The Problem of Political Theory

Thirty years ago, in the conclusion to her study of Ludwig Wittgenstein's significance for political philosophy, Hanna Fenichel Pitkin posed a dilemma arising from Wittgenstein's transformation of the philosophical method and the impact of this change on traditional political theory. Pitkin argued that traditional political theory had been inherently tyrannical. Plato solved the political problem “by eliminating politics” and theorized justice “by eliminating the need for, and thus the real nature of” justice. But leaving aside Plato, often criticized for totalitarian tendencies,¹

Pitkin found that even Jean-Jacques Rousseau and Karl Marx, both committed to the concept of human liberation, followed the same pattern: “[T]here seems to be something in the enterprise of theorizing itself that makes the resulting system seem totalitarian and in that sense nonpolitical. The theorist stands outside the political system about which he speculates and writes; of necessity he deploys and manipulates its citizens without considering their wishes or opinions.”² Attempting to create intellectual order, the theorist imposes order on individuals. It is, Pitkin concludes, not easy “to have an overview of political life, without also seeing in a … basically nonpolitical way—without seeing other men as objects and oneself as the only person.”³

This is the classic issue of the relationship between knowledge and power, and it relates not only to classical political theorists but to current methodological issues in political science. While the behavioral revolution's rift between political science and political theory leads us to separate the normative, the positive, and the empirical, Pitkin's dilemma transcends these divisions, reaching to the basic issue of human knowledge about human affairs and to theory's innate tendency to tell people how they should see the world and their own place in it.

The dilemma arises, however, not because theorists impose order. If that were true, the solution would be to simply stop engaging in political theory altogether.⁴

But this would fail to solve the problem because, as Pitkin writes, “Without some foundation in stable truth, men are unable to orient themselves toward the world or toward each other and hence lack a stable sense of self.”⁵ Her question is whether a Wittgensteinian theory—a “piecemeal” theory that emphasizes “partial overviews” and ad hoc solutions—can be intellectually satisfying and provide “the sense of coherence and mastery” people need in their political lives and actions.⁶ Pitkin sought escape from the dilemma through philosophy, but found it inadequate. Instead, she turned, on the final page of her book, to Wittgenstein's idea that the problems of the human condition can only be solved by individual actions. “If [our concepts or our habits or our institutions] are to change, we must change them in our actions, in our lives; and ultimately that means we cannot change them in isolation.”⁷ “A Wittgensteinian approach … requires of us, that we take other people, and other cultures, seriously, that … we become able to see from the perspective of another. But it also makes possible, and requires of us, that we take ourselves seriously….”⁸ What is required is a framework within which such exploration can proceed, one that provides meaningful guidance without imposing predetermined answers.

Pitkin's challenge has not gone unnoticed by political theorists.⁹

Yet to my knowledge no one has proposed a concrete empirical approach to the problem, and it is toward this goal that the present essay is directed. I suggest that a major barrier against the development of an individual-level, nonauthoritarian political theory and political science appropriate to the democratic experience is the conventional assumption that “the state” is equivalent to “the political.”¹⁰ If political theory is limited to the study and justification of the state, then it is very difficult to see how citizens, in their very narrow civic roles, can be anything but subordinate players. But if we accept the fact that political experience is present in everyday human relationships, then the field of political science in general, and of political theory in particular, expands in interesting ways. Individuals, with all their idiosyncrasies, replace abstract, neutral citizens; a wide assortment of human behavior, from the high-spirited to the underhanded, replaces the right to vote; and individual self-government replaces state-centered hierarchical control. Such a viewpoint replaces an abstract and skeletal political system with fully defined political people, participating in environments that, while having no direct relation to the state and its institutions, are nevertheless political in their determination of “who gets what, when, and how.”¹¹

The idea that strategic interactive behavior is the central phenomenon of politics is not, of course, new. However, developments in the modern social sciences and in social philosophy provide us with the opportunity to pursue the approach with new vigor. I am not proposing we become anarchists. The state is a fact and will continue no matter what philosophers say about it; political theory will continue to pay the state appropriate attention; and citizens will continue to exercise their rights and duties under the state. But for the average person, the modern democratic state, whatever its size, leaves a large amount of empty space. It is often argued that this “private” space is our own, our locus of perfect freedom, our arena for self-satisfactions of whatever sort we choose.¹²

Against this conventional interpretation, I argue that this supposedly empty, personal, nonpolitical space is actually political, and that citizens (in their apparently private lives) are constantly engaged in power relations with one another. These power plays do not occur in abstract conceptual terms, but in empirical and sometimes unseemly political practice, where sharp strategy may prevail even in the most private games.

To explain why this change in viewpoint perhaps seems unexpected requires a brief incursion into the theory of democratic politics. Modern political theory has been defined by its focus on the state as the essence of the political ever since the works of Thomas Hobbes, John Locke, and Rousseau.¹³

Max Weber reinforced this definition in the twentieth century with his idea of the state's having the right of legitimate force over a geographic area,¹⁴ and it entered behavioral political science with David Easton's definition of politics as “the authoritative allocation of values for a society.”¹⁵ Easton's otherwise very broad and useful definition is made conventional by the final phrase—“for a society,” which seems to restrict “allocation” to the level of official government. Recently, the American Political Science Association reaffirmed the centrality of the state in political science through a decade-long survey of the discipline that used the state as a major thematic focus.¹⁶ While postmodernists have sought to deconstruct the state, and historians and political scientists have complained about its unacceptably royal nature and its negative influence on political research,¹⁷ it retains its centrality as the focus of the political.¹⁸

This consensus frequently involves a second dimension: that the state is preferable, or “the solution,” to politics.¹⁹

Yet the well-documented decline in voting, not only in the United States but in other advanced industrial democracies, suggests that many citizens have lost some of their enthusiasm for the government as an object of democratic participation.²⁰ In large part this attitude may result from the growing sophistication of the electorate. Under the aegis of the modern liberal democratic state, citizens have had greater access to economic, social, and educational resources during the past half century. From this new and privileged position they have been able to scrutinize their institutions more carefully than did their forebears and have often found their institutions lacking.

These sociopolitical changes suggest the possibility of expanding the scope of the democratic experience by helping us recognize that politics occurs universally in human relationships.²¹

Defining politics thus more broadly provides the key to solving Pitkin's dilemma: we must articulate a political theory that provides intellectual order and guidance to individual activity, yet is integral to the participants rather than imposed from outside.²² My central framework is that of an open-ended strategic process through which individuals attempt, with uncertain chances of success, to define themselves, their goals, and their place in society. The process is political in that it allocates values, but it is controlled by no outside authority; individuals create it by their own efforts, from the bottom up.

The contemporary philosopher who has shown greatest attention to the widest possible definition of politics is Michel Foucault, who did so much to bring madness, criminality, and sexuality into intellectual consideration, and who late in his career turned to questions of self-government and politics, using power as his central philosophical concept.²³

To those familiar with his work, Foucault may seem an inappropriate choice as a theorist of microlevel politics because he is so well known for his studies of the disciplinary society, from whose surveillance none escape.²⁴ Foucault himself argued against any inconsistency on the grounds that a proper understanding of his ideas showed discipline and freedom to be two sides of the same coin. To make this argument absolutely clear, we must stop treating Foucault solely as a philosopher and recognize his role as a social scientist, which is the realm in which his political analysis proceeds and from which his conclusions follow. Foucault's contribution at this microlevel of politics is to bring a variety of well-known sociological and empirical facts into philosophical focus. He is less successful, however, in tracing the empirical, sociopolitical microlevel processes in which people engage and that determine winners and losers. I therefore bring together several classic works in sociology, and political economy to fill out Foucault's political model. Although students of the state have tried to avoid this kind of micropolitics, viewing it as too confusing or inchoate to be amenable to scientific study, I argue this is no longer so and that philosophy and the social sciences converge to provide a viable empirical model.

Such inquiry will be controversial—an egregious example of mixing French philosophical apples with American sociopolitical oranges. Yet the threat of this criticism should not deter the exercise. The Foucault interview discussed here gains in power as social theory by being given empirical explication; and the American scholars are enriched by being placed in Foucault's ethical framework. In themselves, the American works have been deeply original, but their implications have been difficult to assess. Bringing the two traditions together is therefore an effective way to assess the objective-subjective dilemma of political theory, that is, the search for a theory that is intellectually coherent but not coercive.²⁵

Foucault provides an axiomatic base; Harold Garfinkel adds experiential depth; Erving Goffman reveals the constant challenge of social interaction; and Thomas Schelling reminds us that winning and losing create the realities within which individuals live.

Foucault's Philosophical Argument

Due to the breadth and diversity of Foucault's published books, lectures, and interviews, no unified interpretation of his work has yet emerged. Georges Canguilhem is considered authoritative on the early period; Pierre Hadot essential to the understanding of the final period; and Gilles Deleuze essential if not definitive on the middle period.²⁶

The reception of Foucault among American political theorists has focused on his theory of the disciplinary society, but here, too, scholars have taken several directions in response to the concept.²⁷ At one time classified with French structuralists, Foucault promptly rejected any such affiliation²⁸ and seemed to take pleasure in violating the expectations his audiences sought to impose.²⁹ The social science aspect I emphasize here is supported by two considerations: first, Foucault's publication in French methodology journals; second, the way in which Foucault's politics theme ties together his work from Madness and Civilization through The History of Sexuality.³⁰ The first step in my analysis is to show how strongly Foucault himself emphasized micropolitics.

In a 1984 interview, conducted five months before his death,³¹

Foucault was questioned by interlocutors who persistently presented him with conventional abstract concepts—such as the relationship between truth and power, liberalism and domination, power and evil, subject and truth—while he, as persistently, tried to direct the discussion instead toward concrete behavior. Foucault presented a perspective that was to some degree foreign to philosophical discourse by virtue of its being more consistently analytical at the microlevel than his listeners either expected or were willing to appreciate. Central to an understanding of this novel perspective is Foucault's discussion of “relationships of power.”

Several points in this passage need to be emphasized. First, Foucault has completely abolished the conventional limits on the types of power typically studied by theorists and philosophers, that is, power in the state or its rulers, or systemwide economic power. Foucault instead emphasizes repeatedly that every human interaction involves relations of power, giving as an example the very interview in which he is participating, where he has precedence as the older, more mature scholar, yet also faces the possibility of losing that position if confronted with a critical challenge from an upcoming, younger man.³³

This ubiquity of the relations of power remains a persistent theme to the end of Foucault's discussion in this interview.³⁴

The second point concerns his distinction between “relations of power” and “states of domination,” which Foucault says are frequently mixed up and used interchangeably only because we fail to define our terms.³⁵

States of domination are a subset of relations of power, according to the model Foucault defines here. He does not use the term power as a way of suggesting “political structure, a government, a dominant social class, the master facing the slave.” Instead it is a shortcut for “relations of power,” meaning “that in human relations, whatever they are—whether it be a question of communicating verbally … or a question of a love relationship, an institutional or economic relationship—power is always present … the relationships in which one wishes to direct the behavior of another.”³⁶ The essence of such relationships is that they exist “at different levels, under different forms,” and they are “changeable relations, i.e., they can modify themselves, they are not given once and for all.”³⁷ The criterion that distinguishes relations of power from states of domination is the availability of complete change in the relationship—the ability of participants to invert the relation so that roles are effectively and substantively reversed.³⁸

The third point raised by Foucault in the passage just quoted pertains to the importance of the “whole network of relations”—the “complex field”—in which human beings operate. At this point in the interview one interlocutor asked him whether his approach might be “the center of a new philosophical thought, of another kind of politics than the one we are seeing today.” While Foucault shied away from this suggestion, he did admit that his interest in Greek and Roman experience reflected a shift from the traditional approach to political philosophy as characterized by Hobbes and Rousseau.³⁹

Much of the interview centers on these ancient ethical relations under the second major aspect of Foucault's philosophical model, the well-known “care of the self” that occupied so much of his concern in the late Collège de France lectures. While the individualism implied by the phrase “care of self” might be culturally familiar to an American—if not in quite the sense Foucault used it—his interviewers found it difficult to understand or accept. After being told by Foucault that “care of the self” is “ethical in itself” and quite independent of any associated care for others, one interviewer pushed him to affirm that “care for self always aims at the good for others.” Foucault fought off this interpretation, insisting that “the relationship to self takes ontological precedence” in his model.⁴⁰

Considering that the interview was recorded several decades ago, one cannot help but see in this struggle between a philosopher and his audience the difficulties that arise when paradigm shifts are under way—in Foucault's case the shift from traditional, highly abstract, philosophy to what today might be called “agent-centered modeling.”⁴¹

What Foucault means by care of self is most fully explicated in the second and third volumes of his History of Sexuality.⁴²

However, the 1984 interview highlighted the structure and context of the idea of relations of power perhaps better than the fuller work. In the interview, he restates the basic theme that for the ancient Greeks, care for self meant knowing and improving one's self and the effort “to surpass one's self, to master the appetites that risk engulfing you.” Knowledge is essential to this project, both in the Socratic-Platonic sense and in “the knowledge of a certain number of rules of conduct or of principles which are at the same time truths and regulations. To care for self is to fit one's self out with these truths. That is where ethics is linked to the game of truth.”⁴³

The political conclusion to Foucault's argument is that “relations of power,” defined as the “means by which individuals try to conduct, to determine the behavior of others,” cannot be eliminated from society. “The problem is not of trying to dissolve them in the utopia of a perfectly transparent communication, but to give one's self the rules of law, the techniques of management, and also the ethics, the ethos, the practice of self, which would allow these games of power to be played with a minimum of domination.”⁴⁴

When the interviewers brought up Sartre's claim that “power is evil,” Foucault could not be turned: “Power is not an evil. Power is strategic games.” The essence of an open strategic game is “where things could be reversed”; and the relations of power are “strategic games between liberties,” resulting from “the fact that some people try to determine the conduct of others.”⁴⁵ Power comes at several levels, Foucault adds, but while government is a major source of domination, relations of power between individuals exist everywhere, even to the apparent privacy of the home and in sexual relations.

Care of self enters this dynamic sociopolitical milieu because only through care for the self is the individual capable of entering freely into the strategic games of power and truth. The point of resistance to political power, a state of domination, involves the relationship of self to self. But it also embodies the practices “by which one can constitute, define, organize, instrumentalize the strategies which individuals in their liberty can have in regard to each other.” It is free individuals “who try to control, to determine, to delimit the liberty of others and, in order to do that, they dispose of certain instruments to govern others.”⁴⁶

Finally, to the question of whether philosophy has anything to say about why people try to determine others' conduct Foucault says that the answer depends on the society in question. In some, things are so well settled in advance “that there is nothing left to do.” But in less restrictive societies, such as our own, there are many opportunities for games in sexual and other relations and “thus the temptation to determine the conduct of others is that much greater.” The more people are free in relation to one another, the greater the temptation on both sides to exert power. “The more open the game, the more attractive and fascinating it is.”⁴⁷

The idea that evil is merely an artifact of the libido ludendum is surely a startling conclusion to the interview, and an indicator of how far one must go in the attempt to catch up with Foucault's forays beyond the usual philosophical boundaries. His demonstration that politics is inherent in individual behavior, and not restricted to state institutions, lays down the primary axiom relevant to responding to Pitkin's problem of autocratic theory.

Empirical Sociology and Political Action

Foucault complained in this interview that as a young scholar he had objected to existentialism and phenomenology because of the way they begin with a predefined theory of the subject and construct theories of truth not analytically, but along preconceived lines. As a university teacher, he sought to escape this philosophical trap through specialization in experimental psychology.⁴⁸

He was also familiar with American microsociology,⁴⁹ and his repeated use of game terminology brings in economics as well. While Foucault's concern with power and strategy might suggest a turn in this essay toward political economy, I believe a full appreciation of his argument benefits first from a consideration of the movement within American sociology variously called social psychology, microsociology, ethnomethodology, or the psychology of interpersonal relations.⁵⁰

Perhaps the most radical analysis of the microsocial milieu was undertaken by Harold Garfinkel, most notably in his Studies in Ethnomethodology (1967), with which my analysis begins.⁵¹

Garfinkel's work closely examines the experience of ordinary individuals, revealing their creativity and courage. Traditional sociology defines social systems as “objective” realities. Garfinkel, though, was awed instead by the contingency, the creativity, the subjectivity, and the inherent mystery of social systems. He saw social behavior not as a thing that was simple and concrete and self-evident, but as a complex, ephemeral, and almost miraculous process by which ordinary people contrived to adjust their behavior so that coherent and reasonable social outcomes would occur. Foucault's appreciation of the delicacy of “relations of power” is made concrete in Garfinkel's method.

Garfinkel's analysis often begins with a small group of individuals, frequently strangers, who are brought together to carry out some institutional task, such as admitting patients to a hospital, or conducting an investigation, or agreeing on a jury judgment. He was struck by the fact that people are able to do these things without specific experience or instructions. “Society” has not completely taught them their roles, in other words, so people figure things out through their interactions with one another, using a fund of deep, unstated knowledge upon which their ambiguous discourse depends. People do not say everything they are thinking, but what they are thinking determines their ability to cope with daily situations. This “practical sociological reasoning” works from the bottom up, in Garfinkel's view, rather than from the top down. Each participant tosses out a possible idea for what might be done in a particular case. Yet because phrases are incomplete and ambiguous, a consensus emerges only gradually. One person finds that his remarks are misinterpreted, but, as it is too late to readjust the conversation, the modification is accepted and becomes the premise for further interaction. The picture Garfinkel presents sheds light on the way in which Foucault interpreted social discipline and replaced it with power strategies. Of course, human experience is subject to discipline, but human life is too complicated for that discipline to be complete. Therefore, as Garfinkel shows, each individual must to some extent consciously fend for himself, creating and recreating, or destroying or modifying, the parameters of the immediate social group by the specific actions that are chosen.

The uneasiness among participants and their associates caused by Garfinkel's ethnomethodology results from the need to belong to the groups of various sizes in which circumstances force us to participate. To be “members in good standing” of such groups is important for practical and symbolic reasons, and membership is based almost completely on the individual's ability to show that he knows the same socially sanctioned facts of life as the other members. One's membership card, according to Garfinkel, is one's ability to demonstrate familiarity with the common understandings of “family life, market organization, distributions of honor, competence, responsibility, goodwill, income, motives among members, frequency, causes of, and remedies for trouble, and the presence of good and evil purposes behind the apparent workings of things.”⁵²

Foucault could not have said it better.

Garfinkel's approach to social knowledge clearly brings out the connection between everyday activities and moral certainties. It is not the activities themselves that are important; rather, the importance lies in the conclusions—moral conclusions—that the activities lead us to make about members of the group and outsiders.

Garfinkel attempts to show individuals how they participate actively and constantly in the creation of their own realities, which they view as objective facts—independent of themselves and their activities. He is particularly interested in the degree to which the complex and unwritten “rules” of this activity can never be completely tabulated and thus require further explication by more rules, which in turn require further rules, and so on, until the cultural milieu of even a modestly sized group or organization becomes infinite. Moreover, participants are concerned that the structures be rational, that their accounting practices will create and sustain “recognizable sense, or fact, or methodic character, or impersonality, or objectivity … independent of the socially organized occasions of their use.”⁵⁴

For participants, “the rational properties of their practical inquiries somehow consist in the concerted work of making evident from fragments, from proverbs, from passing remarks, from rumors, from partial descriptions, from ‘codified’ but essentially vague catalogues of experience,” how things are done. “Somehow is the problematic crux of the matter.”⁵⁵ It would seem to be exactly in the “somehow” that Foucault's relationships of power are located. But Garfinkel is not fully sensitive to power relationships in strategic terms, and this aspect of Foucault's vision requires amplification. Individuals do not create their theories alone, but within social communities that have their own logics to contribute.

Collusion and Power

The second step in my attempt to emphasize the practical political infrastructure of the philosophical viewpoint expressed so vividly in the Foucault interview involves the work of Erving Goffman, a social scientist with whose writings Foucault was familiar.⁵⁶

What Goffman adds to Garfinkel's perspective on the “subject” and “knowledge” is an acute sensitivity to the various forms of collusion that shape our social and intellectual lives. Foucault asserts the presence of power in every human relationship, even the most casual; Goffman documents the exact processes by which this power is manifested, and the subtle way in which human weakness and uncertainty become transmuted into strength and intelligence—or into domination and exploitation—within ongoing social interactions. Goffman will not be the end of the story, because he is one-sided on the implications of power as well as its deeper malpractices. However, his work forms an essential step beyond Garfinkel's in bringing Foucault's always evocative but sometimes gnomic principles into concrete view.

Goffman's empirical studies illustrate Foucault's emphasis on rules as constitutive and deeply sociopersonal. In his earliest major work, The Presentation of Self in Everyday Life (1959), Goffman begins with an analysis of social situations, ones in which individuals present themselves to one another, explicitly and implicitly, and begin in concert to develop those social regularities we call systems or institutions.

Goffman uses the metaphor of theatrical performance to emphasize a perspective that was, at the time, unique: “I shall consider the way in which the individual in ordinary work situations presents himself and his activity to others, the ways in which he guides and controls the impressions they form of him, and the kinds of things he may and may not do while sustaining his performance before them.”⁵⁷

There are inadequacies to the dramaturgical model, Goffman says, in that the “stage presents things that are make-believe; presumably life presents things that are real and sometimes not well-rehearsed.” But there is also an eerie similarity between the stage metaphor and “life.” Goffman's many publications, which achieved considerable popularity with lay audiences, attest to the fascination exerted by his unusual viewpoint on human behavior.

Goffman's approach was essentially naturalistic. Where Garfinkel indulged in reality-bending experiments, Goffman simply looked around. In his analysis, there are two levels of individual expressiveness: “given” information, such as verbal statements; and “given off” information, such as actions or impressions resulting from actions.⁵⁸

On meeting someone, for instance, one processes more than just a person's words; one also uses physical characteristics—such as clothing, manner, and the look on the other person's face—as major elements in defining the situation. As Goffman notes, individuals may have many purposes in social interaction. Yet, whatever the specific purpose, the individual seeks to control the conduct of other people, especially their responses to the self. Thus, in any interaction, much will be concealed; participants may know neither the real beliefs, attitudes, nor emotions of the other performers.⁵⁹

In Goffman's model, all knowledge in these interactions is inferential, and participants cannot know anything for certain. Each participant will, however, project (even if passively) a definition of the situation. Ordinarily these definitions are sufficiently attuned so that no open contradiction occurs, although there will not be any consensus—real meeting of the minds—among the participants. Rather, each individual suppresses his own deepest feelings and projects a definition of the situation that he thinks the other will be able to find acceptable, at least temporarily. There is thus created a veneer, a surface agreement, in which everyone participates, and to which they all feel obligated.⁶⁰

Such behavior, of course, is the soil in which everyday politics grows.

Like Garfinkel, Goffman emphasizes that definitions established in such social processes have a moral character. Individuals with certain characteristics have a moral right to expect that they will be valued and treated in an appropriate way. From this viewpoint, the apparently innocent idea of “defining a situation” turns into actions that make a moral claim on other participants.⁶¹

Goffman's distinctive contribution to microsociology is his emphasis on the social structure that controls the roles played by individual members who find expectations waiting for them—and who run into extreme difficulty if they attempt to make major changes. Even actors who seek to change their way of fulfilling a role will find that they cannot create novelty, but can only choose among several already established ways of playing it.⁶² The socialization process tends to idealize roles, according to Goffman, so that performances are “socialized, molded, and modified” to fit the observers' understandings and expectations. Eventually performers exemplify “the officially accredited values of the society” and each performance becomes a ceremony, “an expressive rejuvenation and reaffirmation” of the community's moral values, and a celebration. People who seek upward mobility within the group do this not simply for prestige, but because they want “a place close to the sacred center” of the common values of the society.⁶³ In other words, they play Foucault's cultural power game, the game of truth.

Later, Goffman shifted his approach slightly, moving from the individual to small groups of people who work together and who are, to some extent, forced to “perform” as a group for others. The central group, known as the “team,” is characterized by a sense of familiarity and solidarity among the members, and by a set of shared secrets about how their “act” works (secrets that would “give the show away” if revealed to the audience).⁶⁴

Social behavior, the “show,” takes place within a particular establishment, and the team of performers cooperates in “impression management,” that is, the presentation to the audience of a desired definition of the situation. All members of the group, team members and audience alike, willingly conspire to maintain the desired definition of the situation, which includes the conceptions of the team, the expectations of the audience, and “assumptions concerning the ethos that is to be maintained by rules of politeness and decorum.”⁶⁵ A major requirement involves keeping the “back region” of the establishment, where the performance is prepared, separate from the “front region,” where the performance is actually presented. Audiences do not expect to see the back region, nor does the team want it to be seen.⁶⁶ While Goffman chose to focus on relatively harmless examples of this social behavior, his analysis extends easily to other situations, such as that of overprivileged members of society justifying their position to an audience of the underprivileged, or the tyrant making an impassioned speech about democratic freedoms.

Readers who follow Goffman into his later work expecting to find a larger theoretic perspective are frequently disappointed. The corrosive analytic vision Goffman directed upon everyday society continued to provide striking vignettes of amusing or appalling social behavior, but the hoped-for glimpses of a social philosophy never emerged.⁶⁷

This is one of the reasons I believe it is valuable to bring Goffman and Foucault into the same field of vision; it brings empirical evidence to one and philosophical resonance to the other. Foucault's critics have been frustrated in their efforts to understand how he could simultaneously see society as disciplined and yet free. As abstract terms, freedom and discipline are contradictory; but as concrete processes, they may coexist, as Goffman repeatedly shows. As Wittgenstein argued, actual practice is the only recourse when confronting philosophical dilemmas, and the rich prepolitical infrastructure analyzed by Garfinkel and Goffman is an essential element in unraveling Pitkin's dilemma. This theme is taken further by moving to the work of political economists. They share much of Foucault's emphasis on power and strategy, but from a different tradition and discipline that gives concrete expression to his philosophical viewpoint.

Economic Dimensions of Politics and Theory

Of the classic generation of political economists who first introduced rigorous analytical arguments into political science,⁶⁸

one member in particular worked consistently at the microlevel of political issues and provides empirical referent to Foucault's strong, yet difficult, model of power relations. This was Thomas Schelling, who worked in the tradition of international and foreign policy, but nonetheless directed his analytic attention to individual political behavior, usually far away from state institutions.

The similarities between Foucault's responses in the 1984 interview and Schelling's work in strategy are striking. Both see power in all human relations, though Schelling calls it the “mixture of conflict and cooperation.” Both understand games as being played within social and ideological contexts that shape and constrain them and see that games may mandate winners and losers independently of the merits of the play. Yet the two write in worlds as far apart as economics and philosophy. Consideration of Schelling's work from a Foucauldian point of view shows how serious American social theory can appear when placed next to a major French thinker and demonstrates how empirically acute at least one French philosopher may have been when forced to that position by the intellectual puzzles he chose to pursue.

Schelling's classic work, The Strategy of Conflict (1963), was slightly mistitled; the “strategy of conflict” erroneously implies that his primary emphasis was conflict rather than the many other forms of human interdependence he discusses. Better characterizations of his work turn up in the course of Schelling's discussion—a theory of precarious partnership; a theory of incomplete aggression; or, most accurately, the theory of interdependent decision.⁶⁹

Games of strategy are “those in which the best course of action for each player depends on what the other players do,” and the term strategy “is intended to focus on the interdependence of the adversaries' decisions and on their expectations about each other's behavior.”⁷⁰ Where traditionally social theorists have been expected to make a choice in regard to the nature of society, to define it as either competitive or cooperative, Schelling is at pains to emphasize that neither horn of this supposed dilemma is correct. There is no friendship so close that it cannot contain some conflict, some differences of values, he argues; nor is there any relationship so hostile that it will not contain some mutuality of interests. The “richness of the subject” arises from the “common as well as conflicting interests among the participants.”⁷¹ This firm neutrality of approach is, of course, exactly like Foucault's.

Schelling's use of game theory in The Strategy of Conflict sharply distinguishes him from the majority of political economists, then and now, because he uses the theoretical concepts it provides in a manner so flexible, fluid, and creative that his results often overthrow the theory's own postulates. On the vexed issue of rationality, for instance, Schelling points out what no one else before him had explicitly recognized—that where rationality is pitted against irrationality, irrationality may work better (you cannot successfully bargain with a four-year-old, a madman, or a fool).⁷²

Schelling's insight here rests on his premise that rationality is not a one-dimensional attribute that people either have or do not, but is a multidimensional collection of attributes involving value systems, information systems, and communications systems. His approach to value systems is wide, unlike the narrow self-interest axiom of many political economists, and encompasses the whole range of motives unearthed by psychologists, sociologists, and political scientists.

Bargaining, as Schelling defines it, falls into two basic categories, explicit and tacit, which are separable analytically but inextricably mixed in practice. Both types of bargaining involve the making of threats, promises, and commitments; the creation of credibility for one's actions or threats; the minimization of one's own flexibility (notice how this counters the usual idea that flexibility is desirable); and the search for focal points for settlement. The objective is to avoid actually using one's strengths, exploiting them, instead, as potential forces. Since the classic writings of Carl von Clausewitz, it has been obvious that war was a waste of resources—a successful threat is more efficient because it accomplishes its aim without using up any goods.⁷⁴

Foucault's facing various interviewers and recognizing the dynamics of his own situation is a small example of Schelling's point.

In an argument particularly relevant to Foucault's approach to social relations, Schelling contends that, in observing groups and societies, it is wrong to believe that participants share “trust, good faith, and mutual respect” simply because peace is maintained: apparent agreement may overlay deep hostility, and implicit bargaining is constant among participants.⁷⁵

Achieving success, according to Schelling's analysis, is both conceptually simple and straightforward in practice: it results from individuals so limiting their options that they cannot do what they do not want to do. Obviously this involves some delicacy. An individual must know his opponent well enough to assess what he can get away with; know the situation well enough to determine what will be considered appropriate behavior; and know the extent of his nerves and courage on this particular issue, with this particular person. This last aspect highlights a major facet of political games—the individual's level of determination. Schelling describes “pure” bargaining as the case where “each party is guided mainly by his expectations of what the other will accept,” and where the contest is purely one of will—one participant may concede simply because he is convinced that the other will not do so.⁷⁶ The psychology of this aspect of the political contest has rarely been studied in any systematic way. Foucault does not begin to handle it with Schelling's calm familiarity, which is what makes Schelling so useful in this regard.

Common opinion has it that “bargaining strength” means that the success of a participant is directly related to background resources, and that the weaker competitor may be forced to give way, trapped by the logic of the game's structure. In a pure contest of wills, this need not be so.

If this is true, as both Schelling and Foucault seem to assume, it radically shifts the field within which political interactions occur so that the last have at least an opportunity to become first, and the first are well-advised not to overextend their ambitions.

What gives Schelling's work its particular importance with respect to Foucault's model of power relations is his recognition that the political bargaining process does not merely create individual deals between private persons, but is actually the process underlying the construction of entire political societies. Far from the collegial process some social theorists seem to imply, the construction of social reality is usually a game of intense and perhaps hostile bargaining. The results determine not just who wins in the short term, but who wins in the long-term game of social valuation. The outcome of this civic battle can define, for a long time into the future, one's place in society: who is honored or scorned, who has rights or only duties, who is entitled to justice, and who must eat bitterness. The deeper Schelling analyzes, the more he leaves behind the abstractions of game theory and begins to incorporate elements already presented above—cognitive difference, personal creativity, social inequalities, and the strategic factors so decisive in bargaining.⁷⁸

According to Schelling, by pursuing “too abstract a level of analysis,” “by abstracting from communications and enforcement systems, and by treating perfect symmetry between players as the general case,” game theory has “overshot the level at which the most fruitful work could be done.”⁷⁹

Symmetry between two abstractly rational players is unrealistic because “in too many exciting cases one plays an opponent who is a wholly different kind of person” from oneself.⁸⁰ Even in the classic book that established game theory in American social science, Games and Decisions: Introduction and Critical Survey, Duncan Luce and Howard Raiffa argued that prediction of behavior is foolish “without first having a complete psychological and economic analysis of the players.”⁸¹ The conduct of a game can be altered, for instance, by one player's choice to change his own commitment structure, where this change in commitment alters the opponent's expectations about possible payoffs and therefore constrains the opponent's options.⁸² But psychology tells us little and game theory tells us nothing about what types of players can or will make changes in their own psychological structure. One thus arrives at questions that are the prerogative of political theory, in respect to the nature of the individual and the individual's political goals and capacities.

Democratic Gaming

An academic division of labor between philosophy, sociology, economics, and political science may be the key to an increased depth of scholarship and knowledge, but it may also be procrustean. After learning from philosophers, sociologists, and economists that individual human beings participate “up close and personal” in the construction of their life realities, one may naturally inquire how this model plays out in politics. Frequently the question is left unasked because the issue is not within the disciplinary purview of those who made the initial discoveries. I have therefore tried to pull together some theoretical aspects of the political interactions in which women and men find themselves, where the “political” resides not in government institutions but in the asymmetries—real or potential—of the relations and in the allocations of social values these asymmetries entail. Now it is time to ask whether these ideas help us with Pitkin's problem. Can we use these tools to develop a theory that proceeds from within an individual's experience and control (rather than one imposed from the outside by an “objective” observer) and yet has the capacity to provide meaning to human affairs and a “sense of coherence and mastery” sufficient to guide political action?⁸³

The strategic politics model that is defined at the intersection of Foucault's approach and the detailed empirical context provided by Garfinkel, Goffman, and Schelling may provide a broad epistemological metaphor appropriate to a new form of bottom-up analysis of political issues and systems. It is a game model, but it would be incorrect to identify it with game theory as typically practiced. It might be called a Schelling game, to emphasize the flexibility combined with rigor that is typical of his work.⁸⁴

A microlevel theory of this sort would not simplify events to manipulate them according to rigid mathematical principles. Instead it would amplify our sense of the game as broad and richly detailed enough to provide space for the many kinds of everyday politics we experience and wish to engage, either as researchers or participants. Wittgenstein, who inspired Pitkin's original work, offers examples of this approach in terms that recall Garfinkel's appreciation of people's irrepressible spontaneity.

Wittgenstein took this approach even further in another passage:

A comparable flexibility in the nature of the game is suggested by a political scientist, Kenneth Shepsle.⁸⁷

He argues that where political scientists often implicitly equate “rules” with “official regulations or laws,” it would be preferable to be imaginatively empirical and to consider how “rules” work in such children's games as kick the can or hide and seek (or baseball played not on a regulation lot but in a backyard with trees and brooks). Such games come with “standard” rules received from the past, but “kids are known to alter the rules.” Sometimes these changes respond to an irregular environment (hitting the baseball into the brook is considered to be a double), or the changes may reflect the type of players (young players get special help), and so on. Children may also change the rules mid-game, as the result of conflict over some unforeseen outcome. If the argument grows sufficiently hot, some players may pack up and leave. Such situations are comparable to institution creation, Shepsle argues, because these games are not invented behind any “veil of ignorance,” but by players who know themselves and each other. They will therefore engage in “distributional conflicts” over different definitions of the rules and over the impact of those rules on profits for specific individuals.

These broader approaches to the concept of the game, and these more flexible theories of the game, do not imply that we will stop doing conventional, mathematical game theory and switch entirely into a new paradigmatic framework where games are created “in the field” by human action and ingenuity. Formal game theory is elegant and useful. But the game idea, in all the complexities suggested by Wittgenstein, Shepsle, and the theorists discussed earlier, need not be narrowly held by a single group of scholars. It can, with the perspectives of microsociology, be seen as a philosophical exercise, much of the sort that Wittgenstein proposed.

What would such an expanded game model contribute toward dealing with Pitkin's dilemma?

• Intellectual coherence and order. The game concept provides the everyday person with a framework that is entirely familiar yet newly understood. There are players, strategies, tactics, payoffs, alliances, good play, cheating, and so on. However, the terms apply off the field, court, or game board; and human interactions appear in a fresh light, with new attention to the “other” and a new definition of the self.
• Equalization of participants. There are no authority figures in the philosophically defined game, only winners and losers. The winners have no permanent status and no arbitrary rights; they may have won by virtue or by luck, and will have to face further encounters to prove their position. The losers are also temporary; they can learn from the successful strategies of others, or they can define the game for themselves and, by inventing new tactics, perhaps transform it.
• Contingency. One game does not make a season: People win some, lose some, and live to fight another day. The game metaphor reminds them—even those who have lost regularly because of the foul play of their opponents—not to give up, not to disappoint their fans.
• Strategic awareness. Although an individual may be poor in physical or social resources because of the accidents of past games, he can always turn to his own ability to outwit the opposition or to seize a hitherto unnoticed opportunity.
• Game ethics. Fair play is a virtue recognizable by all players in everyday games. It may sometimes be honored only in the breach, but its value often grows in importance as players gain skill and judgment.⁸⁸

  Even formal theorists have devoted attention to this factor, for instance, in searching for ways of escaping the “Prisoner's Dilemma.” See Axelrod 1984; Axelrod 1997.

  But game theory as a metaphor retains the sharp edge of its rational choice origins; if the opposition cheats, even those who prefer fair play may want to consider going for the jugular.⁸⁹

  It may be useful to insert here a reminder to the reader that these games proceed within established governments. Therefore, it is not a question of disobeying the nation's laws, but rather of strategies that are perhaps unexpected but not illegal.

How is such a microlevel political strategic approach different from standard game theory? It is diametrically opposite on the crucial dimension of analytic purpose. Technical game theory came into its prime in John von Neumann and Oskar Morgenstern's theory of games, where the whole mathematical point of the exercise was to “solve” the game, to devise the right strategy for winning under all conditions.⁹⁰

Decision theory, and at least some of political economics, has a similar concern—rigorous calculation of probabilities in order to achieve desired outcomes.⁹¹

Foucault's approach, combined with the empirical infrastructure of Garfinkel, Goffman, and Schelling, takes as a major premise the idea that there is no solution to the microlevel political gaming world they define. The interaction is ongoing and the outcomes unpredictable; players learn through their engagement in the exercise, and the theorist's role is to watch carefully and intelligently, to understand the creative process as it unfolds. The importance of Foucault stems perhaps less from his emphasis on the ubiquity of politics than on his insistence that women and men become aware of the political nature of the worlds within which they live. The very virtues of modern civilization—its openness and complexity—create new challenges and, according to Foucault, individuals who fail to meet them may fail politically, existentially, or ethically. But the learning process is their own, and is left to their discretion.

Garfinkel's work sensitizes the innocent to the way in which the political field is constituted at the bottom by apparently minor and unimportant behavior—the kind that, despite its seemingly trivial nature, serves as a basis for status within one's personal groups and the human group in general. Goffman documents the small strategies through which groups emerge, cohere, control, and confront one another; he also shows how status hierarchies emerge, maintain themselves, and sometimes collapse as a result of apparently small pressures. Schelling is the capstone of the model, showing that within the appalling richness of human interaction in time and circumstance, it is possible to find analytic focus through close attention to simple strategic concepts, such as players, resources, payoffs, and rules, and the recognition of interactive patterns and strategies that control the outcome of various political systems.

The strategic model defined here is spartan. It makes only one assumption: that people define themselves and their social institutions through their strategic interactive behavior. The model, as developed through the theorists mentioned in this essay, gives no specific advice, sets no preferred goals or strategies, and is silent on outcomes and preferred ethical systems. It provides a picture of how the world appears to work, but the picture is not prescriptive. The model asks only what people do, and have done, throughout various periods of history. Political theorists are then entitled to inquire as to the results of this process, and may wish to point out that certain behaviors entail certain results.

Political science would profit from the same approach.⁹²

It would be a nice reintegration of a discipline too long split by methodological disputes, if political theory and empirically oriented political science worked together on understanding the wider shores of political experience. A great deal of intelligence, creativity, and courage is involved in political activity, and, as I have tried to suggest, it is activity in which all people participate daily. That it be recognized as political, and that theorists may help in showing how it may be done well, expands the realm of democratic theory and raises the possibility that resolution of Pitkin's dilemma may not be completely beyond our reach.