the level of secondary stations, on the other hand, sentiments are much less relevant and the organizing principles have to be impersonal. In coordinating their activities in that setting, personalities impinge upon one another in quite a different way from the affective relations of an intimate group,

Thirdly, persons "influence" one another in the secondary group and the considerations are principally instrumental, Where coordination of activity becomes problematical, so that one or more persons must change plans, the effort to influence actions becomes quite explicit. One searches for goals to dangle before the other. What means end relationships can be demonstrated convincingly that, in his calculus of felicity strike a happy chord? "Ah, now that I see your purposes, sir, I believe I have a plan that will work better for you than your own, in the long run." And so influence limits itself to reasoning in terms of the other man's logic and goals in a context in which the other is free to shift his axioms or drop his plans. "You say you hope for B? But you are going about it in the wrong way; surely C will result instead ." "Then I shall hope for C instead." To this reply influence must fall silent.

Authority, however, has the ability to call up goals and principles to which the other one is committed and which he is not free to shift. It is analytically separable as a fourth mode of relationship imposing social order, from the fifth and last mode, power. Authority uses one's integrity as a lever against him: one is bound by his own voluntary commitments. Power also has a compelling quality, but (except for authorized power, which is authority), it is not voluntary. The power-holder is in a position of instrumental importance in activities of the less powerful one and can give more in the relationship. To be sure, if the subordinate should elect to drop the goal then the source of power would evaporate, but if it is power, the power-holder knows that the goal will not be abandoned. The relationship between him and his subordinate and is based on the structural exigencies of their situation, not upon any translation of a generally applicable rule into a specific call for compliance, as is the relationship of authority.

All these modes of integration are commonly self-administered social control. Even with power, the most involuntary basis for compliance, the subordinate constructs his own approaches to the power-holder, alert to the possibility of retaliation, but nevertheless active. It is in this regard that the model we have outlined here differs most signally from that suggested by the paradigm.

The counter-paradigm is more commonsensical than the paradigm. It is rather un-sociological, but un-sociologists count for something in this world and are worth addressing from time to time. We have avoided many of the difficulties of the paradigm while rendering an account of most everything it explained. The mast interesting theories of other disciplines begin with a model not unlike the counter-paradigm. To be sure, it is a modest little thing and has its shortcomings: the most apparent weakness is that it does not adequately treat expressive or deep psychological complexities. But withal, it is a good starting paint for a scientific revolution.


[1] Thomas Kuhn, The Structure of Scientific Revolutions, University of Chicago Press, Chicago, 1963.
[2] Peter L. Berger, Invitation to Sociology, Anchor Books, Doubleday and Co, Garden City, 1963.
[3] Judith Blake and Kingsley Davis, "Norms, Values and Sanctions," in Handbook of Modern Sociology, Robert E.L Paris, ed. Rand McNally, Chicago, 1964.
[4] George C. Homans, "Bringing Men Back In," American Sociological Review, December 1964, p. 813. and James S. Coleman, "Collective Decisions," in Sociological Inquiry, 34 (1964), pp. 166-180. Coleman's paper is an absolute masterpiece.
[5] We can appreciate the sputters some of these men might utter on seeing how they are grouped, but we seek to highlight only differences in whether they begin with or end with group properties. Smelser belongs in this latter group despite his value-added model, because one of his components is a generalized belief, which is a shared orientation.
[6] Dennis Wrong, "The Oversocialized Conception of Man in Modem Sociology," American Sociological Review, 27, pp. 184-193.
[7] Joseph Bensman and Israel Gerver, "Crime and Punishment in the Factory," American Sociological Review, August 1963.
[8] Carl Friedrich, "Authority, Reason and Discretion," in Authority, Nomos I American Society for Political and Legal Philosophy, Harvard Press, Cambridge, 1958, p. 37.
[9] John Rawls, "Legal Obligation and the Duty of Fair Play," Sixth publication of the N.Y. University Institute of Philosophy, 1963.
[10] Jean Piaget, The Moral Judgment of the Child, Free Press Paperback, 1966, p. 340.
[11] Richard E. Neustadt, Presidential Power, Science Editions, N.Y. 1962
[12] Though the argument between the anomie theorists and the subculture theorists has dominated the field of deviance, it need not continue to do so. Thomas Schelling has broken new ground and writers are hereby urged to plant in his field and let the value-reification problem lie fallow for a while. See "Economics and Criminal Enterprise," The Public Interest, 7 Spring, 1967.
[13] Thomas J. Scheff, "Toward a Sociological Model of Consensus," American Sociological Review 32 (1) Feb. 1967.
[14] Talcott Parsons, The Structure of Social Action, Free Press, N.Y. 1937 page 102.
[15] As examples, we would refer to the work of Anthony Downs, whose economic model of democracy is particularly fruitful, to Shubik and Riker, whose focus has been on coalition formation, using game--theoretical approaches to Olson's work on the logical nexus of collective action, to welfare economics as applied to politics by Kenneth Arrow, to Simon and March's work on power and structural relationships, as well as to simulation techniques, and the approaches of "industrial dynamics" and "operations research. To insulate sociology from these approaches is to miss the twentieth century.
[16] Howard Becker, "Notes on the Concept of Commitment," American Journal of Sociology, 1960 (66), pp. 32-40.
[17] Thomas Schelling, The Strategy of Conflict, Oxford University Press, N.Y. 1963.
[18] James Buchanan and Gordon Tullock, The Calculus of Consent, University of Michigan Press, Ann Arbor, 1962. Introduction and p.p. 189-201.