Archives: Book Chapters

At no time in its history has the Judeo-Christian tradition been able to confine its interests within narrow ethnic, regional, or national boundaries. The Hebrew scriptures tell of a people forced to migrate beyond their own borders in search of food, displacing local gods with a God of the heavens, and recurrently finding themselves caught in the intrigues of warring empires. Jesus constantly ran into the limiting presuppositions of such boundaries in his day, and he repeatedly cut through them to enlarge his followers’ vision. When the lawyer asked him for a definition of neighbor, Jesus pointedly told a story showing that compassion must extend beyond ethnic borders. He was crucified by Roman soldiers. And when he commissioned his disciples, he admonished them to go into all the nations.

Over the centuries, the international dimension has been an integral feature of Western religion. By the end of the fourth century, Christianity had spread to nearly every corner of Europe as a result of the Romans’ conquests. In 1095 the great crusades began, pitting Christianity against the Muslim empire. By the end of the fifteenth century, Spanish and Portuguese explorers had taken their faith to the New World. In 1517 a wave of reforming zeal broke out in central Europe which was to shape permanently the location of national boundaries and the strength of territorial sovereigns. And by the end of the nineteenth century, the Western church was deeply implicated in the affairs of every continent as a result of missionary activity, trade, coloni­zation, and war.

In many parts of the world — Africa, the Philippines, Iran, Lebanon, Nicaragua, Brazil, the United States — the relations between religious institutions and the reigning powers are undergoing dramatic changes. Opposition movements have sprung up in sectors of the religious community that only a few decades ago seemed to have bargained away their political soul. Alliances have been forged between representatives of religious traditionalism and segments of the technical intelligentsia in a way that would have seemed peculiar up until recently. Fighting, physical and verbal, has broken out between religious factions that once stood united against other faiths and creeds.

These developments have challenged policy makers and religious leaders at a very practical level. But they also pose challenges to the ways in which we think about the public place of religion in the contemporary world. They have forced us to question whether our conventional theories are adequate to the task or whether we need to push toward new understandings. They have at the very least raised the stakes for finding effective approaches from which to consider the relations between religion and politics.

In reflecting on the dynamics of religious change, some scholars have shown a tendency to abandon formal theoretical frameworks entirely. They advise that one should go to the field, write up what out sees, and forget trying to mold these observations to fit some preconceived theory. Social science should, in their view, be like good journalism: it should richly describe but keep interpretation and explanation to a minimum.

There is of course much merit in shedding theoretical blinders in order to see with fresh eyes what may really be going on. But, we must also ask, how completely can the assumptions and presuppositions that go with received theoretical frameworks be abandoned? Even when we try to be purely descriptive, we choose selectively to write about some things and to ignore others. As proof, compare books on religion and politics written by political scientists and anthropologists. Even when the subject matter is ostensibly the same, we see quite different questions animating these studies.

It is for this reason that some effort to review the main theoretical frameworks implicit in studies of contemporary religious change seems in order. I have chosen four such frameworks that constitute important perspectives in their own right: modernization theory, world-system theory, what I will call "structural contingency" theory, and some recent work on systems theory and the lifeworld. I shall indicate briefly what each of these four traditions suggests about changes in the relations between religion and politics and then consider some of the more general presuppositions that these theoretical perspectives illustrate. Finally, I offer some observations of a more synthetic nature about factors relevant to an understanding of religious restructuring.

Modernization

Theories of modernization, despite the rather serious attacks to which they have been subjected in recent years, have been so prominent in the social sciences, and have played such an important role in our thinking about social change, that any effort to consider the changing relations between states and religious institutions must begin here. The variation in substantive arguments that have huddled together under the general rubric of modernization is, of course, vast. In considering modernization less as an explicit theory than as a broader set of implicit assumptions, it is nevertheless possible to speak of it more or less as a coherent framework.

The central presupposition of modernization theory is that something can be identified that distinguishes "modern" societies from those that are less modern (i.e., "traditional" societies). Among the characteristics that might signal the presence of greater modernity are higher levels of industrialization, a greater use of advanced technology, overall indicators of higher economic development, more literacy, a more comprehensive educational system, greater urban density, and more extensive administrative capacities on the part of the state. It is argued that these characteristics may be associated with one another at a sufficiently high level that overall portrayals of various societies as more or less modern make a good deal of empirical sense.

The modernization perspective also suggests that a kind of wholesale movement in the world has been going on for some time — movement along the continuum from less modern to more modern. Some societies, like Great Britain and the United States, started along this path relatively early and thus have progressed considerably, while others (e.g., South Korea) are relative latecomers with much remaining to be accomplished, and still others (e.g., Ethiopia) have hardly started down the track at all.¹

The progression from traditional to modern is thought to have serious repercussions for religion, since religion has everywhere been an integral feature of traditional societies. Moving toward more modern social arrangements is presumed to entail a displacement or devaluation of traditional religious institutions, or at least some concessions on the part of such institutions to the secular environment. "Secularization" refers either to the fact that religion, in this process, comes to have a less prominent or influential position in modern societies or to the fact that it retains its influence only by conforming increasingly to such norms as rationality and relativism or by making compromises with science, economic concerns, and the state.²

Many specific doctrinal and organizational changes have been associated with the secularization process. As societies modernize, doctrine is expected to focus more on happiness in this life than on other-worldly compensations. It is also expected to become less dogmatic, taking on live-and-let-live orientation conditioned by the fact that people become more aware of the realities of competing worldviews. The gods themselves may undergo a transformation, as people cease to attribute miraculous deeds to them or conceive of them as unquestioned authorities, viewing them instead as symbolizing love and redemption or sanctioning ethical systems. More emphasis is in fact expected to be placed on symbols because people come to recognize the differences between religious symbols and the underlying truths these symbols are supposed to convey. Religious functionaries lose some of their unique claim to power as a result of competition from secular professionals and in conjunction with a more general rise in education and values stressing individual discretion. Religious organizations, in turn, come to focus on a narrower range of social functions and are likely to adopt marketing strategies and formal bureaucratic procedures in order to compete effectively.³

For the most part, modernization theory emphasizes the long-term direction of such changes. But it also provides for assertions to be made about the rate, timing, and severity of shorter term adaptations. An especially important "take-off" period, lasting only several decades, may occur within the much longer transition from tradition to modernity. In this period, the economy may grow by quantum leaps, a new mode of production may come into being, or the political regime may change in a way that promotes subsequent economic change. The consequences for religious institutions may be exceptionally severe. Old patterns may be overthrown within a relatively short period, making new patterns all the more visible and controversial.

There may also be reversals in the modernization process, or at least segments of the society who mobilize temporary resistance. Thus, some of the turmoil one sees in the religious sphere may be understandable as protests against modernization. Should industrialization spurt ahead, for instance, religious traditions rooted in agrarian lifestyles are likely to mobilize sentiment against these economic changes. After a rapid advance of this kind, any temporary economic setback may play directly into the hands of such opposition movements, giving them greater credibility and funneling dissidents into their ranks.⁴

Modernization theory has been applied most widely, of course, to societies in the Third World that have undergone rapid change in recent decades and experienced religious turbulence in conjunction with this change. It forces analysts to pay attention to the ways in which political development undermines religious tradition and therefore provides an implicit framework in which even many studies that have not bought into all of its assumptions have been cast.⁵

The Modern World-System

Critics of modernization theory have focused particularly on its tendency to treat societies as isolated units. Rather than viewing each society as a separate entity moving along a track from less modern to more modern, the critics of modernization theory argue that societies interact with one another as parts of a larger system. The reason one Society is a laggard may have something to do with the fact that other societies are not.

World-system theory has emerged over the past fifteen years as leading contender with modernization theory.⁶ As the name suggests, world-system theory emphasizes the larger set of social, economic, and political relations that link societies together. According to world-system theorists, these relations began to emerge in the sixteenth century, chiefly as a result of international trade and diplomacy among the European states. Gradually, this system became the driving force of modern capitalism, and its influence spread over most of the globe by the end of the nineteenth century. World-system theorists maintain that in today’s world the contours of societal change on virtually every continent must be understood in terms of the dynamics of this larger system.

The intellectual origins of world-system theory can be traced most directly to various offshoots of Marxism, including studies of political economy, theories of imperialism, and ideas about dependent development. Because of the epistemological assumptions in these traditions, world-system theory has paid little attention specifically to the role of religious beliefs or religious institutions. It has thus been necessary for other theorists to suggest ways in which the world-system perspective might be useful for understanding changes in these beliefs and institutions.

Applications of world-system theory to questions of religious change have focused to a great extent on the ways in which short-term changes in the world economy may affect the stability of religious institutions. Some of the arguments that have been advanced do not differ markedly from those advanced within the modernization framework. Generally, though, world-system theory has placed greater emphasis than modernization theory on the abrupt, disruptive, and conflictual nature of changes in the world economy. Because societies’ fortunes are said to be so closely connected with the dynamics of the broader system, many things can happen over which societies themselves have little control. For instance, a periodic downturn in the business cycle can have severe repercussions for a tiny country whose economy depends heavily on exports. Or an outbreak of war may cut off trading channels, resulting in equally serious disruptions in exporting societies. Religious institutions may be caught in the middle of such changes: pro-Western religious orientations may suddenly become unpopular because of changes in trading alliances, peasants may turn to millenarian or folk religions to revitalize economically threatened communities, communist or nationalist movements among oppressed urban workers may strike out at traditional religious organizations, and so on.⁷

World-system theory’s roots in Marxism have made it particularly sensitive to the cyclical and conflictual nature of modernization. Rather than conceiving of modernization as moving happily toward greater prosperity and enlightenment for all, world-system theorists depict it as moving in fits and spurts, as a kind of Hobbesian drama. As capitalism spreads through the world economy, it produces war, oppression, and hardship for many, even though it may generate prosperity for a few.⁸ Capitalism also becomes subject to its own internal contradictions.

Cycles of rapid expansion in economic production are likely to be followed by downturns conditioned by slackening demand; the costs of acquiring and protecting new markets through diplomatic deals and military intervention eventually outweigh the gains to be had from these markets; dominant countries gradually lose their hegemonic power; and the whole system becomes subject to the strains of realignment as new countries or new modes of production rise to prominence.

Within a single society, the social dislocations attendant on these broader strains may look very similar to the observer who focuses only on that society and the observer who emphasizes world-system dynamics. What the world-system theorist insists on bringing into the picture, though, is some understanding of the external forces that bring about these dislocations. Two lines of inquiry are likely to follow. One stresses the ways in which regimes and elites in a particular society respond to these external forces. Rather than seeing religious conflict strictly as a domestic issue, the observer looks at it in terms of the military obligations, foreign debt, trade advantages or disadvantages, coups d’etat, and so on that may be inspired by broader diplomatic, military, and economic considerations. The other line of inquiry stresses ways in which such conflicts and dislocations in particular societies may exemplify patterns of a more general or systemic nature. Rather than viewing upheavals in Central America in isolation from those in the Middle East, for example, observers are inclined to ask how both may reflect debt patterns in the world economy or the competition among superpowers for energy supplies or the arms race.

Perhaps the key attraction of world-system theory, overall, is that it sensitizes social scientists to the growing global interdependence that now exists among nation-states. For studies of religious and political change, this interdependence has clearly begun to be a critical consideration.

Structural Contingency Models

A third perspective — which, for lack of a common term already in use, I will call "structural contingency" — can be identified in a variety of work that has arisen over the past decade or so in criticism of both the modernization and world-system perspectives. The common denominator in this work is a conviction that modernization and world-system theorists paint with too broad a theoretical brush. Certainly Central America and the Middle East are both subject to the broader forces of modernization or the broader dynamics of world capitalism, say these critics, but look at the differences. We need some way to take account of the fact that the colonial histories of the two regions is quite different. And we should also pay heed, especially in the present case, to differences in the religious histories of the two.⁹

The structural emphasis in these studies might be described as an interest in institutional arrangements. The relations among various agencies of the state, organized economic actors, the ecclesiastical elite, and opposition movements or parties have been of special interest. Rather than simply emphasizing levels of institutional differentiation, as many modernization studies have, proponents of structural contingency have stressed the importance of resource flows, overlapping memberships, mutual and competitive interests, and organizational interactions. And rather than focusing on a society’s general placement in the world system, they have devoted attention to the dynamic relations between this position and the opportunities, resources, and constraints under which influential actors may operate. Thus, for example, it may be especially important in the Central American case to know that dominant land-owning families also hold important positions in the government but not in the church, while in the Middle Eastern case it may be important to know that political officials have relied heavily on money from exports and have therefore needed to depend less on the merchant elites in local markets that dominated power structures in the villages.¹⁰

The contingency emphasis can he seen in the tendency of these analysts to argue that variations in societal development depend on the ways in which institutional arrangements were configured historically Societies that have retained a strong relation between landowners, religious elites, and the central governing bureaucracy, for instance, may respond to new crises or economic developments in vastly different ways than societies that at some point in their history threw the landowners out of power, reformed the church, and gave mass representation to the industrial working classes.¹¹

Studies exemplifying structural contingency ideas have been especially prominent in the literature on religious movements. Starting with observations about the relations between the state and established religious institutions, these studies show how religious movements position themselves in relation to both. When the state holds powerful control over the religious hierarchy and other public institutions, popular religion may become a nursery for political opposition. The appeal of popular religious groups in Brazil, Chile, India, and Poland has been examined in these terms.¹²

Lifeworld Colonization

The tendency that seems evident in studies influenced by world-system theory and in studies emphasizing structural contingencies is a shift away from abstract theoretical models toward somewhat more down-to-earth approaches that give considerable room for empirical induction. These approaches, of course, gain ready support from regional and area specialists and ethnographers whose studies rely heavily on firsthand observations and extensive familiarity with the traditions of particular societies.

The danger in moving entirely to strategies of empirical induction is that integrative, and even prescriptive, perspectives lose out.

The recent work of German sociologist Jurgen Habermas, in which questions about the formal characteristics of social systems in general and the dynamics of the lifeworld are the focus, exhibits a clear preference for deductive theory of a prescriptive sort.¹³ Habermas has drawn eclectically from modernization theory and Marxism to create what he calls a reconstructive model of cultural evolution. In this model, the modern epoch is characterized by an abandonment of the three-tiered or dualistic universe of traditional religion, a reliance on scientific and technical reasoning, and increasing intervention by the state to promote advanced industrial capitalism and to combat its ill effects on social life. We are, however, on the verge of transcending the modern epoch and moving into a post-modern period. Habermas regards this as a vital step that we must take in order better to master the contingencies we face.

According to Habermas, the twin evils that beset us resemble the evils identified by Weber and Marx in the nineteenth century. From Weber, Habermas borrows a concern for the effects of bureaucratization, and from Marx, a focus on the evils of the capitalist market. The former is associated mainly with the modern state and, strategically, gives Habermas a means of criticizing his neighbors to the east, while the latter conjures up the dangers of rampant free enterprise which, to many Europeans, the United States epitomizes. Underlying both tendencies, however, is what Habermas calls "technical reason." This is a reliance on instrumental logic and the means of adapting to the material environment. It contrasts with an emphasis on open and free debate about the goals and values of society itself — what Habermas calls "communicative action." To gain command of our collective destinies, Habermas believes we must cultivate communicative action.

From this perspective, many of the religious movements we see emerging in various parts of the world — especially those in advanced industrial societies — can be understood as protests against the growing bureaucratization and monetarization of the lifeworld. Habermas suggests that we are finally becoming aware of the threats that confront our quality of life, our sense of ourselves, and our natural environment. Consequently, we see an increasing number of movements attempting to combat these threats. As examples, he cites the various mystical and human potential groups that have arisen in opposition to the impersonality of modern life, the efforts mounted by established religious groups to advocate equality and social justice in the name of traditional or divine values, communal experiments with the reshaping and redefinition of work, and special interest groups concerned with gender roles, the family, and environmental pollution.

Habermas takes a critical view of all these movements because he regards their own theoretical vision as being too narrow. The solution, he argues, must come from a better understanding of the communication process itself. Thus, some developments in theological hermeneutics attract him as examples of such progress. But in the meantime, he predicts a heightening of social unrest in which various short-lived religious movements play an important role.

On the Uses of Theory

I have, in sketching the contours of these four theoretical perspectives, deliberately refrained from offering evaluative remarks. I do not thereby mean to suggest, however, that the choice is merely a matter of indifference or personal taste, or that a synthesis of all four perspectives may be best. On the other hand, these clearly are not alternative theories that may be decided among on the grounds of which one "best fits the data." Indeed, the scientific model of theory testing is likely to lead us very much astray in the present context. Specific hypotheses could perhaps be derived from the literature reflecting these various perspectives, and then evidence could be adduced to see whether one hypothesis made better sense than another. But that kind of exercise in positivist social science lies at an entirely different level of discussion than the one we are engaged in here.

For present purposes, the idea of perspective carries more appropriate connotations than the notion of hypotheses. I have chosen these four theoretical perspectives because they help make explicit some of the assumptions that are likely to influence the ways we think about the relations between religious institutions and state structures. We need, therefore, to pay closer attention to these assumptions, now that the general frameworks have been described.

One of the choices these frameworks set before us concerns the fundamental stance we choose to take toward the general thrust of economic development in the world today. At the extremes the choices are clear. One extreme emphasizes the light; the other extreme, darkness. Implicit in many variants of modernization theory is the assumption that economic development is both inevitable and desirable. While the transition to modernity may be painful, perhaps especially so for practitioners of traditional faiths, the overall gains must be positive. Physical health, prosperity, greater individual freedom, and cultural sophistication are the measures of these gains. At the other extreme, many variants of world-system theory and some variants of the other perspectives regard economic development as inherently productive of conflict, oppression, and exploitation.

What we look for in studies of religion and the state depends greatly on the stance we take in relation to these two extreme interpretations of the development process. Studies conducted during the 1950s and early 1960s often took an optimistic view of economic development and, in keeping with this outlook, showed how traditional religions were adapting to westernization and saw value in the accompanying cultural shifts toward rationalization and individual piety. More recent studies, particularly those informed by the war in Vietnam and concurrent critiques of neocolonialism and dependency, have taken a more pessimistic view of economic development. It has become much more common, therefore, to see analyses focusing on the role of religion in resistance movements, on exploitative alliances between regimes and established religions, and on the political implications of millenarian, messianic, and other grass-roots religious movements.

A second choice concerns what temporal and spatial framework we think is most useful to emphasize. The perspective implicit in modernization and lifeworld colonization theories, and in some interpretations of the other two frameworks as well, takes centuries as the appropriate time frame for studies of social change. How societies have evolved since, say, the thirteenth century is the central issue. Moreover. the spatial framework is often left unspecified, except for references to the West, Europe, or the capitalist system, or a concept of society (meaning the territorially bounded nation-state) is adopted. Over against these highly macroscopic designations, some studies have taken a much more specific spatial and temporal orientation, focusing for example on a specific event such as the overthrow of the shah in Iran or a series of related episodes such as the conflicts between Jews and Palestinians in Israel.

I shall suggest below that a more useful strategy than either of these extremes lies in focusing on religious changes over a period of several decades and involving at least several levels of spatial organization. The problem with the purely long-range macro-level theories is that they often ride too easily over the complex terrain of historical reality. The result is a closed theoretical system that seldom stands corrected by new empirical research. The problem with the more descriptive studies of single episodes is that such episodes are never in reality isolated from all others and, together, they may add up to changes of more massive proportions than any of their component elements might have suggested. Many of the religious conflicts we read about in the newspapers have prehistories and antecedents that need to be traced over a period of several decades or longer, and many represent reactions and counterreactions to other movements both in the same society and in the larger international system.

A third choice hinges on the extent to which we think events can usefully be organized into some coherent system or pattern. Modernization theory, world-system theory, and lifeworld colonization theory all assume that social change follows certain general patterns. Structural contingency theories assume that patterns of relations can sometimes be reconstructed in retrospect but emphasize the nonrecurrent nature of historical events to a much greater extent. Many descriptive studies and ethnographic reports, of course, deny the value of searching for broader patterns at all.

The social sciences are currently in much ferment over these questions in general. There is, however, at least one promising development — a development that perhaps registers the heightened sensitivity to language that has emerged in the social sciences in recent decades. This is the tendency to distinguish more sharply between the conceptual apparatus of the social scientist and the lived experience of social actors. Or put differently, it involves more nuanced distinctions among layers of abstraction. Consequently, one may speak of "bureaucratization" as a master trend in societal evolution and yet recognize that the specific manifestations of this trend in religious hierarchies and in government agencies may be quite different. Similarly, one may argue for the importance of economic resources to any episode of religious change but acknowledge that these resources may be constrained by organizational arrangements specific to particular societies.

Stated differently, the various concepts now available from a number of competing theoretical perspectives provide sensitizing devices rather than empirical indicators. To account for the shifting locus of public religion in a specific society, one uses these sensitizing devices to orient oneself to the relevant factors that must be considered. Much like an artisan’s tools, these concepts help produce a nuanced, multifactoral account of what happened. One does not expect to find changes in a different situation conforming to the same processes. Rather than working with a parsimonious specification of master trends, social laws, or correlations among variables, the analyst works with a complex set of concepts and empirical observations, piecing them together in ways that reveal underlying processes and interconnections.¹⁴

Fourth, there are choices to be made about where religion, or cultural patterns in general, fit in. At one extreme, world-system theory, as I have indicated, would at most regard religion as epiphenomenon subject to the more profound influences of political economy. At the other extreme, some variants of modernization theory have examined religious change solely within the context of its own internal unfolding; in such treatments, specific forms of doctrinal innovation seem to depend more on previous levels of ideological development than on any features of the economy or polity. Lifeworld colonization theory credits secular cultural patterns (e.g., rational communication processes) with an active role in social change, but it minimizes the importance of religion. Structural contingency theories are more likely to emphasize the institutional qualities of religion than some other approaches, but they often neglect patterns of value and belief.

At present, the rediscovery of culture in the social sciences, at the debate over methods of studying culture empirically, promises to shift studies of religion and politics more in the direction of looking at religious and political culture. Already, studies have been appearing in which greater attention is given to religious and political symbolism.¹⁵ Some studies have also begun to locate symbolism more squarely in the domain of everyday social practice, including discourse. sermons, speeches, and the emergence of dialogue within and between religious communities.¹⁶

Finally, in all of this, different emphases depend to some extent on how interested we are in change itself. There is, to be sure, a bias in the social sciences toward assuming that modern societies are faced with unprecedented crises. The language of predicament and paradox is altogether common, as are diagnoses that emphasize radical departures from the past, Significantly, a host of recent studies have challenged these biases and have shown the importance of continuities with the past even in situations of seemingly radical religious innovation.

Studies in which timeless truths are sought, which consequently pay little attention to questions of change at all, also remain of value. The relevance of the debate about change, however, is likely to deepen rather than abate. The question itself will undergo some degree of redefinition in the process. Whereas the earlier studies of modernization focused primarily on questions of movement from one point to another, more recent studies prompted by the other approaches suggest more complex and erratic trajectories of change.¹⁷ In examining the relations between religion and the state, therefore, it becomes especially important to consider what in particular has changed under certain conditions and what in particular has not changed.

Some Observations on Religious Restructuring

I have suggested, then, that many of the choices governing the ways in which we understand contemporary relations between religion and politics are contingent on the broader assumptions we find built into our major theoretical perspectives. How we evaluate particular studies depends as much on these assumptions as on more empirical issues and matters of evidence or validity. No single practitioner can dictate the choice among these broader assumptions. All that can be hoped for is greater sensitivity to the existence of these choices.

At the same time, I have suggested that certain shifts of perspective have been taking place in studies of religion and its public locations. We may hope that some of these shifts represent learning from past mistakes and a closer reckoning with the complexities of the subject itself. Let me, therefore, conclude by outlining a framework — a focus on religious restructuring — that draws together several of these recent emphases.¹⁸

For any consideration of the relations between religion and politics, the symbolic boundaries that divide institutions are clearly important. We need to know how religion is defined, how the political is defined, and how the two are separated from or connected with another. We may also need to understand other cultural divisions that bear on the relations between religion and politics. For example, the division between public and private may be important to understand, especially if politics is seen as the domain of the public and religion is a private concern.

An emphasis on institutional boundaries can be derived from modernization theory. It tells us to pay heed to the growing differentiation that characterizes modern society in general and modern religious systems in particular. What I have in mind here, though, is not the seemingly inevitable progression toward ever more complex cultural distinctions that has fascinated modernization theorists. It is instead the concrete demarcations that are constantly subject to negotiation in public discourse. Religion and politics may indeed be more formally differentiated as two distinct institutional domains in modern society than they are in traditional societies, but the events happening in country after country reveal that these domains are also in constant contact with one another. They interact and, perhaps more importantly, the content of each and the proper relations of each to the other are subject to continuous processes of negotiation. Officials stake out their turf, religious leaders contest these definitions, struggles ensue in tin courts or in assembly halls, movements mobilize to sharpen their claims, and the newspapers and television provide commentary.

Sensitivity to the ways in which cultural categories are defined can provide valuable clues to understanding the dimensions and directions of religious restructuring. One can think of these cultural definitions as fissures which, when subjected to stress, become major fault lines along which changes in religion take place. A minor rift between two ethnic factions, for example, can break into open warfare when aggravated by other social changes. More subtly perhaps, ambiguities about the character of public morality can provide a basis for religious conflict when one faction chooses to take a hard line on motivations and another takes an equally hard line on results.

The main point is that symbolic boundaries provide an empirical context in which to trace the likelihood of religious change or its aftershocks. When established boundaries fade away, space may be created in which new demarcations can arise with special significance. When the rights and responsibilities of individuals and groups are defined ambiguously, other social changes may render these tensions more opaque. And when such ambiguities have consequences for the distribution of social resources, acute struggles may be set in motion to determine how they are resolved.

If cultural boundaries can be likened to the fissures that run through the social terrain, then any effort to understand religious restructuring must also pay attention to the great forces that bring pressure to bear on these fissures and that effect a reshaping of the entire landscape. At the broadest level, some of these forces are likely to be understandable only in terms of dynamics in the larger world-system. At more proximate levels, one can also focus on domestic societal dynamics and on the shifts internal to religious organizations themselves.

Changes in the larger world-system are likely to consist of shifts in overall rates of economic growth, changes that reverberate from the rise and fall of great powers, alterations in international relations, variations in uncertainty and conflict, and even modifications of the extent to which people are aware of these larger relations. As suggested earlier, many of these changes occur in the relative near term and thus may have serious repercussions on religious and other institutions. These are likely to be exogenous shocks to any particular society. And they may have especially serious fallout for religious organizations. Even though religious leaders themselves may not be in the forefront of international trade or diplomatic negotiations, religious traditions are likely to be a significant part of what separates or joins two societies’ cultural orientations. Religious groups can facilitate or undermine the legitimacy of alliances, and any such alliances may spell victory, defeat, or at least minor alterations in the opportunities of population segments, whose identities are defined by religious commitments.

Domestic social changes are likely to constitute the more immediate conditions under which religious restructuring takes place. Many of these changes may ultimately have international dimensions as well, but it is the more proximate contexts to which religious loyahtst"o respond. Examples include the more or less routine transitions from one regime to another that accompany the electoral process, more abrupt political transitions that may come about as a result of coups or assassinations, policy initiatives that open new jobs in certain sectors of the economy and close down others, programs designed to alleviate human suffering and satisfy demands for minimal levels of social welfare, and social changes associated with the educational upgrading, professionalization, or gender redefinition of the labor force.

Shifts internal to religious organizations themselves are likely to focus on schisms and mergers, the resolution or reopening of conflicts between organizations, the assumption or abandonment of certain functions by religious organizations, or the emergence of new organizational forms. Social scientists have traditionally focused on a rather limited range of organizational forms in religion — specifically, churches, sects, and cults. More attention needs to be directed toward the conflicts and competition between organizations, federations as supraorganizational means of resolving some of these conflicts, and more innovative organizational forms such as parachurch agencies and special interest groups. It is within these organizational contexts that the opportunity to redirect religious energies lies. One can think of religious restructuring as partially motivated enterprise in which organizational leaders and followers produce ideas and actions within given institutional constraints.

The state is likely to be an especially important actor in all these changes. As the foregoing examples suggest, most of the social changes that happen in the relative short term (say, a decade to a half century or so) result from specific policy initiatives. These initiatives, in turn, may well represent responses to broader social conditions; and their consequences may be for the most part unforeseen and even undesired. But social change in the modern era is increasingly characterized by the intended and unattended consequences of planning, and such planning typically involves agencies of the central bureaucratic state.

We must not, however, assume that religious actors and organizations merely respond to changes set in motion by the state or by other agencies in the larger society. In our own context we can think of numerous instances in which major movements oriented toward social reform have been initiated or encouraged by religious organizations. Such initiatives, and similar movements oriented in opposition to government programs, have served as nettlesome reminders of the continuing ability of religious leaders to make their voices heard in the public square.

It is important to consider both the social and the cultural resources that religious actors may have at their disposal. In some societies private charity has kept religious organizations prosperous long after government contributions were cut off. The significance of an educated clergy, colleges and seminaries under religious auspices, and even places of worship that can double as meetinghouses for political purposes should not be overlooked. Religious actors also have at their disposal the narratives and rituals that continue to speak to primordial needs and concerns even in otherwise secular societies. Visions of hope and statements about meaning and purpose in life remain very much the preserve of religious functionaries.

Among these resources are the kinds of relations that structural contingency models have emphasized. It is, again, naive to assume that religious leaders operate entirely within their own domains while political functionaries stay exclusively within theirs. Preachers and mullahs in public office provide only the most visible exceptions. In one country after another, periods of peaceful coexistence between church and state have been made possible by carefully orchestrated deals. Government policies would not be criticized by religious leaders as long as religious organizations’ finances were not questioned. Central regimes would build roads, but local religious functionaries would still administer welfare relief. Religious organizations, therefore, are likely to have influence over at least some agencies of the state, and vice versa.

If religious organizations are strong enough to make their voice heard in public affairs, this voice will nevertheless be heard in ways that reflect religion’s social position. In the United States, conservative religious leaders have made themselves heard on decisions of abortion, homosexuality, and public morality. But their ability to carry the day has been limited by broader cleavages in the culture that put them against religious liberals on all these issues. Neither the conservative leaders nor their liberal counterparts have been able to speak with unquestioned authority as a result of this larger division. Where such conflict has existed in the past, it has often given other actors an opportunity to suggest alternatives to religious appeals of all kinds.

The concerns evident in Habermas’s framework gain special relevance here. Among its other roles, religion may be in an especially privileged position to raise questions about fundamental collective values. This position can be compromised, however, by religion’s relations to the state on the one hand, and to the marketplace on the other hand. Religion is a lifeworld that can be colonized by the forces of bureaucratization and monetarization. In some contexts, the temptation has been to build ever larger bureaucracies in hopes of becoming more effective at lobbying for religious causes. In other contexts, religious leaders have been seduced by the market model of success: striving for larger and larger churches, bigger operating budgets, more effective television programming, and so on. In either case, a concern for institutional aggrandizement subverts the capacity to think openly and critically about basic values.

All of these considerations assume, of course, that the relations between religious institutions and their host societies are not static. That assumption may be more appropriate in some societies than in others. Certainly it appears that the position of religion in a great many societies has taken new and unexpected turns in recent years. The specific turns can perhaps not be predicted beforehand, but some of the factors influencing these dynamics can be anticipated. Clearly, it is within the canon of most religious traditions themselves to regard their followers as a pilgrim people whose tents must always, in a sense, be in motion through the wilderness.

END NOTES

1. For an overview, see C. E. Black, The Dynamics of Modernization: A Study in Comparative History (New York: Harper & Row, 1966).
2. The work of Peter Berger represents one of the principal foci of modernization theory as applied to matters of secularization; see especially "A Sociological View of the Secularization of Theology," Journal for the Scientific Study of Religion 6 (1967): 1-18, and The Sacred Canopy (Garden City, N.Y.: Doubleday, 1967). For an overview, see Phillip E. Hammond, "Religion in the Modern World," in Making Sense of Modern Times: Peter L. Berger and the Vision of Interpretive Sociology, ed. James Davison Hunter and Stephen C. Ainlay (London: Routledge, 1986), pp. 143-58.
3. A theoretical treatment of these changes that remains most useful can be found in Robert N. Bellah’s Beyond Beliefi Essays on Religion in a Post-Traditional World (New York: Harper & Row, 1970), pp. 20-50.
4. On the dynamics of modernization and demodernization, see Peter Berger, Brigitte Berger, and Hansfried Keilner, The Homeless Mind: Modernization and Consciousness (New York: Random House, 1973); see also Daniel Bell, "The Return of the Sacred? The Argument on the Future of Religion," in The Winding Passage: Essays and Sociological Journeys, 1 960-i 980 (Cambridge, Mass.: Abt Books, 1980), pp. 333-45.
5. For a strongly stated argument against modernization theory’s assertions about religion, see Mary Douglas, "The Effects of Modernization on Religious Change," Daedalus 111 (Winter 1982): 1-20.
6. The basic perspective is outlined in Immanuel Wallerstein’s The Modern World-System, vol. 1 (New York: Academic Press, 1974), and The Capitalist World-Economy (Cambridge: Cambridge University Press, 1979). For a critical review, see Daniel Garst, "Wallerstein and His Critics," Theory and Society 14 (1985): 469-96.
7. For examples of applications and adaptations of world-system theory to questions of religious change, see Gary Nigel Howe, "The Political Economy of American Religion: An Essay in Cultural History," in Political Economy: A Critique of American Society, ed. Scott G. McNall (Glenview, Ill.: Scott, Foresman, 1981), pp. 110-36; Phillip E. Hammond, "Power Changes and Civil Religion: The American Case," in Crises in the World-System, ed. Albert Bergesen (Beverly Hills: Sage, 1983), pp. 155-72; George M. Thomas, Christianity and Culture in the Nineteenth-Century United States: The Dynamics of Evangelical Revivalism, Nationbuilding, and the Market (Chicago: University of Chicago Press, 1989); and Robert Wuthnow, Meaning and Moral Order: Explorations in Cultural Analysis (Berkeley and Los Angeles: University of California Press, 1987), pp. 2 15-64.
8. Without adopting the "world-system" label, other studies of the expanding world order have also focused on these disruptive consequences; see, for example, Eric R. Wolf, Europe and the People without History (Berkeley and Los Angeles: University of California Press, 1982), and Peter Worsley, The Three Worlds: Culture and World Development (Chicago: University of Chicago Press, 1984).
9. For specific criticisms of the world-system perspective along these lines, see especially Mayer N. Zald, "Theological Crucibles: Social Movements in and of Religion," Review of Religious Research 23 (1982): 317-36.
10. For two examples that account for historic variations in church-state relations in Europe by examining institutional arrangements, see Mary Fulbrook, Piety and Politics: Religion and the Rise of Absolutism in England, Wurttemberg, and Prussia (Cambridge: Cambridge University Press, 1983), and David Zaret, The Heavenly Contract: Ideology andOrganization in Pre-R evolutionary Puritanism (Chicago: University of Chicago Press, 1985).
11. David Martin’s A General Theory of Secularization (New York: Harper & Row, 1978) can be read as an application of structural contingency logic to the question of national variations in secularization.
12. For example, see Scott Mainwaring, The Catholic Church and Politics in Brazil, 1916-1 985 (Stanford: Stanftrd University Press, 1986), and Amrita Basu, "Grass Roots Movements and the State: Reflections on Radical Change in India," Theory and Society 16 (1987): 647-74.
13. See Habermas, The Theory of Communicative Action, 2 vols. (Boston: Beacon Press, 1984-87). For a brief critical overview of Habermas’s arguments, see Hugh Baxter, "System and Life-world in Habermas’s Theory of Communicative Action," Theory and Society 16 (1987): 39-86.
14. I believe this is one of the lessons to be learned from the recent theoretical work of Jeffrey C. Alexander in Theoretical Logic in Sociology, 4 vols. (Berkeley and Los Angeles: University of California Press, 1983) and Action and its Environments (New York: Columbia University Press, 1988). For an appreciative essay on this contribution, see Randall Collins, "Jeffrey Alexander and the Search for MultiDimensional Theory," Theory and Society 14 (1985): 877-92.
15. For an empirical study that argues strongly for mapping out the cultural terrain of religion and politics, in opposition to Marxist interpretations, see Karen E. Fields, Revival and Rebellion in Colonial Central Africa (Princeton: Princeton University Press, 1985).
16. Studies of liberation theology have been especially cognizant of these relations between ideology and practice. See, for example, Clodovis Boff, Liberation Theology from Confrontation to Dialogue (New York: Harper & Row, 1986). In a different setting, see the emphasis on "symbolic practice" in Jean Comaroff’s Body of Power, Spirit of Resistance: The Culture and History of a South African People (Chicago: University of Chicago Press, 1985).
17. These erratic trajectories receive special attention in the essays included in Religious Resurgence: Contemporary Cases in Islam, Christianity, and Judaism, ed. Richard T. Antoun and Mary Elaine Hegland (Syracuse: Syracuse University Press, 1987).
18. My own attempt to utilize a perspective of this kind is The Restructuring of American Religion: Society and Faith since World War II (Princeton: Princeton University Press, 1988).

The theoretical tradition initiated by Max Weber nearly a century ago has remained a popular perspective from which to examine religion, ideology, and, in general, the processes of change in these cultural systems. The Weberian perspective gives credence to the importance of ideas in social life. It poses specific arguments about rationalization, charisma, political legitimacy, and religious ethics. It offers a methodological orientation that emphasizes subjective meanings. In disagreeing with Marx, Weber countered what in his view was a one-sided materialism with a more complex model in which ideas could also exercise causal influence. In contemporary sociology, Weber has often been a source of inspiration for interpretive, ethnographic, and historical approaches to religion advanced in opposition to narrowly reductionistic, positivistic research designs.

Yet certain limitations have been evident in the Weberian tradition. The manner in which religion and ideology are conceptualized has resulted in what sometimes appears to be an overly subjective view of their nature and functioning. In the extreme, they become moods and motivations operating at the level of internalized values instead of observable features of discourse, language, or institutions. Supporters of the Weberian perspective have countered that this is, indeed, an extreme view, based more on a narrow reading of The Protestant Ethic and the Spirit of Capitalism than on an understanding of Weber’s more extended and mature works, especially Economy and Society. Yet even these works have been criticized on the same grounds.

Causal statements sometimes revolve around notions of "elective affinities" among subsets of otherwise highly complex ideological systems. Why particular ideas become operative in particular situations remains unexplained. Moreover, comparisons with other contexts cast doubt on whether these affinities are universal. In addition, the social conditions bringing about changes in ideas are often left unspecified or are attributed too readily to notions about the interests of particular status groups.

Critics charge that Weber’s arguments go beyond notions of elective affinity in emphasizing the role of status groups as carriers of new ideas, but the intervening mechanisms relating particular ideas to particular status groups still leave much to be explained. Peasants, for example, are said to believe in God because they see supernatural acts in nature; soldiers are said to believe in fate because they risk an early death. But these explanations focus on psychological factors rather than providing insight into the institutional contexts in which ideas are actually produced, paid for, brought into contact with an interested audience, enacted in collective rituals, used to mobilize resources against competing ideologies, and embedded in social arrangements.

These limitations have been particularly evident in Weberian treatments of English and American Puritanism. Inspired by Weber’s thesis concerning the relations between inner-worldly ascetic religious values and the spirit of capitalism, numerous inquiries have examined the ideological components of Puritanism, its social origins, and its consequences. This body of literature has generated some useful hypotheses about the effects of Puritanism on the rise of science, its legitimation of revolutionary dissent, and the qualities of religious rationality, to mention only a few of the themes that have been pursued. Beyond these specific historical questions, this literature has also stood as an example of how one might use the Weberian tradition to derive generalizable propositions about the nature of religious change. Scholars have looked to it for inspiration in thinking about such questions as the origins of new theological ideas, the relations between changing religious views and changing views of the state, and the religious legitimation of secular knowledge.

But this body of literature also remains the subject of much controversy. Whether Puritanism as ideology had significant independent effects on related social developments remains in dispute, as do matters associated with its content and the conditions facilitating its appeal to particular groups. The issue of Puritanism casts into sharp relief all the difficulties of relating the sacred to its social environment — difficulties of measurement, difficulties of accounting for change, and difficulties of establishing relationships between complex systems of ideas.

Much of the controversy surrounding Weberian arguments about Puritanism has focused on substantive issues. For instance, historical inquiries have variously identified the radically uncertain Puritan soteriology, the Puritan doctrine of the elect, and the implicit Puritan humanitarianism as its key ideology. Others have asked whether Puritanism was truly a seventeenth-century innovation or whether its roots extended not only to the magisterial Reformation but also to the highly rationalized religious concepts of the Middle Ages. Still others have debated the specific involvement of Puritans in such developments as the early scientific academies or the cadres that advanced ideas about civil insurrection.

So numerous have these studies been that scholars have on occasion expressed doubt about the likelihood of much of value coming from further work of this type. Interest in the Weberian thesis, however, appears to remain strong in large part because, for all the historical particularity of Puritanism, it constituted one of the most formative episodes in our own nation’s history and the history of the modern West. Interest in the Weberian study of Puritanism also remains strong because of the theoretical implications of Weber’s work for understanding the sacred in our own time.

Recently, several important studies have appeared that, in addition to contributing historical evidence on Puritanism, also offer some interesting new ways of thinking about the theoretical assumptions concerning religion and ideology in the Weberian tradition. The theoretical roots of these studies tap directly into Weberian arguments, but the studies provide insights about religion and ideology that should be of value well beyond the Weberian school itself.

The primary contribution of these studies is to suggest some of the specific social processes that intervene between general social or cultural conditions and the appearance of an ideology. They suggest, in different ways, that ideologies are not related to the social environment through some diffuse mechanism such as affinity, legitimation, or isomorphism but are produced in specific institutional contexts by organizations, movements, and agents that draw resources from the environment and take specific actions that facilitate or impede the adoption of particular ideologies. Given this, these studies are clearly relevant to current debates inspired by traditions other than Weber’s as well — for example, neo-Marxist discussions of the state and ideology and poststructuralist treatments of cultural production.

The Heavenly Contract

David Zaret’s The Heavenly Contract is an effort to examine in greater detail the origins of one doctrine in English Puritanism that Weber thought particularly significant in the legitimation of rational industrial capitalism — the idea of a covenant between the believer and God.¹ "Covenant theology," as it was called, made it possible for individuals to gain evidence of their divine election by cultivating a disciplined life of pious devotion and godly deeds. Thus, it tempered the earlier predestinarian emphases of Calvinism by introducing notions of contract, exchange, and reciprocity into the relations between God and the individual.

It was this doctrine that Weber emphasized in arguing that the inherent uncertainties in Calvinist theodicies had been modified in Puritanism to the point that the fruits of disciplined activity in an ascetic calling could be taken as a sign of one’s salvation. Zaret argues that the importance of this theological development for the Weberian treatment of religion and capitalism suggests a need for an adequate account of its origins.

In contrast to previous accounts that focused on theological antecedents of covenant theology, Zaret’s study offers a sociological approach that draws inspiration from recent attempts to rewrite social history from a more popular, bottom-up perspective. He focuses mainly on the changing relations between laity and clergy and the ways in which religious ideas were produced.

The explanatory model for The Heavenly Contract starts with an assertion of the importance of social-structural, technical, and cultural changes in English society. These changes produced an active laity more interested in taking part in popular dissent movements. Among these changes, crucial developments included the penetration of market forces into the countryside, where the vast majority of the population still lived; the appearance of a popular press and the growth of literacy; and a set of ideas conducive to popular dissent, such as anticlericalism and a more rationalized orientation toward material and spiritual life.

Then, on the clergy side, the model focuses on the manner in which preachers responded to and attempted to seize control of the emerging popular dissent. Ideas about the sanctity of conscience and the importance of godly edification were emphasized and subtly modified by clergy as they attempted to secure popular submission to the secular authorities, regain ecclesiastical control of the laity, forge alliances with the state, and secure patronage from the aristocracy.

With these background characteristics of the laity and clergy considered as preconditions of religious change, the model shifts to particular movements that developed as competing ways of organizing the new relations between laity and clergy. Competition among three specific groups — lay Puritans, Separatists, and radical heretics — seems to have been most important. Zaret argues that the controversies among these groups hinged chiefly on the ambiguous position in which clergy found themselves. On the one hand, they served as ordained ministers subject to the authority of the church; on the other hand, they played an increasingly active role as leaders of a popular movement. In short, clergy were caught between the expectations of their ecclesiastical superiors and a restive lay following. It was in this context that covenant theology emerged.

Clergy were hard-pressed to retain their control over the lay movement and often found themselves advancing doctrinal compromises with the more extreme demands of the laity or, in response to organizational pressures, advancing tentative solutions that turned out to have unintended consequences. Covenant theology developed from a mixture of intellectual precedents that were articulated and modified in response to organizational pressures. It drew heavily on contractarian ideas concerning the believer’s initiative in responding to divine election. Such practices as reading the Bible, hearing sermons, practicing family worship, and examining one’s spiritual life all came to be emphasized as evidences of divine election.

These elements of covenant theology not only solved some of the immediate dilemmas of church polity that the clergy faced but also made sense to laypersons because of their similarity to contractarian ideas gaining prominence in the world of finance and trade. Ultimately, therefore, the affinities between Puritan ideas and economic ideas that Weber identified are important features of the argument, but Zaret specifies in greater detail than Weber had the conditions under which these ideas came together.

Zaret’s study is based on an extensive reexamination of sermons, lay memoirs, and historical studies of ecclesiastical debates and broader social conditions. Although it is limited to the English case, and therefore does not benefit from the broader kinds of comparison Weber made, it utilizes a rich set of sociological conceptions and pieces them together to form a detailed process model involving interactions among social-structural, cultural, organizational, and interpersonal factors.

The virtue of Zaret’s approach is that it demonstrates the relevance of broad social conditions as effects on religious ideas, but instead of pinning these effects on hidden psychological states or vague generalities about interests and legitimation, it traces the factors actually involved in the specific contexts in which ideas were produced, modified, and disseminated. One comes away with the feeling that a plausible set of connections between social conditions and religious change has been identified.

Piety and Politics

Mary Fulbrook’s comparative study of Puritanism and Pietism in England, Wurttemberg, and Prussia also contributes to the current reassessment of Weber’s ideas.² Asking why religious ideas favored absolutism in Prussia in contrast to a politically passive orientation in Wurttemberg and an anti-absolutist attitude in England, Fulbrook is led to examine the interaction between religious ideas and the social contexts in which they take shape. Like Zaret, she concludes that ideological outcomes cannot be explained simply on the basis of theological antecedents. Indeed, she fails to find support for a Weberian style argument that ideas bear an elective affinity with the particular social, economic, or political structures they help to legitimate. Nor does she find any support for a straightforward effect of social class or status groups on ideologies. She is forced, like Zaret, to examine religious ideologies within a broader, more complex pattern of "sociopolitical environments."

 She observes that Puritanism in seventeenth-century England and Pietism in eighteenth-century Wurttemberg and Prussia were quite similar in general theological orientation. All three emphasized personal purity and conversion, a life of diligence in attending to the scriptures, a clear vision of God’s kingdom becoming a moral community on earth, and an active inner-worldly approach to social improvement and moral asceticism. Yet the three varied dramatically in the specific attitudes that developed toward absolutist government. Why?

Her argument suggests that general ideas are transformed into specific political ideologies by the social obstacles that religious communities confront in pursuing their goals. She recognizes that in the cases under consideration, religious sentiments were embedded largely within the institutional context of a state church. The opportunities for religious reform movements to achieve their goals were largely determined, therefore, by the relations present among the state, state church, and other powerful interest groups.

English Puritanism developed an anti-absolutist ideology because it found a relatively strong popular base among the laity but was opposed by the state church, which was closely linked to the king who gave it patronage in return for ideological support. Pietism in Wurttemberg took a politically passive turn because it was largely tolerated by the state church, which was somewhat independent of the king and capable through the involvement of the aristocracy of incorporating new ideas about the nature of the polity. In short, Pietism became anti-absolutist but not revolutionary because mechanisms were available for gaining its ends peacefully. Prussian Pietism, in contrast, developed in the context of a feudal aristocracy that gave it a relatively weak economic base. Indeed, Prussian Pietists found their greatest support from the centralizing state itself, which was struggling to gain control over the feudal aristocracy; so Pietism gradually took on political attitudes that supported absolutism.

Fulbrook’s study does not have the nuanced series of variables linking social conditions and religious ideologies that Zaret’s does. But she is able to advance a highly parsimonious model for the differences she seeks to explain. Rather complex sets of historical variables are summarized in the relations among state church, state, and aristocracy, and these relations then determine where a religious reform movement is likely to turn for support, which becomes a decisive factor in shaping the movement’s views of the state.

Wallington’s World

Wallington’s World, by historian Paul Seaver, represents a contribution to the Weberian legacy of a quite different sort.³ Nehemiah Wallington was a Puritan artisan who lived in London from 1598 until his death in 1658. Seaver’s study is based on Wallington’s extensive memoirs, religious reflections, and letters — some 20,000 pages in all, of which 2,600 survived. Seaver uses these materials to develop an intensive portrait of the mental outlook typified by this Puritan layman.

The topics covered range over some of the terrain most frequently trod by Weberian scholars: the doctrine of the calling, questions of divine election, evidences of salvation, attitudes toward money, sexual and family values, standards of morality. Seaver denies that Wallington can be regarded as a typical Puritan artisan (the fact that he wrote so much was itself unusual), but he presents enough material in addition to that of Wallington — from Puritan sermons and other autobiographies — that the reader is likely to come away feeling that he or she has learned something about Puritans in general.

In contrast to Zaret’s and Fulbrook’s studies, Seaver’s pays little attention to the broader social conditions that may have shaped Puritan ideology. The value of his study lies mainly in its portrayal of the ideology itself. If we agree with Zaret’s claim that popular ideology among the laity has been a neglected feature of Puritanism, then Wallington’s memoirs take on added significance: they give us a glimpse beyond the reconstructed categories that Weber and others have applied to Puritanism. Religious ideology ceases momentarily to be a set of ideal types and becomes actual discourse. We are able to see the words and categories that Puritans actually used to create their sense of reality. This, I believe, is of particular value for rethinking some of Weber’s ideas about social-psychological factors in religion.

Religion and Rebellion

In an essay in Bruce Lincoln’s volume Religion, Rebellion, Revolution, Christopher Hill also considers the lay aspect of Puritanism.⁴ In a sense, Hill’s story picks up where Zaret’s and Seaver’s leave off. He is concerned more with the second half of the seventeenth century than with the first. He concedes the importance of covenant theology but focuses on its breakdown rather than its origin. Like Fulbrook, he considers the radical tendencies in Puritanism, but he looks more at the conflict between rich and poor than at that between new merchant elites and the state. His discussion emphasizes the fluidity of ideas, the way in which resolutions in one period become problems in the next, and the importance of paying attention to inequalities among groups concerned with ideologies.

The phenomenon Hill seeks to explain is the shift away from a covenant theology that applied to the elect minority toward a more universalistic doctrine of salvation that applied to the majority. He suggests that this shift reflected the changing social positions of rich and poor. By 1640, the inequality between rich and poor had apparently become accentuated. Covenant theology was attractive to the rich because they had sufficient wealth to lead the godly life it required. The covenant idea, however, also required a clear doctrine of sin as a means of controlling the poorer segments of society.

Even so, Puritanism only partially solved the problem of social control, and in the 1640s the poor were able to voice their demands more insistently. An emerging antinomianism temporarily allowed competing ideas to flourish, and universalism became increasingly popular among the poor. It was a solution to the problem of sin that had been imposed on the lower classes in conjunction with the covenant theology of the elite. The elite’s response between 1640 and 1660 was increasingly to downplay the idea of sin and, indeed, to discourage reflection about the nature of God more generally. What emerged was a deistic philosophy in which the ideas of sin and God receded in favor of new social control mechanisms provided by law and legitimated by conceptions of the lawfulness of nature.

 Two Treatises of Government

The lawfulness of nature was, of course, a central argument in Locke’s political philosophy. By the end of the seventeenth century, when Locke’s ideas began to flourish, the influence of Puritanism (if Hill’s thesis is correct) was already on the wane. Yet there has also been much speculation, often prompted by explicit considerations of Weber, that Puritanism had paved the way for the Lockean theory of democracy through an elective affinity of ideas.

One of the most ambitious, and for the most part successful, attempts in recent years to examine religion and ideology within its historic context has focused squarely on this connection between Puritanism and democratic theory. Taking Locke’s Two Treatises of Government as the particular form of this theory requiring explanation, political scientist Richard Ashcraft has reexamined every relevant piece of information to provide a novel new account of the social conditions under which this influential contribution was made.⁵

The broader purpose of Ashcraft’s study is to demonstrate by example the virtue of taking political and religious ideology out of the realm of abstract philosophical discussion and considering it "in relation to a socially defined audience whose members seek to obtain certain practical advantages through social action."⁶ Ashcraft pursues this objective by demonstrating that Locke did not write the Two Treatises of Government as ivory tower reflections; to the contrary, the treatises grew out of the dynamics of a radical political movement that began in the 1670s, incorporating the exclusion crisis, the Rye House plot, and Monmouth’s rebellion, and culminating in the Glorious Revolution of 1688.

Ashcraft also argues that ideology needs to be understood, as Zaret suggests, within a matrix of previous concepts, slogans, tracts, and arguments that establish the framework in which discussion can take place. He supplies a vivid example by demonstrating that Locke’s use of the phrase "an invasion of rights" (from which our own phrase "invasion of privacy" developed) grew out of an encoded form of discourse used by Shaftesbury and his fellow conspirators. The agents sounded out potential recruits (usually in pubs) by engaging them in discourse about "invasions." The meaning of this ambiguous term was clarified to indicate a literal invasion only when the recruit’s loyalties had been established.

The book also provides some useful examples of the ways in which ideologies become articulated with, or reflective of, their social environments. For example, Ashcraft argues that the interest of Locke (and his contemporaries) in epistemology emerged in relation to claims advanced by the clergy and other defenders of the crown that knowledge of the common good could only come from professionally institutionalized sources — namely, the church and state bureaucracies. Ashcraft also shows how electoral politics and religious divisions resulted in an emphasis on certain kinds of ideologies, such as those asserting rationality or arguing for legal guarantees of property relations.

Weberian Themes

These studies are for the most part compatible with, and even sympathetic toward, Weber’s general approach. Indeed, they draw repeatedly on Weber’s insights. For example, Zaret not only adopts Weber’s idea of Puritan emphasis on evidences of salvation as his main dependent variable but also borrows directly from Weber in identifying price-making markets as a source of practical rationality.

Yet there are also efforts in these studies to suggest explicit reformulations of Weber. Zaret believes that his conclusions are consistent with Weber’s general views on the importance of rationalization in Western societies but suggests that his study challenges Weber’s specific ideas on the relation between ultimate values and practical ethics. According to Zaret, his evidence documents a strong causal effect running the other way — from rationalized contractarian norms in economic relations to the more abstract doctrine of covenant theology. From this finding, he adduces that Weber was probably wrong in assuming the necessity of rationalized religion for establishing a uniform hierarchy of values that challenged traditionalism and supplied motivation for the pursuit of rational economic activities. Contrary to Weber’s effort to show "how abstract ethical doctrines could influence everyday life," Zaret concludes that "practical ethics in profane activities can be no less influential for the formation of abstract doctrine."⁷

Fulbrook also suggests that her study challenges a strict Weberian interpretation of the relations between religious ideas and practical outcomes. She casts doubt on the assertion that there is an intrinsic affinity between certain religious ideas and political orientations by showing how similar ideas lead to different outcomes in different settings. An extreme interpretation of her findings is that ideology does not matter at all. A more cautious interpretation is that it matters but does so because it is malleable rather than fixed, changeable instead of static. Abstract ideas take on meanings within specific contexts that then motivate people to engage in one kind of action or another.

None of these studies is pitched at a general theoretical level, but they do offer theoretical implications that are worthy of more general consideration. Weber’s interest in the relations between Puritan ideology and broader economic development is clearly evident in these studies. So is his interest in the relations between Puritanism and ideas reflecting rationality in the political sphere. But whereas Weber emphasized the correspondence between ascetic Calvinism and acquisitive capitalism as ideal types, the later studies we have considered here examine the actual social conditions under which Puritan ideas were produced and modified. Attention is shifted away from general relations to intervening processes, from motivations and legitimation to the contexts of ideological production, and from functionalist or consequentialist arguments to considerations of social mechanisms that bring ideology into contact with relevant audiences and actors.

We have here a set of variables that begins to make the connections between social structure and religious ideology more plausible. Instead of being left to rely on deductive theoretical formulations about the economic or political interests served by ideologies, we are offered attempts to spell out a credible set of social factors that relate the actual producers and disseminators of ideas with their social environment, with interested audiences, and with sequences of action that put ideas into effect. It is possible to identify several types or clusters of variables that may, in various ways, be relevant generally to the study of change in religion and ideology.

An Approach to Religious Change

Beginning with broad environmental conditions that may serve indirectly to produce a climate conducive to the emergence of new ideas, we can identify variables that affect the likelihood of a potential audience being present in the first place. Zaret identifies three types of variables that probably have general relevance to the consideration of audiences: (1) social-structural variables that concern the size, social location, and access to material resources of a potential audience; (2) technical variables that specify some medium by which ideas can be communicated to this audience; and (3) cultural variables that deal with the general symbolic or mental capacities that the audience has at its disposal for processing the ideas to which it is exposed. Examples of these variables might include (1) the diffusion of markets and the integration of vocational groupings into market networks, (2) the technical capacity to produce and distribute printed materials, and (3) individuals’ familiarity with styles of reasoning and engaging in rational theological or political argumentation.

Zaret’s emphasis on audiences probably needs to be expanded in thinking about the relevance of social environments to ideology at the most general level. Audiences are not the only factor involved in changing or sustaining an ideology. Other resources are also required: material resources, such as the wherewithal to pay for printing, buildings, and salaries; and symbolic resources, such as culturally defined categories of thought and ritualized enactments of ideas or formal legitimation from scholars, professional experts, economic elites, and public officials. In addition, resource environments can be characterized in other ways than by the sheer supply of specific kinds of resources. More abstract attributes of the environment, such as its heterogeneity or instability, are also likely to be important. For example, an increase in environmental heterogeneity may facilitate the emergence of ideologies having more loosely connected elements, such as an individualistic ideology that decouples specific tenets by regarding them as matters of personal preference.

At a level more proximate to the actual production of ideology, institutional contexts also become significant theoretical variables. They interact with the broader environmental conditions in ways that shape the eventual form and content of ideas. These variables help greatly to specify the intervening mechanisms by which larger social-structural conditions affect ideologies. Instead of having to posit a direct, psychological, often mysterious effect on disaggregated individuals, the investigator is able to show that ideas are actually produced by specific sets of people who have access to necessary material and cultural resources, are influenced by particular organizational constraints, and bring their ideas into contact with an audience that has reasons to be receptive.

In dealing with the institutional contexts in which ideologies are produced, one’s attention must include consideration of inequities in the distribution of social resources. These inequities are not only important for determining who has access to means of cultural production; they also become problems with respect to social control that may result in attempted ideological resolutions. Even though ideologies may be advanced in the name of a small minority, rationalizing tendencies of the kind Weber observed may require conceptions of the relative statuses of people more generally. These kinds of analyses — studies of the relations between ideologies and status groups — tend to become unsatisfactory to the extent that they stop with mere assertions about interests and legitimation. Further efforts are generally needed to specify how the interests of particular status groups are brought to the attention of the producers of ideology, how these producers respond, and what constraints influence their response.

A clue that may often be helpful in addressing these questions lies in the notion that cultural producers themselves have a stake in maintaining their position and may engage in creative compromises in order to do so. Periods of rapid social change, such as revolutions, may expand the opportunities for such innovations by altering distributions of resources, upsetting institutionalized roles, generating uncertainties in moral obligations, and setting in motion competing ideological movements.

Fulbrook’s study demonstrates the value of recognizing not only that ideas are produced within institutions but also that these institutions have important relations with other institutions. An established church may be the immediate context in which a religious movement appears, and it may be the focus of the movement’s ideology, but this institution is in turn likely to depend on the state and the economic hierarchy for support. The broad environmental changes that Zaret identifies are mediated not through a single institution but through a set of related institutions. The effect of these broader changes is not only to create a lay audience for ideological reform but to influence the relations among institutions. Fulbrook’s study illustrates especially well the importance of resources such as political support and economic patronage, both of which may depend less on the nature of the popular audience than on macro-level linkages among institutions. Her work, therefore, points to the possibility of making genuine contributions to the study of religious ideology by focusing on broad institutional relations.

An additional set of relevant variables may be termed action sequences. These are concerned with the fact that cultural producers actually do something to generate cultural innovations. These action sequences include the responses of actors to stimuli in their immediate situations — specifically, responses that involve attempts to articulate ideas, to make sense of things, to offer interpretations acceptable to relevant constituencies. The significance of conceptualizing these responses as a distinct set of variables is that cultural innovations are seen as a result of specific actions, not just general phenomena that emerge automatically and mysteriously from structural conditions. These actions also incorporate an element of agency or voluntarism into the theoretical framework, like Weber’s idea of the charismatic leader, only in a broader sense.

Several considerations of a more specific kind are relevant to the analysis of action sequences that result in religious change. One is the importance of specific crises or triggering events that lead actors to utter pronouncements or offer interpretations. Zaret calls these pressures. For example, Puritan clergy had to respond to a series of specific demands or controversies, such as nonconformity in wearing the surplice and using the cross in baptism. These immediate crises may, of course, also become symbols that stand for positions on more general issues.

Another valuable consideration in treating action sequences is the idea of compromise or synthesis. The new religious views that are produced often represent a bringing together of several formerly distinct strands of thought. In some instances, close analysis of this process can actually show how specific cultural categories were drawn upon, combined, and modified. The result can be the emergence of a new dominant idea (the idea of "covenant" in the Puritan case), a new model or metaphor for thinking about moral obligations (contractarian metaphors in Lockean theory), or a new form underlying the relations among ideological elements (individual conscience as a decoupling mechanism). In other words, the outcome of an action sequence may be a specific idea or concept distinguished by explicit substantive content, or an underlying metaphor that influences ideas without becoming a key concept itself, or a formal structure that has to be inferred from the relations among ideas.

An implicit consideration in the analysis of action sequences is the existence of several competing ideologies, interpretations, or interpreters. Actors are likely to compete with one another to gain control over a situation by making sense of it or imposing an ideological definition on it. Their competition may stem from their having different organizational locations, different interests, or different cultural materials with which to work. In the Puritan case, for example, several radical Separatist groups emerged alongside the more moderate Puritan clergy. Once such competition has developed, its very presence is likely to influence the manner in which ideas are articulated. In addition to stating their own views, actors are likely to engage in real or imaginary dialogue with their competitors, attempt to debunk or neutralize competing ideas, or use them as illustrations of wrong-headed or unworkable proposals. For example, the idea of a heavenly contract gained cogency among Puritan clerics at least in part because it was used to support specific arguments against radical heretics’ ideas about adult baptism and free will.

Whenever action sequences involve competition between different theological views, the role of immediate institutional constraints and environmental resources is likely to be that of influencing the selection process among different ideas. That is, certain ideas over time are increasingly "selected for" because they fit the situation better (make sense of it and have resources at their disposal), while other ideas are "selected against."

For instance, Puritan emphases on self-control and introspective examination were increasingly selected for after 1590 when hope was lost that the state could be used as a resource for imposing ecclesiastical controls on the laity. From the standpoint of specific clergy or theologians, this sort of selection process may well have resulted in unforeseen outcomes. Actors respond to events with partial knowledge of relevant conditions and articulate short-term visions of what needs to be done, and the significant cultural outcome is more likely to depend on the cumulative trajectory of these decisions than on any clearly articulated plan set forth at the beginning.

What we have, then, as a general framework implicit in these studies is a three-factor model of cultural production. It consists of environmental conditions, institutional contexts, and action sequences. Considering all three sets of variables as well as the relations among them enables the analyst to look at the possible effects of very general social conditions on ideas and yet to identify some plausible intervening links between these conditions and the specific ideas that result. Weberian notions about master tendencies in modem culture (e.g., the tendency toward rationalization) and the mutual relations between these tendencies and other developments (e.g., industrialization and bureaucratization) can thus be subjected to concrete empirical investigation. We need no longer leave these relations at the level of correlational affinities if the process by which these affinities develop begins to take on specific shape.

Weber, Marx, and the Subjective

One further issue is Weber’s emphasis on subjective meanings. Continental sociologists have often been critical of their Anglo-American colleagues for stressing too heavily Weber’s interest in the subjective. Yet it is clearly not an exaggeration to say that Weber’s emphasis on the Protestant ethic, on the spirit of capitalism, and on meaning, motivation, and verstehen all underscore an interest in the subjective. This interest continues to make Weber attractive to many sociologists, but the feature is not without its problems. In the case of Protestant capitalist theses, it has resulted in emphasis being placed on subjective affinities between sets of ideas rather than on the institutional settings in which ideas are actually produced, as in the work of Zaret and Fulbrook. More generally, it has led to an approach to religion and ideology that is methodologically problematic because it requires analysts to probe subjective meanings buried away in people’s heads. How was it, then, that Weber was able to pay so much attention to the subjective?

The answer generally tendered is that Weber’s concern with the subjective developed in response to, and can be taken as a corrective to, Marx’s economic determinism. According to this interpretation, Weber wanted to demonstrate that people do not live by bread alone, as it were, but are guided by the internal motivations and meanings that come from their ideas. This explanation may be valid, although the contrast it suggests between Weber and Marx is probably overdrawn, given Marx’s interest in the subjective problem of alienation. Nevertheless, the question remains of why a critical observer such as Weber was able, from his early interest in the Puritan ethic, to sustain the conviction that one could make reliable statements about subjective meanings. An answer to this question can be found in Seaver’s treatment of Wallington.

We know a great deal about Wallington’s subjective worldview because he wrote everything down. His penchant for writing, while perhaps more extreme than that of his contemporaries, was itself a reflection of an important Puritan concern — what Seaver calls "the examined life." Wallington wrote everything down as a way of objectifying his own thoughts and feelings. He did so because the Puritan idea of a covenant between God and the elect providing evidence of one’s election motivated the believer to keep accounts, to know his thoughts and feelings in order more carefully to discipline them and to guard against temptation.

Moreover, Puritanism provided a rich set of available scripts to facilitate turning one’s inner feelings into objective narratives. In Wa!lington’s case, for example, we are bombarded with direct quotations and paraphrases from the Bible, with explicit references to Puritan doctrines (such as justification and election), and with rich metaphorical and allegoric images (about beasts, trees, illnesses, Journeys, and so on). Wallington seldom speaks of a subjective state at all; rather, he obje tifies his self, using a host of concepts such as "saint," "soul," "bosom," "the godly," "the Lord’s dear one," and so on. It is almost as if Wallington is writing about someone other than himself, or about himself only as a member of a broader category. He expresses deep emotion and inner feelings, but in highly codified ways: he has "peace and comfort," "takes delight," "strives mightily in this heart," "lays hold upon," "experiences conflict between the flesh and the spirit." What we have, then, is a situation in which introspection is demanded on the one hand, and made possible by a readily available and exceedingly rich set of cultural codes on the other.

The Puritan setting was perhaps unique in this respect. Indeed, Hill argues that by the end of the seventeenth century theological emphases had shifted in a way that militated against self-examination of this kind. The Puritan case, in this sense, provided Weber with a rather unusual glimpse of the subjective. It was not at all like the modern research interview in which an observer attempts to elicit information about subjectively held attitudes from individuals who have never reflected on their feelings until the moment when they are presented with preceded questions that are not part of their own subculture. If anything, the Puritan case was more like a clinical setting in which clients talk about feelings in highly codified terms that have been provided by psychoanalytic theory, learned by interacting with the therapist, and sanctioned by ideas about the value of self-examination.

The special qualities of Puritan discourse probably need to be kept in mind in attempts to imitate Weber’s concern with subjectivity. What he was getting were messages about subjective meanings that were filtered through a very formalized, socially constructed set of cultural categories. It may be, therefore, that less attention should be given to the idea of subjective meanings and more research should be devoted to the socially constructed categories used in discourse about these meanings.

This conclusion also leads to a more general point about the nature of culture that is consistent with the previously mentioned criticism leveled by Zaret against the idea of abstract values legitimating practical ethics. Sociological theories of the 1940s and 1950s tended to portray culture as a hierarchy of values that was integrated by higher-order religious values and was internalized by the person, thereby giving unity and direction to the person’s behavior. Although some of these discussions saw inspiration for this view in Weber, it can be argued that it was more generally derived from common notions in American political theory. More recent discussions of culture have largely abandoned this emphasis on a hierarchy of values and have focused to a much greater extent on discourse, symbolism, and communication. From this perspective, the essence of culture lies less in its internalized capacities for integration and motivation than in its communicative potential. On theoretical grounds, therefore, the subjective aspect of culture becomes less important than the objective components evidenced in texts, discourse, and expressive behavior.

Drawing together the various arguments suggested by these studies, then, we see a view of religion and ideology rather different from that advanced by Weber. Religion is envisioned not as a set of internalized values that influences an individual’s moods and motivations but rather as a codified set of concepts and categories that is evident in discourse, reinforced by practical commitments, and advanced in institutional settings. To understand how ideas change, we need to consider not only subjective needs and values but also the relations between actors who articulate ideas and actors who provide an audience for these ideas, the institutional contexts in which these dynamics take place, and the larger social resources that institutions have at their disposal.

END NOTES

1. Zaret, The Heavenly Contract: Ideology and Organization in Pre-Revolutionary Puritanism (Chicago: University of Chicago Press, 1985).
2. Futbrook, Piety and Politics: Religion and the Rise of Absolutism in England, W/irttemberg, and Prussia (Cambridge: Cambridge University Press, 1983).
3. Seaver, Wallington ‘s World~ A Puritan Artisan in Seventeenth-Century London (Stanford: Stanford University Press, 1985).
4. Hill, "Popular Religion and the English Revolution," in Religion, Rebellion, Revolution, ed. Bruce Lincoln (New York: St. Martin’s Press, 1985), pp. 46-68.
5. Ashcraft, Revolutionary Politics and Locke’s "Two Treatises of Government" (Princeton: Princeton University Press, 1986).
6. Ashcraft, Revolutionary Politics and Locke’s "Two Treatises of Government," p. 7.
7. Zaret, The Heavenly Contract, p. 208.

How do we assess the changes that are taking place in contemporary religion? Is it possible to see in these changes some broader patterns that conform to — or deviate from — the theories we have at our disposal for understanding modern religion?

There are theories — theories of religious and cultural evolution — that are relevant to these questions. Rather than depicting religious change as a simple decline in the importance of religion, these theories emphasize qualitative changes in the character of religion. They depict broad evolutionary stages in the development of modern religion and provide clues about the possible future of modern religion. These are, of course, speculative frameworks, based as much on stylized notions of intellectual currents as on factual evidence about either the past or the present. Nevertheless, they can be useful as a starting point for thinking about changes in contemporary religion from a broader perspective.

Evolutionary Patterns

One need not accept the strong statements that some theorists have made about cultural evolution (let alone "sociobiological" evolution) to find value in theories that have tried to organize what we know about historical development according to some broad evolutionary schema. The virtue of such theories is that they provide general frameworks that highlight certain aspects of the modern situation and permit comparisons to be made with earlier periods. In tracing broad processes of social and cultural change, these theories also offer some guidance in thinking about the possible direction of changes in the future.

It is, of course, necessary to recognize that such theories depend to a great extent on the kinds of values and presuppositions built into the larger cultural environment from which they emerge. One finds, for example, that the present period is often portrayed as an infinitely better arrangement than anything that has been experienced previously — or, in other cases, that the present period is fraught with deep crises that pose a dire threat to modern society but that new ideas are just now being formulated that will save humanity from sure destruction. We must recognize value judgments of this kind for what they are rather than confusing them with the historical record itself. In providing an Archimedes point from which to view the major contours of an entire cultural epoch, however, such theories can be enormously useful.

Many efforts have been made to describe modern religion in broad evolutionary terms. Indeed, the sociological study of religion came into being largely as a result of such efforts. Marx, Weber, and Durkheim all formulated implicit theories of the direction of modern religious evolution, not to mention the more explicit scenarios advanced by Comte, Tonnies, Spencer, and others. A more recent generation of social theorists — including Talcott Parsons, Marion Levy, Jr., E. 0. Wilson, Peter Berger, Thomas Luckmann, and Bryan Wilson, to name a few — also contributed to this literature. Rather than trying to bring this entire theoretical armamentarium to bear on the question of religious change, however, I will content myself with a brief foray into the ideas of three representative theorists whose work on modern religion is unrivaled in subtlety of argument and appreciation for the nuances of cultural analysis: Robert N. Bellah, Jurgen Habermas, and Niklas Luhmann. It will be helpful to have a brief overview of each theorist’s claims in mind before trying to consider contemporary religion in relation to these claims.

Robert N. Bellah’s theory of religious change was published as a brief article in the American Sociological Review in 1964. Although Bellah has elaborated on the initial framework (even to the point of repudiating some of it) in more recent lectures and publications, the original outline has remained a central and provocative feature of most theoretical treatments of modern religion. Cast in a broad evolutionary framework, Bellah’s outline distinguished five ideal-typical patterns of religion, each of which is a distinct stage or "relatively stable crystallization" in the development of religion from early societies to the present. Bellah labeled the five stages primitive, archaic, historic, early modern, and modern. The progression from each stage to the next involves a process "of increasing differentiation and complexity of organization that endows the organism, social system, or whatever the unit in question may be with greater capacity to adapt to its environment than . . . its less complex ancestors."¹

Archaic religions, in BeIlah’s view, were more internally complex and differentiated than primitive religions in their view of the world and the kinds of religious action and organization they embodied; historic religions were more complex, in turn, than archaic religions, and so on. For our purposes, however, only the modern stage requires consideration, since Bellah traces its origins well back into the nineteenth century and argues for its prevalence, even at the popular level, in the contemporary United States.

Modern religion, Bellah asserts, is distinguished from all previous varieties by a "collapse of the dualism that was so crucial to all the historic religions." Instead of there being a sharp distinction between "this world" and some "other world" — between the natural and supernatural — modern religion tends to mix the sacred with the profane. Religious claims are grounded more in considerations of the human condition than in arguments about supernatural revelation. The "divine," in this sense, becomes more approachable, more imminent.

On the surface, this collapse of the supernatural into the natural would appear to represent a return to the more undifferentiated type of worldview evident in primitive or archaic religions. However, Bellah claims it actually consists of an increase in symbolic differentiation. Whereas reality was once perceived in dualistic terms, it now takes on a "multiplex" character. Religious symbols stand not so much in contradistinction to the secular world but as signals of deeper meaning. They point toward a form of transcendence that nevertheless penetrates everyday life.

According to Bellah, modern religious symbols are also held in a uniquely self-conscious manner. Rather than "looking through" symbols to the truths they convey, we "look at" symbols with greater appreciation of their capacity to shape reality. In Bellah’s view, this capacity to differentiate symbols from the truths they convey is a prime example of the more complex, more highly differentiated character of modern religion. There is, in a sense, a heightened capacity to work with symbols, to manipulate them to our advantage, to subject sacred symbols to textual criticism and yet to recognize them as truth, to distinguish myth from literal fact and at the same time to gain meaning from both. In other essays Bellah elaborates these points, both descriptively and normatively, by suggesting that modern culture develops an attitude of "symbolic realism" toward religion that recognizes the humanly constructed nature of religious symbolism and affirms the importance of such symbolism as a source of ultimate meaning and personal integration.

It is no accident that Bellah perceives a close connection between religious symbolism and personal integration. The "self" occupies a role in modern culture that is not only prominent but problematic. Gradually becoming emancipated from all sustaining collectivities, it is free to determine its own destiny — even choose its own identity. But it is also shaped by a myriad of experiences that in a complex society render it fragile and in need of integration. In religious matters the frightening responsibility that grew out of the Reformation and, in Weber’s words, left the believer face to face with God has been greatly expanded. Now the individual must accept responsibility not only for the duties that God prescribes but also for the very choice of gods to worship. Creeds must not only be lived up to but also interpreted and selectively combined, modified, and personalized in a way that the individual finds meaningful. And because the self tends to be multidimensional, if not transformational, the process of choosing and adapting religious symbols may acquire the quality of a lifelong "journey" rather than a once-for-all decision, let alone an ascribed status.

Religious action and perforce religious organization, according to Bellah, become more open and flexible in the modern era. No longer can churches and other religious organizations maintain a monopoly over the individual’s relation to the ultimate. Membership standards become looser as emphasis shifts away from uniform codes of belief and moral conduct toward individual choice and freedom of conscience. At the same time, the definition of what it means to be "religious" or to have ultimate meaning expands, becoming in the process more a product of individual interpretation. As a result, ethical conduct in the secular world replaces narrow definitions of salvation, and a larger number of people pursue their salvation through specialized, short-term commitments outside of the church entirely. Bonhoeffer’s "religionless Christianity" provides a theoretical image of the potential outcome. And at the organizational level Bellah foresees the emergence of "an increasingly fluid type of organization in which many special purpose subgroups form and disband."

The process by which a thoroughly modernized form of religion makes headway is, in Bellah’s view, not accomplished strictly at the expense of more traditional variants of religious expression. Fundamentalist religions and authoritarian religious cults that seek to impose uniform orthodox beliefs and standards of membership on believers will continue and perhaps even experience periodic revivals as reactions to the ambiguities inherent in modern society. There will also be people for whom religious symbols become so thoroughly unsatisfactory that a purely secular worldview — one that denies the possibility of ultimate meaning — is sought as a replacement. In the long term, however, Bellah is doubtful that either the fundamentalist or the purely secular mode of accommodation to modernity is as viable as an open, multiplex version of the sacred.

Bellah’s description of modern religion is grounded in considerations about long-term social evolution, but these considerations remain largely implicit in his work. Since the publication of the original essay — which compressed those theories into single sentences — Belllah has written widely on modern religion but has never spelled out in sufficient detail the dynamics of the underlying evolutionary model. Indeed, he has at times appeared to abandon even the premises of progress and linear development on which evolutionary theories are based. It is, therefore, of some interest to turn to a second theorist who has also taken up the idea of religious evolution: Jurgen Habermas.²

Trained as a German social philosopher with intellectual roots in the Marxist tradition and in the Frankfurt school of critical theory, Habermas departs from Bellah in significant ways in his depiction of social evolution. For instance, Habermas pays more explicit attention to economic development and to the state, credits the social sciences with a more prominent role in cultural evolution, and stresses secular procedures as elements of legitimation rather than emphasizing sacred or religious values. Yet Habermas, like Bellah, is deeply indebted to Weber, Heidegger, and Parsons; he ultimately rejects the Marxist emphasis on historical materialism; and he seeks an evolutionary framework that provides not only a description but also a normative guide for the discussion of modern culture. Indeed, Habermas’s treatment of religion draws directly from Bellah, contributing to it mainly by recasting Bellah’s efforts into a somewhat more explicitly elaborated view of cultural evolution.

Habermas distinguishes four stages of cultural evolution: neolithic, archaic, developed, and modern. The first three correspond roughly to Bellah’s primitive, archaic, and historic stages; the last subsumes Bellah’s early modern and modern stages. In principle, Habermas distinguishes his stages from one another on the basis of different principles of institutional organization, different levels of productive capacity, and different capacities for societal adaptation to complex circumstances. In practice, the stages are distinguished mainly, as are Bellah’s, by ever-increasing levels of cultural complexity and differentiation.

Neolithic cultures manifest relatively low levels of cultural differentiation. Motives and behavioral consequences remain undifferentiated, as do actions and worldviews, human and divine events, natural and social phenomena, and tradition and myth. Archaic cultures make greater degrees of differentiation possible among all of these categories, thus permitting greater clarity about the linearity of history, greater opportunities for calculated action with respect to the control of nature, and expanded opportunities for the development of rational law and the state. Developed cultures contribute additional layers of differentiation, replacing myth and tradition with unified cosmologies and higher religions, articulating well-codified moral precepts, and positing universalistic principles as modes of legal and political legitimation.

Modern culture, which coincides in Habermas’s view with the period since the Reformation, is typified chiefly by an erosion of confidence in the validity of higher-order principles. Reason, as a process of systematic reflection, replaces absolute axiomatic laws as the basis on which cultural meaning and coherence rests. As a result, faith is no longer taken for granted but becomes objectified as a focus for the application of reason. With reason also comes a greater degree of self-consciousness about the procedures used to arrive at and test the validity of statements concerning religion, morality, and nature.

Habermas is here alluding to more than simply the proverbial "warfare" between reason and faith that scholars since the Enlightenment have tended to emphasize. He is also asserting that faith itself increasingly becomes subject to tests of reason, logic, and even empirical procedure. As Bellah also observes, this process leads to greater self-awareness about "statements," as distinct from the meanings conveyed by these statements. Whereas Bellah sees a somewhat ready solution to the inherent tension between reason and faith, each maintaining its own validity within separate realms, Habermas sees greater evidence of tension between the two. His view suggests considerable, but by no means total, retreat on the part of religion against the onslaught of rationality.

But religion does have a continuing role to play, according to Habermas (here in agreement with Bellah), in providing personal integration, unity, subjective meaning, and a unified self-identity. Like Peter Berger, whom he quotes, Habermas perceives a requirement for some form of symbolic universe — a sheltering canopy — to integrate the various systems of action in which an individual engages. Religious worldviews played a prominent role in fulfilling this requirement in "developed" cultures. Greater differentiation of human action from nature allowed a greater sense of individual identity to develop; a unified picture of the forces governing the universe contributed to the internalization of a coherent set of moral principles which in turn facilitated greater unity of the self, and, as Weber observed, religious explanations for the misfortunes of nature helped individuals and collectivities to function more effectively in the face of risk, grief, and ultimate doubt. In the modern period religion continues to perform many of these traditional functions.

However, the role of religion in the modern period has, in Habermas’s view, become progressively restricted. With modernization has come greater control over the exigencies of nature, thus diminishing the need for religious explanations for these exigencies and increasing the importance of specialized knowledge devoted strictly to the empirical understanding of nature. Religion has accordingly come to focus less on metaphysical assertions about the world and more on exclusively subjective concerns about individual meaning and integration. In short, religion has been pushed increasingly into the realm of what Habermas calls "civil privatism."

The "privatization" of religion is an emphasis that Habermas shares, in some respects, with Bellah and indeed with many other observers of contemporary religion. But Habermas remains too concerned with social processes to perceive much value in a religion that merely gives interior solace to the individual. Like Bellah, he recognizes that much of the traditional importance of religion lies in its capacity to instill a sense of community and perform socially integrative functions. Bellah maintains that this function continues to be performed largely through the application of religious values — not through the institutionalized church but through the culturally differentiated set of religious symbols he calls the civil religion. For Habermas, however, the notion of civil religion seems to be an unworkable throwback to a less modernized set of values.

Perhaps because of his roots in the more thoroughly secularized setting of western Europe, Habermas sees little continuing significance in the role of traditional religious values as a means of social legitimation and integration. The social sciences, in his view, have seriously undermined the plausibility of such values, rendering them relativistic and secondary to a kind of technical reason devoted to the solution of social problems through the application of technical knowledge. Habermas admits that the social sciences have been relatively ineffective in actually resolving social problems, let alone providing substitutes for religious answers about death, grief, illness, and tragedy. Nevertheless, he maintains that by creating awareness of the social constructedness of religious symbols, the social sciences have made it impossible to rely on absolute religious claims for social integration. Instead, social integration will, for better or worse, depend chiefly on the rational-legal procedures that have been developed in modern societies for the conduct of business and government and for the resolution of disputes.

Still, Habermas does point out one function that religious symbols may be able to perform — the function of facilitating communication. This is at first glance an anomalous position to find Habermas taking, because he generally appears to be skeptical of religion, treating it as a form of ideology that systematically distorts communication by virtue of the fact that religious people generally seem to be unable to abandon their own suppositions long enough to truly consider the interests and values of others. Yet in a few scattered passages that remain largely undeveloped, he makes reference to the possibility that some understandings of modern religion seem to promote rather than inhibit communication. Noting the theological work of Metz and Pannenberg in particular, he argues that a new concept of God is evident in this work that recognizes the socially constructed character of all conceptions of the divine but nonetheless asserts the utility of such symbols because of their emphasis on community and reconciliation.

Like Bellah, Habermas perceives a serious degree of erosion associated with institutionalized religion — so much so in fact that he seldom bothers to discuss organized religion. In his evolutionary perspective, the growing differentiation of modern culture has placed religion in competition with reason, the natural sciences, and most recently the social sciences, all of which have taken over many of the topics on which religion traditionally spoke with authority. In addition, modern religion has become more internally differentiated such that religious knowledge has become more distinct from religious experience and meaning, religious techniques have become more distinct from religious values, and religious statements have become objectified over against the largely privatized meanings and interpretations they are given. Finally, growing awareness about the situational relativity of religious symbols, while reducing their absolute plausibility, has perhaps enhanced their role as objects and facilitators of communication about social values.

A third perspective on the evolution of modern religion has been provided by Niklas Luhmann, another leading German sociologist, often in critical debates with Habermas. A professor of sociology at the University of Bielefeld in West Germany, Luhmann has emerged as one of the major theorists in contemporary sociology. Known until recently only through his debates with Habermas, he has begun to attract considerable attention among American scholars in his own right. Like the classical theorists, he treats religion as a basic feature of modern society.

Luhmann is a systems theorist, somewhat in the same genre as Talcott Parsons, who understands social evolution chiefly in terms of social differentiation. Thus he contends that modern societies differ from traditional societies mainly in terms of having a greater number of clearly differentiated social spheres or subsystems. The basic institutional pattern of modern societies was laid down, in his view, between the sixteenth and eighteenth centuries with the emergence of a relatively autonomous political system which was accompanied by increasingly autonomous systems in other realms as well, such as science, law, education, and art. In the process, religion gradually came to be viewed as just one sphere among many. Like Habermas, Luhmann believes that modern religion has largely lost its capacity to legitimate and unify the society as a whole. Religion is not likely to die out, he says; it may even flourish among reactionary groups critical of political and economic processes. But it must function increasingly without significant connections to other spheres of social action.

This, at least, was Luhmann’s position until recently — a fairly pessimistic view of modern religion that did not differ substantially from that of many social theorists in the nineteenth century. But Luhmann’s broader framework has been in flux, and so have his views on religion. Like Habermas, he has immersed himself in the growing literature on language, discourse, and communication, coming increasingly to conceive of society itself as a vast system of communicative action, and this perspective has given him a number of novel ideas about the nature of religion.

At the 1984 meetings of the American Sociological Association Luhmann presented a paper entitled "Society, Meaning, Religion — Based on Self-Reference" in which he outlined the core of his ideas about modern religion. Building on his previous work, he continued to describe modern society as a product of evolutionary development, but he also suggested that a fundamental characteristic of modern society is its inevitable and enduring confrontation with paradox At one level, the basic paradox confronting modern society can be seen in the fact that there must be closure for communication to occur, yet there must also be openness in order to cope with the high degree of complexity and change in modern society. At another level, paradox is seen in the fact that any statement about the world implies that other possibilities were also available, thus rendering even definitive statements contingent. In short, Luhmann claims that we live in a world where, paradoxically, contingency is necessary. It is from this simple recognition of paradox that Luhmann develops his views of the functions of religion.

"The plentitude and void of a paradoxical world is the ultimate reality of religion," he writes. "The meaning of meaning is both richness of references and tautological circularity." Since modern society is both a closed system and yet inevitably constrained to remain open to unanticipated contingencies, it must cope with the paradoxical nature of its situation. Luhmann defines religion as the set of forms that society develops to "deparadoxize the world." Religious forms "absorb" paradox, resolving it by revealing that seeming opposites are really one and the same.

The essential insight here is seen most vividly in contrast with Weber’s description of the function of traditional religion. Weber argues that traditional religion serves to "make sense" of the universe by demonstrating that everything can be related to everything else, can be "unified" into a single whole. Thus good and evil, suffering and joy, life and death are all reconciled in a single overarching framework. But Luhmann’s argument is that even this framework cannot be accepted as final. It is a system that necessarily recognizes possibilities for meaning beyond itself. This is the paradoxical tension with which modern religion must always deal.

Any concrete religious symbol system or organization is thus inevitably precarious. It exists as a form that resolves ambiguity, and yet this form must always grant the reality of ambiguity. Religious systems therefore undergo nearly continuous transformations, both altering their social contexts and being altered by them. They are endangered by their own successes, resolving paradox to their long-term detriment and necessitating new conceptions of paradox itself. Luhmann thus conceives of religion as undergoing an evolutionary process but disagrees with Bellah and Habermas about whether this process can be reduced to a model of discrete stages.

Nevertheless, some significant landmarks in the evolution of modern religion can be identified. Since the paradoxes that religion deals with pertain, in Luhmann’s view, to the nature of communication, these landmarks have mostly to do with developments in the nature of communication. One was the invention of the alphabet. This invention led to the capacity to produce written religious texts which then, along with the gods and their utterances, became objects of religious reflection. No longer could priests and prophets simply offer their own renditions of the supernatural; they were now constrained by what the scriptures said. Luhmann suggests, not entirely facetiously, that the development of Christianity itself may be seen as a desperate attempt by religious specialists to survive the invention of the written word.

This of course happened centuries ago. It continues to affect the course of modern religion insofar as scripture and "the word" have remained to a high degree objects of self-conscious reflection in the Christian religion. As Luhmann notes, the New Testament canon itself seems to reflect a pattern of faith that is more closely circumscribed by religious texts than is the Old Testament. But none of this gets at the distinctive developments that have characterized religion in the twentieth century.

Luhmann’s depiction of contemporary religion includes many of the themes already noted in Bellah and Habermas — privatization, differentiation of beliefs from organizations, special purpose organizations, and a declining capacity to legitimate political and economic action. He also specifies Bellah’s notion of greater differentiation between the supernatural and natural realms by observing again the importance of communication: God remains an object with whom we can communicate; in contrast, we do not normally talk with nature. Most of these are, however, developments that Luhmann traces back as far as the origins of Christianity or even to the Greek city-states.

For the contemporary period, the main problem facing religion is, in Luhmann’s judgment, that of communicating with a god who has withdrawn into silence. Can we, given our own increasing awareness of the nature of symbols, realize the functions of our beliefs and yet hold firmly to them? Can paradox be resolved if we are aware that the role of our beliefs is to resolve paradox? His answer is that we probably cannot. Yet he also admits that this degree of self-reflectivity is probably limited mainly to the better educated segments of the population.

There is, however, a curious paradox within Luhmann’s own discussion of the paradoxical nature of religion. Recognizing both the value of the process of resolving paradox and the manner in whjch communication functions, we may be able to reconsecrate the churches for this very purpose. That is, the function that the churches may be best able to fulfill, even in modern society, is that of promoting communication with God. Within the insulated sphere that the churches provide, it may be possible to evoke a form of reality in which communication with an otherwise invisible deity makes sense. It may not be possible for such a reality to be constructed for everyone or on a broad scale, for communication with an invisible deity clearly runs counter to the norms of modern culture. Yet this contrast may be important in itself. As Luhmann argues, "churches seem to cultivate countermores, depending for their success on being different. Religion may have become counteradaptive, and this may be the very reason for its survival and for its recurrent revival as well. The church itself, by now, may have become a carnival, i.e. the reversal of normal order."

This vision of "church as carnival" contrasts sharply with the ideas of either Bellah or Habermas. Starting from the same premises concerning social complexity and cultural differentiation, Luhmann nevertheless is led along a different route by his emphasis on the paradoxical nature of social relations. Although silent in this perspective, God has by no means collapsed into a monistic worldview or been replaced by secular-rational norms of discourse. Religion fails to provide a clear sense of broader societal integration in all three of the models, but in Luhmann’s the emphasis on communication tends to preserve the idea of religious community, if only as counterculture. He maintains that the churches have a lasting, albeit restricted, social function to fulfill.

These are of course broad theoretical visions not to be used, as Bellah reminds us, "as a procrustean bed into which the facts of history are to be forced but a theoretical construction against which historical facts may be illuminated." How well, then, do these theoretical constructions illuminate the historical facts of American religion in recent decades?

From Theory to Evidence

In varying ways Bellah, Habermas, and Luhmann all stress the importance of greater self-awareness with respect to symbolism — or to culture, we might say — as a feature of modern religion. There is, in fact, much evidence that religious culture has become increasingly objectified for self-reflection at least at some levels. Theological reflection has, for example, converged to a remarkable degree with some aspects of the social sciences in its concern for the symbolically constructed character of reality. Munich professor of theology Wolthart Pannenberg, sounding strikingly like a social scientist, has observed that "it is only by symbols and symbolic language that the larger community to which we belong is present in our experiences and activities."³ And he goes on to argue that the church not only uses symbols but is itself symbolic. In a similar manner, Yale theologian George A. Lindbeck has asserted that "a religion can be viewed as a kind of cultural and/or linguistic framework or medium that shapes the entirety of life and thought."⁴ Both writers develop their arguments, as Bellah might have predicted, not from metaphysical first principles but from anthropological considerations of the nature of symbolism. In applications of discourse analysis and deconstructionism, other theologians have taken the investigation of religious symbols even further.⁵

At the popular level, evidence of greater self-consciousness about the nature of religious symbolism is naturally less apparent. To broach the subject of mythology or textual criticism remains a mark of extreme heresy among the third of the public who believes that the Bible is not only divinely inspired but to be taken literally. Yet there are also indications that a substantial number of believers have achieved some degree of mental differentiation between their faith and the symbols with which it is expressed. For example, among those who believe the Bible to be divinely inspired, only half regard it as absolutely free of errors. Or for another example, a study of Lutheran church members showed that only one member in three felt it possible to prove the existence of God, and of these only about half felt this could be done from evidence in the Bible.⁶

Studies of new religious movements are also replete with evidence of the self-conscious application and manipulation of symbols. Because many of these movements grew from small groups in which symbols were either invented or synthesized from other sources, members tend to be keenly aware of the power of symbols. Being in a position to remake their own rituals and ideologies, and seeing the immediate effect of these symbols on group life, they quickly develop a heightened sense of "symbolic consciousness." Movements that seek to block out the effects of axiomatic constructions of reality through meditation, drugs, or religious experiences also sharpen their members’ sensitivities to the nature of symbolism.

Bellah’s argument that greater self-consciousness about religious symbolism tends to be accompanied by a greater emphasis on personal interpretation and a decline in tacit acceptance of official creeds is also supported by a variety of evidence. In the Lutheran study mentioned previously, only half of the respondents felt that God has given clear, detailed rules for living that apply to everyone; most of the remainder felt that individuals have to figure out how to apply God’s rules to their own situations. The study also included an effort to determine how much agreement exists between individual members’ views and the official theological positions of the church, first by interviewing theologians to determine what the official positions were, then by surveying pastors to see if these positions were taught, and finally by surveying members about their own beliefs. The core theological tenets of the church, as described by its theologians, consisted chiefly of three simple propositions: that Christ was fully God and fully man, that Christ was crucified to forgive our sins, and that men and women are sinners whom God loves and is giving new life. These tenets were uniformly accepted by the clergy and, partly because of the prescribed schedule of sermon topics, emphasized from the pulpit. Yet the laity survey found that only one member in three affirmed all three of these propositions. On other teachings, such as the church’s views of baptism and communion, agreement was equally low. Given these tendencies, it may not be surprising that denominational boundaries seem to be weakening.

The evidence is less clear with respect to Bellah’s claim that modern religion is principally characterized by a collapse of the dualistic worldview that distinguishes God from man, the supernatural from the natural, this world from a world beyond. As Bellah himself observes, more than ninety percent of the American population affirms some belief in the existence of God. Such affirmation scarcely answers the question of whether there has been, as he puts it, "a massive reinterpretation" of the nature of God. But more refined questions suggest that a sizable number of Americans still express their faith in dualistic terms. For instance, nine persons in ten believe Jesus Christ actually lived, seven in ten believe he was truly God, and six in ten think one must believe in the divinity of Christ to be a Christian. The results of studies documenting consistently high levels of belief in life after death, heaven, and Christ’s presence in heaven also point to the survival of a strong element of religious dualism in American culture. Indeed, the persistence of these beliefs seems to be one of the more stable elements of American religious culture, in contrast with the serious restructuring that has taken place in many other beliefs and practices.

But if dualism continues, evidence also suggests that believers have, in a sense, "subjectivized" the divine rather than continuing to conceive of God as a metaphysical, transcendent, or omnipotent being. A study conducted in the San Francisco Bay Area in the 1970s found, among persons who said they definitely believed in God, that eight out often believed in God’s influence on their personal lives, but only about half felt God influenced social events.⁷ In the Lutheran study, only three in ten believed that God "shapes events directly through nations and social affairs."

A good deal of speculation — and some research — has also suggested that God is relevant to contemporary Americans mainly because the sense of God’s presence is subjectively comforting — that is, religion solves personal problems rather than addressing broader questions. This is perhaps especially true among evangelicals: historically evangelicals have tended to emphasize God’s sovereignty in all things, but more recently much of their literature focuses on emotional and psychological concerns.⁸

While it may be, then, that a high degree of supernaturalism remains in American religion as a formal tenet, the operational relevance of the supernatural may have largely collapsed into the interior concerns of the self. This conclusion tends to be supported by the high degree of interest that surveys have documented in questions of personal meaning and purpose and in the number of quasi-religious self-help movements that have developed since the 1960s.

At the level of religious organization there is also much to support Bellah’s contention that religious expression has become increasingly differentiated from traditional religious institutions. High rates of denominational switching and interdenominational marriage, reduced levels of denominational identity and cross-denominational tensions, and pervasive contacts across denominational lines all point toward a declining monopoly of specific religious traditions over the securement of religious convictions. Evidence on the numbers of individuals who consider themselves religious or who hold certain tenets of faith and yet do not belong to religious organizations or attend religious services regularly point in a similar direction. The idea of "special purpose groups" also gains confirmation from the evidence on rising numbers of such groups and the extent of participation in these groups.

At the most general levels of societal integration and legitimation, the evidence, while subject to various interpretations, suggests the continuing relevance of religiously inspired ethical concerns but also reveals the diminishing weight of religious arguments as such, relative to the weight these arguments carry at the individual level. In many respects, the most obvious religious development with respect to societal integration has been the rise of the New Christian Right. On the surface this movement — and the countermovements it has provoked — suggests a continuing tendency for religious values to find their way into the public domain as a part of debates over societal goals. Liberals and conservatives alike have resorted to religious arguments in defense of claims about public morality and the role of the state in defending public morality. Yet the very dissension that has been produced by the religious right points up the difficulties of gaining any kind of broad consensus on the basis of traditional religious values. Where tacit agreement has been achieved, it has been achieved mostly with respect to the underlying rational-legal procedures to which political action must pay heed. Moreover, secular myths having to do with individual freedom, material success, and perhaps especially the wonders of technology may be an even more powerful source of societal legitimation than, traditional religious arguments by virtue of being grounded in many assumptions that are virtually axiomatic in the culture.

Habermas’s assertions overlap considerably with those of Bellah, especially with regard to the privatization of religious expression, the decline of religious orthodoxy, and the erosion of religious institutions. Thus his claims illuminate many of the same bits of evidence as those just considered. His general emphasis, however, tends to suggest that rationality, natural science, and the social sciences have an even more pervasive effect on religion. The evidence suggests that all of these influences have indeed exercised a negative effect on traditional religious beliefs and practices. Not only do scientists — and especially social scientists — demonstrate radically low levels of religious commitment, but scientific and social scientific meaning systems appear to operate as functional alternatives to traditional theistic ideas for a number of people, and technical rationality plays an increasingly important legitimating function in the wider society. Highly publicized reactions to science and social science on the part of religious conservatives, as evidenced by lawsuits concerning the teaching of evolution in public schools and court cases challenging the influence of "secular humanism" on school textbooks, suggest that Habermas’s forces of "secular rationality" have by no means carried the day. And yet, the very grounds on which these controversies have been fought — arguing for the "scientific" basis of creationism, making use of the "rational-legal" procedures supplied by the modern court system, and drawing on social scientists for "expert testimony" — all point to the considerable degree to which even religious conservatives have accommodated to the norms of secular rationality.

At the same time, a clear case cannot be made in support of Habermas’s claim that the sciences have so reduced the physical and social contingencies of modern life as to make religious worldviews largely irrelevant. To the contrary, the sciences seem only to have contributed to a greater degree of sensitivity about the contingencies of life. Indeed, Luhmann’s suggestion that modern society is inevitably confronted with the paradoxes of its own contingency seems to be more applicable. Thus the concerns that continue to inspire deep religious discussion, such as the prospect of nuclear annihilation, the rights of the unborn, euthanasia, world hunger, Third World dependence, and so on, are clearly evidence of lingering contingencies in a technologized world.

At a more theological or philosophical level there is, however, a very significant development that Habermas’s discussion helps to illuminate: the manner in which religion itself has been redefined in the face of advances in the realm of natural reason. Modern definitions of religion, as Bellah suggests, have come to focus increasingly on symbolism and meaning. Writers such as Clifford Geertz, Peter Berger, and Bellah himself conceive of religion as a special kind of symbol system that evokes a sense of ultimate, transcendent, encompassing meaning. Drawing on the social sciences, this conception actually manages to save religion from the onslaught of post-Enlightenment positivism. It accomplishes this feat by positing religion as a type of symbolism concerned with the meaning of the whole of life. The meanings of anything less — of selected aspects of the world — can be identified by the contexts or frameworks in which those aspects are located. But the meaning of the whole lies beyond any specific context. As Wittgenstein observed, "the meaning of the world lies outside of the world."⁹ Thus the "world of facts" with which the empirical sciences deal must be seen ultimately in another context — a context given meaning by religious symbols — which is beyond the scope of the empirical sciences.

Thus it is not irrelevant to Habermas’s argument that modern religion tends to be defined the way it is. Not only has there been a greater degree of differentiation between symbols and truth in modern religion but there has also been an increasing degree of differentiation among kinds of symbols. As a result religious symbols have been put beyond the reach of rational and empirical criticism to the extent that they have been identified with a different type of reality construction. It should be noted that Habermas himself has contributed to this development by identifying different types of validity claims that may be embodied in ordinary discourse — some of which can be subjected to empirical criticism, others of which remain matters of nonempirical metaphysical or philosophical reflection. Some of the difference evident in American society between religious liberals and religious conservatives can be understandable in these terms, especially if some part of the more liberal population can be assumed to have differentiated in their religious discourse between symbolism oriented toward holistic meanings and symbolism subject to empirical criticism.

Luhmann’s discussion, lastly, raises distinct questions mostly in relation to his ideas about communication with God and the role of the church as carnival. Although he recognizes the precariousness of efforts to communicate with an invisible, silent God, the abiding paradox of living in a world that recognizes its own contingency forces him to concede that such communication will likely continue as a feature of modern life. Much evidence in fact suggests that, despite considerable erosion of religious practices in other areas, attempts to communicate with the divine remain strikingly prominent. Prayer in particular seems to have remained a strong feature of contemporary life in comparison with other kinds of religious behavior. For example, one of the Gallup surveys mentioned earlier showed that 60 percent of the American public personally considered prayer to be very important and another 22 percent regarded it as fairly important; by comparison, only 39 percent thought that reading the Bible is very important, 38 percent thought that attending religious services is very important, and 28 percent thought that being part of a close religious fellowship group is very important. Other surveys have documented high levels of interest and involvement in prayer, a high degree of belief in the efficacy of prayer, and a strong tendency to regard prayer as actual communication with God.

Apart from prayer, evidence also suggests that many people continue to value highly their sense of a relationship with God, and many feel they are close to God. The Gallup survey showed that "growing into a deeper relationship with God" was considered very important by 56 percent of the public and fairly important by an additional 26 percent. In another Gallup survey, 64 percent of the public indicated that their relationship to God was very important to their own sense of self-worth, and nine out of ten expressed satisfaction with this relationship.

Luhmann observes with some interest that modern religion seems to depict God chiefly as an all-loving being, thus reducing much of the motivation for salvation that has historically been associated with Christianity (viz., to avoid damnation). This depiction may in fact serve a positive role in sustaining the plausibility of communication with an invisible God in the modern era: communication may be easier to sustain when God is envisioned not as distant judge but as a loving friend, an intimate "God within." At any rate, there are clear indications that contemporary imagery envisions God in such terms. Eight of every nine persons say they feel that God loves them, 80 percent say they feel close to God, and only 16 percent say they have ever felt afraid of God.10 In the Lutheran study cited earlier, nine in ten said God loved them and was giving them new life; only a quarter felt they were sinners under the wrath and judgment of God. And evidence from a 1984 National Opinion Research Center survey indicates that, although images of God as judge and king persist, substantial numbers of Americans lean toward more intimate images such as lover and spouse.¹⁰

Luhmann’s idea of the church as a kind of counterculture devoted to maintaining the plausibility of communication with God also appears consistent with a variety of evidence. Although this is by no means a new role for the church to fulfill, it is a role that the church seems to have carried on with surprising success as the culture has become increasingly secularized. To be sure, the religionless Christianity of humanistic ethicalism that Bellah speaks of is evident in many mainline churches. But the importance of the religious community gathered for worship and fellowship with God is also strikingly evident. Protestants and Catholics alike have shown increasing interest in liturgy as the heart of such communal activity. Pannenberg has suggested that "the rediscovery of the Eucharist may prove to be the most important event in Christian spirituality of our time, of more revolutionary importance than even the liturgical renewal of our time."¹² As the sense of guilt and sin that became prominent in the teachings of the Protestant Reformers erodes, he suggests, the church will increasingly find its reason for existing is to serve as a symbol of wholeness in a broken world. This is the purpose of the eucharist: to dramatize communion with God and to evoke the healing presence of God in the world. Moreover, in an argument with which both Habermas and Luhmann have shown familiarity, he suggests that the eucharist can be interpreted in distinctly modern terms as a symbol that dramatizes freedom by casting ossified structures in doubt and that enhances adaptiveness and communication by emphasizing openness and provisionality: "the human predicament of social life is not ultimately realized in the present political order of society, but is celebrated in the worship of the church, if only in the form of the symbolic presence of the kingdom to come."

The typical congregant may well not participate in "the worship of the church" with the sense of sophistication that Pannenberg suggests. Yet in some form the church — whether liberal or conservative — does continue to attract participation largely as a place in which to experience the closeness of God and the communion of fellow worshipers. Among the gratifications from church mentioned most often in a national survey of regular church attenders, for example, were feeling close to God (77 percent), the experience of worshiping God (60 percent), and a sense of companionship or fellowship (54 percent).

Conclusion

If the foregoing is any indication, American religion demonstrates many of the characteristics that theorists have identified with modern culture. Many religious beliefs and practices remain much in evidence, contrary to simpler predictions that have envisioned a simple decline in religious vitality. These beliefs and practices may have retained their vitality through accommodations to the contemporary cultural situation. In becoming more oriented to the self, in paying more explicit attention to symbolism, in developing a more flexible organizational style, and in nurturing specialized worship experiences, American religion has become more complex, more internally differentiated, and thus more adaptable to a complex, differentiated society.

For several reasons this is a fairly speculative conclusion, however. In the first place it has to be defended largely without comparable evidence from other times or places. Some of the characteristics of American religion that bear directly on the theories of Bellah, Habermas, and Luhmann can be shown to have intensified even in recent decades. But it can legitimately be asked whether many of these characteristics are unique to the modern period, whether they are intensifying, or whether they might not also have characterized Western religion a century or even a millennium ago.

Another difficulty is that evolutionary theories tend to be cast in such broad terms that essentially ambiguous data can be readily manipulated to support them. It often seems less than clear what counts as evidence of increasing differentiation and what might be regarded as counterevidence. Habermas has argued specifically against trying to make such connections with concrete historical examples, suggesting that evolutionary theories are better viewed as normative guides toward the future than as testable theories. Thus it may be that American religion only seems to have accommodated itself to modernity because of a selective interpretation of the facts.

This criticism, however, should not overshadow the positive role that evolutionary perspectives can play. If we admit that their purpose is not to provide us with testable hypotheses, then we can make use of them, as suggested earlier, to illuminate what might otherwise seem to be disparate or insignificant developments. We are led to think about the possibility that what appear to be signs of decay in contemporary religion may actually have beneficial consequences for its survival over a longer period. For example, the decline of orthodoxy may beassociated with a rise in personalized religious interpretations that make religion more adaptable to changing circumstances. We are also led to think about the relations among certain developments and the significance of these developments in a broader context. If it is true that religion is becoming more highly differentiated, then it may be especially important to develop greater effectiveness in dealing with new distinctions, with new understandings of symbolism, and with new kinds of religious organizations.

The more serious limitation of existing evolutionary approaches to religion is that they fail to illuminate much about the relations between religion and the broader social environment. Bellah, Habermas, and Luhmann all relate religious evolution to the growth of complexity and subsystem differentiation in the larger society. But none of the three draws explicit connections between these two levels of development of the kind that would indicate how a particular form of religious differentiation might be related to a specific example of societal complexity. And since all three leave open possibilities for maladaptive reactions, it becomes exceedingly difficult to pin down what constitutes complexity and what the effects of complexity may be.

Beyond this general problem, Bellah, Habermas, and Luhmann seldom identify the mechanisms of cultural change. At times it appears that each theorist regards religious evolution, like other dimensions of cultural evolution, as resulting from its own internal dynamics. This seems to suggest that previous symbolic structures set the constraints and provide the opportunities for new cultural developments. Thus one is forced to look mainly at the internal logic of Christianity, the legacy of Reformation Protestantism, and theological debates of the eighteenth and nineteenth centuries to understand what has shaped the character of American religion. Implicit in this approach is the view that institutional differentiation has progressed to such a high degree that religion is no longer significantly affected by anything other than developments within the religious institution itself.

Habermas himself, though, has openly challenged this view in some of his more recent work. Although he believes that the general patterns of cultural evolution are internally determined, he observes that other aspects of social structure are likely to be particularly influential during the transition from one general phase to another. He specifically mentions protest groups and religious movements as examples of mechanisms that may play a critical role in times of transition.

In the American case, evolutionary theories are most deficient in interpreting religious characteristics in relation to elements of the broader social fabric. Most of the empirical characteristics that fit in one way or another with these theories do not apply uniformly to the entire population. Many of them pertain most clearly to the young and to the better educated — factors that suggest the growing prominence of these characteristics. But the theories provide little help in explaining why education is a factor, and why now. Nor do they cast into sharp relief, except in very general statements about adaptation and reaction, the kinds of conflict that have emerged between religious liberals and religious conservatives. The assertion that differentiation marches forward may be an accurate appraisal of long-term tendencies, but the apparent inevitability with which this process is portrayed fails to account for either the more specific tensions that develop in the short run or the precipitating events that engender these tensions. Greater differentiation between values and behavior or between the kingdom of God and institutional programs, for example, can scarcely be understood in any kind of adequate way apart from the specific societal pressures that reinforce these changes.

What we need to complement highly abstract evolutionary theories are concrete historical comparisons from which to gain a sense of the kind of environmental factors — as well as the kinds of internal institutional responses — that may result in religious restructuring. The recent period has been fraught with rapid changes in education, technology, and the character of the state. To understand how these developments have affected religion in the past, and how they may shape religion in the foreseeable future, we need comparisons drawn from concrete historical material.

END NOTES

1. All references and quotations are from a reprint of the original article that appears in Bellah’s Bryond Belief Essays on Religion in a Post-Traditional World (New York: Harper & Row, 1970), pp. 20-50.
2. The essays of Habermas that deal most specifically with religious evolution include Communication and the Evolution of Society (Boston: Beacon, 1979) and "History and Evolution," Telos 39 (1979): 5-44.
3. Pannenberg, Christian Spirituality (Philadelphia: Westminster Press, 1983), p. 35.
4. Lindbeck, The Nature of Doctrine: Religion and Theology in a Postliberal Age (Philadelphia: Westminster Press, 1984), p. 33.
5. For example, see Vincent Brummer, Theology and Philosophical Inquiry (Philadelphia: Westminster Press, 1982); and Mark C. Taylor, Erring: A Postmodern A-Theology (Chicago: University of Chicago Press, 1984).
6. See my article, "Religious Beliefs and Experiences: Basic Patterns," in Views from the Pews: Christian Beliefs and Attitudes, ed. Roger A. Johnson (Philadelphia: Fortress Press, 1983), pp. 10-32.
7. Thomas Piazza and Charles Y. Glock, "Images of God and Their Social Meanings,’ in The Religious Dimension, ed. Robert Wuthnow (New York: Academic Press, 1979), pp. 69-92.
8. See James Davison Hunter, American Evangelicalism (New Brunswick, N.J.: Rutgers University Press, 1983), pp. 91-99.
9. Ludwig Wittgenstein, Tractatus Logico-Philosophicus (Atlantic Highlands, N.J.: Humanities Press, 1974).
10. The figures cited in this section are from two studies conducted by the Gallup Organization: Jesus Christ in the Lives of Americans Tod4y (1983) and How Can Christian Liberals and Conservatives Be Brought Together? (1984). Some of the results of the first study have been published in George Gallup, Jr., Who Do Americans Say That I Am? (Philadelphia: Westminster Press, 1986).
11. Andrew M. Greeley, "Religious Imagery as a Predictor Variable in the General Social Survey," paper presented at a plenary session of the Society for the Scientific Study of Religion, Chicago, 1984.
12. Pannenberg, Christian Spirituality, p. 47.

Every community is awash in words. Religious communities are no exception. Sermons, prayers, singing, creedal recitations, and discussion groups make up the very being of such communities.

In recent years the flow of religious discourse has spilled into the public arena with increasing intensity. Religious broadcasts fill the airwaves and direct-mail solicitations fill our mailboxes. Bishops issue statements on social issues such as nuclear disarmament and economic justice. A pope stumps the country delivering homilies. Preachers become presidential candidates. And media specialists try to make sense of it all.

Social scientists have in recent decades developed a fairly standard way of studying the relations between religion and public affairs. Opinion polls are the method of choice, supplemented by occasional applications of content analysis, in-depth interviews, and discussion of broader social developments to provide context. As a result of this often valuable research, we have a good sense of the public’s tolerance for religious leaders making statements about various kinds of social issues. We also have some data about the issues clergy say they speak about. And we have many studies of the ways in which religious beliefs and attitudes toward social issues correlate — fundamentalism and bigotry, parochialism and conservatism, conservatism and views of the priesthood, moralism and attitudes toward abortion, religious preference and voter orientations, to name a few. We even have frequency counts of the kinds of themes that are expressed on religious television shows or in religious books.

But on religious discourse as discourse we have virtually nothing. It is as if our standard methods have trained us to think of religious communities (and not just religious communities) as silent worlds. People have religious beliefs, convictions, and sentiments. They harbor predispositions, orientations, and commitments. They behold religious symbols, and these symbols give meaning to their lives, help them construct reality, and provide them with security and a sense of belonging. But they do not speak.¹

Or if they do speak, our standard methods register only the surface features of their discourse. For instance, we may cull through the transcripts of religious broadcasts to see how many of them touched on abortion, school prayer, the Supreme Court, or politics in general. We may scan the titles of religious books to see how many fall into various preconceived categories: theology, family, self-improvement, sexual relations, and the like. Or we may ask church and synagogue members or clergy whether they have discussed topics such as personal crises, moral issues, politics, or the federal budget with fellow parishioners. But none of this gives us any indication of the ways in which religious discourse is actually put together.

It may, of course, require more than a leap of religious faith to argue that the actual composition of religious discourse is itself important. To someone trained in the social psychology of opinion research, discourse is likely to be relevant only as a means of tapping into the deeper attitudinal predispositions that supposedly govern behavior. Discourse is in this view ephemeral, unpredictable, superficial; only the underlying mindsets are meaningful. We want to discover how personalities are put together, not to invest time in the study of meaningless chatter.

 Discourse Rediscovered

There has for some time been a movement in the social sciences to bring discourse back into the picture. In addition to the contributions of small coteries of ethnomethodologists and conversation analysts who have always studied discourse, we now have the formidable (and often forbidding) legacy of Foucault’s decentered poststructuralism, Habermas’s borrowings from speech-act theory, Derrida’s language-focused deconstructionism, and a more scattered array of empirical investigations focusing on public discourse.

We need not become camp followers of the esoteric theoreticians to appreciate the importance of understanding religious discourse, however. Much of this discourse is highly codified in sacred traditions. Its practitioners gain competence in its use through long years of training and experience. Homiletics and hermeneutics are required features of most pastoral educations. How-to books abound — for preaching, leading discussions of religious texts, proselytizing. Even the sacred traditions themselves recognize the importance of the word, the kerygma, as the vehicle of creation, reconciliation, and community.

That much we could discern by immersing ourselves for any length of time within a religious tradition. Any competent practitioner of the faith could testify to the importance of discourse. But when religious discourse enters the public sphere — when it becomes public rhetoric — we confront another, perhaps equally compelling, reason for trying to understand it. Some of it seems to affront common sensibilities so deeply that we find it difficult even to focus on what is being said; in other instances this is less of a problem. For example, I usually spend some time having students look at direct-mail solicitations from religio-political organizations in a sociology of religion course I teach. Sometimes I also ask them to sample a few religious broadcasts on television or to watch a short video of fundamentalist dialogue in class.

Generally the reaction from my privileged, sophisticated, tolerant, upper-middle-class white juniors and seniors is repulsion. They find fundamentalist discourse so alien to what they are used to thinking that their processing capacity breaks down. Why?

Put the same students in an upper-middle-class white Episcopal church or Jewish synagogue and the response, of course, is quite different. But why "of course"? Close inspection of the content of the discourse in these different settings may reveal a great deal of overlap: talk of God, love, forgiveness, faithfulness, and so on. Apparently the discourse is packaged — framed, structured — in a more meaningful way in one context than in the other.

This, it seems to me, is the heart of the matter when we consider religious discourse as public rhetoric. Is it that Jerry Falwell’s ideas are so alien to the American democratic tradition that thoughtful intellectuals dismiss them on rational grounds after careful consideration? Or does the structure of Falwell’s discourse itself lead them to dismiss his ideas out of hand? We may be correct in saying that FaIwell’s ideas are indeed alien to the ways in which most academics think. But I suspect there is more to it than that. The reason we know they are alien is probably, in part, because of the way in which his discourse is put together. By the same token, we may find the U.S. Catholic bishops’ statement on nuclear disarmament much more compelling (my students do), and part of the reason is probably that it contains a discursive structure with which we are more comfortable.

Communication about Societal Goals

Let me broaden the scope a bit more. The issue is really not whether academics can appreciate Falwell’s or the bishops’ discourse better, although that may be important too. The issue is whether different segments of the society can speak effectively to one another about broad issues of societal importance. At present, much evidence (including some from the kinds of opinion surveys I have just maligned) indicates that religious conservatives and religious liberals in the United States are deeply divided — on nearly everything. The two groups are also about equal in numbers, each constituting about 40 percent of the adult population (judging from the ways in which people categorize themselves). And both sides express enormous hostility and misgiving toward the other.²

The reasons for this hostility and misunderstanding are extremely complex. They include historic precedents, different organizational trajectories, and even class differences. But they also reflect different styles in the use of public discourse. Like my students, religious conservatives can walk into a church and sense almost instantly that it is "too liberal" for them. How?

The issue can be broadened even further. Surely this is an instance of what Habermas has called communicative action.³ It exemplifies the problem of different segments of society trying to engage in competent communication about basic societal goals. One side wants to curb abortions; the other side wants to restrict restrictions on abortion. Both sides want to reduce the risk of nuclear war. And both sides would benefit by being able to realize the religious goals of reconciliation and forbearance. But if Habermas is correct, communicative action often breaks down because of unrecognized barriers built into the very speech acts it comprises. We need to understand discourse itself, he argues, in order to participate effectively in the public sphere.

Where to start? I have used the phrase "public rhetoric" in the title of this chapter, and there is in fact a growing literature on public rhetoric to which we might turn. Efforts have been made to suggest ways in which public rhetoric may differ in stable and unstable settings. Content analysis has been done on presidential speeches to test these ideas empirically, for instance. Some effort has been made to determine whether popular television preachers consistently express themes presumed to be compatible with values in the larger society. Other studies have examined moralistic themes, success motifs, and images of the collectivity in public speeches.

Though promising, the literature on public rhetoric is not what I wish to consider here.⁴ I wish to focus instead on the possibility of mining recent works in the field of literary criticism for some insights into the structure of religious discourse. A very ample tradition exists here as well, especially because religious texts have been fair game for literary analysis for a long time. Two works of fairly recent origin, however, seem particularly valuable.

Literary Models

Northrop Frye’s The Great Code: The Bible and Literature is a masterly analysis of the biblical canon by one of the foremost literary analysts of our time.⁵ It is an attempt to say, from a literary standpoint, what is distinctive about the biblical texts. Frye focuses on language, myth, metaphor, typology, imagery, narrative, and rhetoric. In so doing, he demonstrates the importance of discursive structure to the communication of religious meaning. He also supplies some general concepts, as well as numerous substantive hints, about how to analyze religious discourse. Although the book deals specifically with the Bible as a written text, the analytic framework is sufficiently broad to be applied to many other kinds of religious discourse as well.

Susan Rubin Suleiman’s Authoritarian Fictions: The Ideological Novel as a Literary Genre is quite different.⁶ With the exception of a rich twenty-page section on "exemplary narratives," which focuses on biblical parables, the book is not about religious discourse at all. Its examples are drawn primarily from the works of nineteenth- and twentieth-century French novelists (Balzac, Aragon, Bourget, Nizan, and others). She is concerned, though, with a particular kind of novel — the novel that attempts to persuade its readers of the validity of a doctrine. And in this, her work is immediately relevant to the study of religious discourse.

The two studies complement one another. Frye’s creates a stage; Suleiman’s fills in the props. From Frye we learn some of the ways in which the arrangement of words in religious texts influences their meaning. From Suleitnan we discover some of the particular strategies that writers and speakers may use to shape that meaning. Both are concerned with the restriction of meaning — that is, with the ways in which the relations among words influence the variety of interpretations that can be drawn from those words. Together they give us clues about the ways in which religious discourse may function in public settings.

It is important, before we launch into more specific arguments, to understand what these writers believe to be distinctive about religious discourse. Suleiman includes religious discourse within a larger set of communication that she describes as "ideological" or "authoritarian." This kind of communication attempts to persuade readers or listeners f the correctness of a particular way of interpreting the world. Usually it refers explicitly to, and identifies itself with, a recognized body of doctrine or a system of ideas. This defines a very broad category of communication including philosophical and religious discourse. But it apparently does exclude many other kinds of communication, such as nversation that is not aimed at persuading someone of a particular point of view and discourse oriented solely toward description or enrtainment (though this is not to say that these other forms of discourse ight not have some ideological overtones).

Frye Sets the Stage

Frye perceives religious discourse as more distinctly differentiated. Biblical discourse at least (he does not attempt to generalize to, say, primitive myths) makes use of poetic and metaphoric imagery but also purports to tell a historic story. And yet these stories are not merely descriptions of the past; they are told to convey specific ideas about the sacred and its relation to society. Frye claims that it is extremely important that biblical stories be regarded as "historically true," even though external sources of validation are generally lacking. A distinctive feature of religious discourse is that it presents itself as truth through the ways in which the discourse itself is internally arranged. This means that a biblical text must avoid making certain kinds of claims that would render it subject to external verification and must also demonstrate certain kinds of internal coherence. For example, no evidence (Frye claims) exists for the life of Jesus outside the New Testament; consequently, the writings that refer to Jesus must conform to certain criteria to avoid making this a problem: "Evidence, so called, is bounced back and forth between the testaments like a tennis ball; and no other evidence is given us. The two testaments form a double mirror, each reflecting the other but neither the world outside" (p. 78). Frye stops short of trying to capture these criteria in any simple formula. But his view that religious discourse depends heavily on its own internal arrangement constitutes the basis for one of the central themes of his analysis.

He expresses this theme metaphorically. Religious discourse revolves around itself, Frye asserts, creating both centripetal and centrifugal motion.⁷ The centripetal aspect refers to the "primary" or "literal" meaning of the text. It depends on not questioning the words, not looking for deeper meanings or applications or connections, but simply taking the story at face value, as for example, in reading the Exodus story as an account of a historical episode. The centripetal aspect is also illustrated by the quotation about the two testaments forming double mirrors. Centripetal meaning derives from what other biblical scholars have called its "intertextual coherence." One text within the biblical canon refers to another, and that one refers to another, thereby providing a kind of closed system — a system that reinforces itself. Centrifugal meaning, in contrast, refers to the more numerous connotations and layers of interpretation that "spin off" from a religious text. The Exodus story may be taken not simply as a historical account but as a message of hope, an illustration of redemption, a metaphor of new life or even of revolutionary possibilities. Frye suggests that religious discourse invites both these kinds of meaning, that it encourages both a closed reading and an open horizon of broader meanings, and that it functions effectively only when these two forces are held in tension.

Frye provides numerous examples of the ways in which the construction of religious discourse sets in motion these two opposing tendencies. He observes, for example, that the biblical texts are filled with metaphor, especially metaphors of the "anti-logical" A-is-B variety (e.g., "Joseph is a fruitful bough"). Others — most notably Ricoeur — have made the same observation, arguing that metaphor contributes to the multivalency of biblical meaning and thus to the enduring appeal of biblical texts.⁸ But Frye’s argument is different. He agrees that metaphor can open up multiple layers of meaning — can, in his terms, generate a centrifugal force that opens the text to larger interpretations. At the same time, Frye suggests, metaphor also closes down possible meanings. When a text states that "Christ is God and man," it must also be read literally despite its reason-violating message. Metaphor, in this sense, sets in motion a centripetal force that restricts the text to a narrow interpretation. Always there is the possibility of both kinds of interpretation. And yet the very way in which the metaphor is constructed also predisposes it to be interpreted in one way rather than another. We will return to this point, but for the present it will do to establish the point that metaphor illustrates one of the ways in which the relations among words (as much as the words themselves) are important to the interpretation of religious discourse.

Paradoxes of Self-Reference

Frye also examines the dual, opposing tendencies evident in biblical uses of imagery about time, history, nature, and even discourse itself. The last of these is especially interesting. There is, as Niklas Luhmann has recently observed, a paradoxical quality in religious discourse that imitates the paradoxical character of life in general.⁹ Frye’s analysis emphasizes the extent to which this paradoxical quality is suggested by the Bible’s own references to itself. It is a written text and therefore it has a beginning and end and is subject to questions of internal coherence

— all features of centripetality, of the closing down of possible meanings. Yet the Bible also points to "a speaking presence in history." It is a text that implies an absence of boundaries, an openness to new revelations a centrifugal tendency. Moreover, religious discourse is deliberately ambiguous as to which of these orientations is being referred to at a given point. The phrase "word of God" refers both to the Bible itself and to that "presence," for instance. Only by examining the context in which a phrase like this is used can we determine which meaning is implied. Again, it is the relationships among words — the broader configuration — that determines what they signify.

One other basic point to be gleaned from Frye appears in his final chapter: religious discourse tends to evoke either the centripetal orientation or the centrifugal orientation as a kind of overarching gestalt from which all its internal content is viewed. Religious discourse has both these tendencies inherent within it, says Frye, but its practitioners have generally gravitated toward one pole or the other. Some have felt more comfortable emphasizing the closed aspect of biblical meaning; others, its more open, expansive interpretations. Frye himself seems inclined to favor the latter, if one or the other must be chosen. But he attributes the tendency toward polar orientations to the ambivalent quality of biblical discourse itself. On the one hand, he suggests, the nonrational, metaphoric construction of religious discourse invariably draws the practitioner toward additional layers of meaning. The texts themselves seem to invite the reader to relate ordinary experience to them, thereby developing continuously elaborated interpretations of both. Frye writes that "the dialectical expansion from one ‘level’ of understanding to another seems to be built into the Bible’s own structure, which creates an awareness of itself by the reader, growing in time as he reads, to an extent to which I can think of no parallel elsewhere" (p. 225). It is for this reason, Frye believes, that religious faith can never be reduced to simple doctrinal statements but requires re-creative action and thought to the point that it becomes too complex to understand and thus produces an inevitable degree of inherent doubt. Those who regard religious discourse in this manner, Frye suggests, emphasize its multiplicity of meaning, its centrifugality. In their view, the proper approach to religious discourse is one that says "there is more to be got out of this" (p. 220).

There is, however, the opposing tendency as well. Religious discourse does fold back on itself. It does exclude many interpretations. It damns heresy, hypocrisy, and wrongdoing. It begs for a literal reading. And this, coupled with the uncertainties inherent in its centrifugal interpretations, encourages some to emphasize only the centripetal orientation and find security in delimiting the range of biblical interpretations. As Frye puts it, "man is constantly building anxiety-structures, like geodesic domes, around his social and religious institutions" (p. 232). This orientation deemphasizes freedom, variety, multivalency. It seizes on those metaphors that are most conducive to exact, literal renderings. To some, this itself is heresy, of course. But to Frye it is a normal reaction to religious discourse. Using a graphic image, he concludes, "the normal human reaction to a great cultural achievement like the Bible is to do with it what the Philistines did to Samson: reduce it to impotence, then lock it in a mill to grind our aggressions and prejudices" (p. 233).

Suleiman Supplies the Props

I mentioned earlier that Frye fulfills a kind of general stage-setting function for thinking about religious discourse. Suleiman is more helpful for supplying the specific props. Her book is replete with examples of the ways in which texts restrict possible interpretations in order to drive home the validity of a particular ideological position. She helps us understand, in Frye’s terms, the centripetal forces at work in religious discourse. And by contrast, we can also appreciate how different structures may reinforce centrifugal tendencies. There is, in fact, a striking resemblance between Suleiman’s interest in meaning and Frye’s (although she cites Frye only once, and then in a different context). She contrasts two kinds of novels — one that exhibits centrifugal tendencies, the other in which centripetal forces predominate. "Modernist" novels, she says, seek to "multiply meaning" (or even, as Barthes observed, to pulverize it). The roman a these, in contrast, "aims for a single meaning and for total closure" (p. 22). Her concern is with the latter.¹⁰

One of the ways in which certain kinds of texts or discourse close down the array of possible meanings, says Suleiman, is through sheer repetition. By saying things over and over, texts reveal the way in which we should interpret them. If there is confusion or ambiguity the first time around, by the nth time we should be clear. This sounds like an obvious — and simple — point. But Suleiman demonstrates that the study of repetition in texts is anything but obvious or simple. If something is repeated exactly the same way at different points in a text, it actually fails to achieve its goal of creating greater clarity. There has to be redundancy, but it has to appear in different settings and across different features of a text before we will get the point. To show how this might work, Suleiman develops an elaborate taxonomy describing the "principal constituents" of any narrative text. At the level of the story told, she distinguishes characters, contexts, and events. And at the level of discourse (the telling of the story), she distinguishes narration, focalization, and temporal organization. She also offers a number of subdistinctions, such as those describing the various qualities of a character and the kinds of narrative functions fulfilled by characters’ actions. Having made these distinctions, she is then able to show all the possible permutations of circumstances under which redundancy might or might not be present. For example, redundancy might consist of the same event happening to two or more characters in a story. Or it might consist of the same event happening two or more times to a single character. At a wholly different level, redundancy might consist of the narrator pronouncing the same commentary several times about a single character, or indicating several times that his or her information has come from a single source. In all, Suleiman discusses twenty-three types of redundancy.

Whether all these redundancies are equally important and whether they can even be distinguished neatly in specific cases remains open to question. We can, however, benefit from the general thrust of her argument without agreeing with it in every detail. She shows us two things about discourse analysis that go well beyond the kind of word counting that has been used by content analysts to measure repetition and thematic emphases. First, many kinds of redundancy may be built into texts in ways that are more subtle than the usual methods of content analysis would reveal. For example, we may need to consider verb tenses and personal pronouns to ascertain whether characters generally use the same "voice" or not. We may want to examine the settings in which dialogue occurs to see how similar or different it is from one instance to another. We may even need to see if the order in which narrations and their interpretations are given tends to be the same or different throughout a text. Second, we learn from Suleiman’s discussion that an entire discourse cannot be treated as a single text for purposes of examining repetition. For example, we could learn only a little about repetition by counting the number of times the word abortion appears in the transcript of a religious broadcast. Meaningful analysis entails dividing discourse into a number of analytic categories into different narrations, characters, events, sequences, and the like — to determine how much redundancy exists across these various categories. We would want to know, for example, whether the use of the word abortion in a religious broadcast occurred only within narratives or in a wider variety of discursive forms, whether it was spoken by more than one narrator, whether it was spoken in the same "voice," whether it occurred consistently in a particular kind of sentence structure, and so on. All of these could reinforce a particular meaning attached to the word on the one hand or could call forth a variety of interpretations on the other.

Exemplary Narrative and Apprentices

Other chapters of Suleiman’s book deal less formalistically but nevertheless effectively with particular patterns of discourse that seem to be employed frequently in ideological texts. For example, many such texts contain what she refers to as an "exemplary narrative" — that is, a story embedded within the larger text that reveals by example how we are supposed to think, act, or feel. She examines religious parables as one illustration of these kinds of narratives. Her purpose is not to show, as others have, that parables lend themselves to multiple interpretations. Rather, she means to show that the capacity of parables to make any point at all depends on a particular style of construction. In the first place, we have to be able to recognize a parable as an exemplary narrative. Usually we are able to recognize them as such because the speaker makes an abrupt change in tense and mode, thereby setting off the narrative from what has preceded it. Often there is a direct imperative as well that enjoins the listener to pay attention and to look for the point ("If you have ears, then hear"). And sooner or later the story itself becomes sufficiently enigmatic to demand an interpretation. Indeed, Suleiman asserts, the whole purpose of a parable is to set up a situation in which an interpretation is needed. Often the audience in the text actually asks the narrator to supply an interpretation, thereby speaking for the listener outside of the text. For example, Jesus’ disciples routinely asked him to say what his stories meant. And typically the narrator supplies an interpretation. The exemplary narrative works because it conforms to this identifiable construction. It establishes a relation between a sender and a receiver within the text that evokes a similar relation between the text and its actual reader or listener. It tends to be sufficiently general to allow for a wide range of identification: Jesus’ parables are about "a sower," "a woman," "a father" who had "two sons." It occurs within a larger textual context that invests it with intentionality. It also tends to be interpreted by an authoritative narrator who experiences little or no challenge to his authority from other characters or voices in the larger text.

Another specific literary device that Suleiman discusses is the use of apprentices and stories about apprentices to help drive home the author’s ideological intent. Here is her definition of an apprenticeship story: "two parallel transformations undergone by the protagonist: first, a transformation from ignorance (of self) to knowledge (of self); second, a transformation from passivity to action" (p. 65). Or, more simply, an apprenticeship story is about a hero who goes forth into the world to find himself and who achieves this goal by undergoing a series of adventures or tests. Religious discourse is, of course, replete with such stories — from the biblical stories of Jacob, Joseph, Jesus, and Paul, to modern equivalents about religious converts, to such fictionalized variants as Luke Skywalker and Indiana Jones.

Like the exemplary narrative, apprenticeship stories work because they conform to certain rules of construction. In addition to the apprentice, various antagonists must be introduced to provide contrasts and to present hurdles to overcome. Often there is a guide or mentor who functions not only to help the apprentice but to make explicit the lessons the apprentice has learned. Above all, a virtual identification must be created between the reader and the protagonist. This is often accomplished by dialogue between the guide and the apprentice that parallels the dialogue going on between the narrator of the text and the reader. For instance, Jesus counsels his disciples and receives questions from them in a way that permits the reader of the text to ask the same questions and receive the same answers. This sort of thing also takes place when the apprentice is a figural actor who exemplifies general characteristics of a certain social class or a particular time. Bunyan’s Christian in Pilgrim’s Progress evokes images of the Puritan artisan more generally, just as do the communist heroes of the twentieth-century fiction that Suleiman analyzes.

Applications to Ordinary Discourse

All of Suleiman’s examples come from formal texts (novels, fables, parables), and all of Frye’s examples come from the biblical canon itself. The question thus arises as to whether any of their conclusions are valid with respect to the analysis of more ordinary kinds of religious discourse. Can we expect to find the same level of importance associated with complex structural arrangements in simpler and more spontaneous discourse, including verbal discourse, or is this the stuff of which only the sophisticated texts of great traditions and talented writers is composed? One way of gaining at least a partial answer to this question is to consider whether or not the professional clergy who transform the more sophisticated literature of religious traditions into verbal discourse have any training in such matters.

We would have to query clergy on this issue to find out, but we do know that seminary training routinely includes courses in sermon preparation and hermeneutics, and we know that the material used in these courses often reflects a great deal of sophistication about the matters we have been discussing. Consider, for instance, the advice given in two of the most popular texts on preaching, Elizabeth Achtemeier’s Creative Preaching: Finding the Words and Leander E. Keck’s The Bible in the Pulpit.¹¹ Both are highly self-conscious about the importance of words — not just the deeper truths words convey but the words themselves. Achtemeier writes, for example, that "if we want to change someone’s life from non-Christian to Christian . . . we must change the images. . . in short, the words by which that person lives" (p. 24). Both are also mindful of many of the specific issues raised by analysts such as Frye and Suleiman. Keck, for instance, decries the tendency seen in religious broadcasting and among fundamentalists to reduce religious truths to simple doctrinal formulas. For the modern highly educated person who believes "that truth is glimpsed momentarily and in fragments, that it lacks symmetry, that it is awkward and angular as it breaks through to us," he suggests, the very phrasing of religious discourse in consistent, propositional statements will sound unreal (p. 42). The kind of text that asserts a single imperative (Suleiman’s "ideological" text) should be avoided at all costs, he says. Instead, emphasis should be placed on simply relating the biblical stories, and conscious attention should be given to letting the diversity of different characters shine through. Rather than using the authoritative voice of the traditional narrator, he suggests, the preacher should tell stories that reveal his or her own vulnerabilities and thereby permit greater identification by the audience. Achtemeier, also writing primarily for clergy in liberal mainline churches, offers similar advice. Speak about actions, she urges, rather than propositional truth; retell the biblical stories, but tell them in conjunction with — and parallel to — stories from everyday experience so that parishioners can identify with them.

Another, more direct way of illustrating that the analysis of discursive structures is relevant to understanding religious discourse is to look briefly at some actual examples of religious discourse. I suggested earlier that this kind of analysis might prove useful for understanding the cultural chasm currently separating religious liberals and religious conservatives. Frye’s distinction between centripetal and centrifugal emphases seems relevant to this distinction. And the advice of Achtemeier and Keck seems to suggest a certain affinity between the centrifugal emphasis and the more liberal orientation. Are the same kinds of differences evident in actual sermons?

Two Examples

Consider the following: the pastor of a liberal Protestant church preaches a sermon on the story of the prodigal son called "Intolerable Love"; across town, a preacher at a small fundamentalist church delivers a sermon called "The Meaning of Life." If we had not been told one was given in a liberal church and the other in a conservative church, could we have placed them correctly merely from examining the two transcripts? Or, perhaps more importantly, what do the two texts reveal about the differences between liberal religious discourse and conservative religious discourse?¹²

This is not the place to examine the two texts in detail, but we can illustrate how some of the points we have been discussing might be applied.¹³ From the surface content of the titles alone, we would get little clue as to the underlying differences between the two. Both speak of broad existential, psychological themes; neither focuses specifically on a religious or biblical phrase. Moreover, when we consider the structural arrangement of the words in each title, we also see similarities: both are quite brief, both contain a primary noun and a modifier of that noun, and both relate the noun and its modifier in a way that seems sufficiently paradoxical or contradictory to evoke a question: how can love be intolerable, how can life have meaning (specifically, a single meaning — "the meaning of life"). Yet (with a little sleight of hand that comes from peeking ahead and thinking of Frye’s basic distinction), we can already sense that one text is going to emphasize centrifugal meanings and the other, centripetal meanings. We sense this from the fact that "intolerable love" genuinely opens up all sorts of questions and possible answers, whereas "the meaning of life" implies that something as vague and complex as "life" is going to have a simple interpretation that can be called "the meaning" (consider the quite different implications that would be evoked by the phrase "the meanings of life" or even "meaning in life").

Pushing into the body of each text, we find further similarities and differences. For example, both employ some of the devices Suleiman discusses for creating an identification between the audience and either the narrator or characters in the story. In one, the speaker begins: "Who of us can read the story of the Prodigal Son nowadays without a catch in our throats?" The pronouns are all plural; they categorize us and the narrator together. In the other, all the pronouns are singular. But they occur in a sequence of questions that collectively encompasses everyone in the audience: "How can I handle death? How can I overcome my feeling of loneliness? How can I better manage my time?" As the sermons progress, though, we increasingly see two different patterns in the relations suggested between narrator and audience. In the sermon about intolerable love, the narrator seldom refers to himself; he consistently uses plural pronouns; and when he does refer to himself, he refers to someone who is himself struggling, learning, uncertain, weak. He objectifies the story and relates it to himself and to the collective "we" with phrases that make the latter dependent: "like a resentful child," "broken," "caught." In the other sermon, the narrator tells numerous anecdotes about himself; he becomes a much more intrusive object in his own narratives; and in these stories his authority is never challenged. Indeed, they are often apprentice stories in which the speaker as apprentice encounters other people and then observes their confusion and supplies them with answers. But they are not genuine apprentice stories of the kind Suleiman analyzes. They show the speaker as one who has already found the answers or who instantly recognizes them or sees their applicability to others’ problems. In short, the role of the narrator in the text proves to be a key to the relatively more "ideological" or "authoritarian" tone of the second sermon in comparison with the first.

Opposite Flows

Most interesting, though, is the manner in which the two sermons move between simple, univocal meanings and complex, multivocal meanings. The narrator of the sermon on love observes near the outset that the story of the Prodigal Son is a "simple little story." And in retelling the story, his sentence structure underscores its simplicity: "First we have the younger son, the prodigal. We remember him most vividly. He is hungry for life. . . . He acts. . . . He is the experimenter These are simple, short, declarative statements. They reinforce the story’s surface simplicity. The same sentence structure is present as the narrator continues his description of the other characters in the story, the elder brother and the father. But then the narrator switches to interpretation. He announces the switch by stating that "we learn some things" from the story and that this was Jesus’ point in telling the story. Now the meanings conveyed become more complex. The story, it turns out, is not so simple after all. It is a story about envy, alienation, forgiveness, searching, the self, grace, reconciliation. And now the very complexity of the sentences forces the listener to abandon any conception of simple, straightforward interpretations. Here is the key sentence that summarizes the main point of the sermon: "The biblical notion of the wrath of God is not so much that of the anger of a just God, but it has to do with God’s passionate intolerance towards all forms of sin and what sin does to the world which is loved." Forty-four words! Delivered orally, it is little wonder that the story evokes, as the narrator said it would, "a catch in our throats." Meaning is not so simple after all. One probably cannot even grasp it on one hearing. The sentence does not invite clarity. It invites a sense of mystery to be probed, re-examined, and experienced.

Contrast this movement from the simple to the complex with that in the, other sermon. Here the flow moves in the opposite direction. As noted previously, the text begins with some simple questions that draw the audience into their orbit. Each question is short and beguilingly simple. But there are nineteen of them! And they are delivered as a single unbroken chain. Together they signify the complexity of life — the problems of stress, decision making, communication, fear, pain, and the like. Then there is a series of short narratives. Each describes a seemingly simple answer to the question of the meaning of life and then negates this answer with statements that typify confusion, mystery, openness, searching: "he does not know real answers," "you don’t know where you’re going," "we find ourselves never getting anywhere," "what are the real answers," "knowledge does not contain answers," "there are no answers in power," and so forth. The words themselves open up the complexity of the subject. But, as in the other sermon, so does the sheer length of many of the sentences. For example, here is a sentence that describes the confusion among the followers of Jim Jones’s People’s Temple: "Every one of them was steeped in traditional religion, and they left it because they found a group of people that loved them, that gave them more answers, that cared for them, that generated warmth — when all they did was sit in pews and have meaningless things told to them so they could go out and live meaningless lives." Fifty-nine words. The very fact that this is grammatically a run-on sentence places the listener in a situation of openness: the specific clauses of the sentence could go on endlessly. Finally, though, the narrator asks directly, "where’s the answer?" And then he gives the answer: "Jesus is the answer." Note the simplicity of the sentence. Moreover, the sentence is actually framed, set apart from all other sentences in the text, enclosed: the narrator reveals that he saw it on a billboard along the highway when he was driving home from his grandmother’s funeral. Finally, lest there be any confusion on the audience’s part, he makes explicit toward the end of the text, just as the other narrator did at the beginning, that things are really quite, simple: "Does life ever seem mysterious?" he asks. The answer, he says, "is simply found," and then he underscores the point by saying it is "not hard to understand," it is "so simple," and it is "profound in its simplicity." Again he relies on slogans to underscore the point — simple refrains from well-known hymns, short biblical quotations. And twice he puts these statements in the mouths of children.

The two sermons illustrate strikingly the contrast between Frye’s centrifugal and centripetal tendencies in religious discourse. In the sermon on intolerable love, the movement runs from simplicity to complexity, from a restricted literal reading to a figurative multivocal reading. In the sermon on the meaning of life, the direction of movement is from complexity and searching in many directions to simple, succinctly codified answers. The one "opens out" into broader meanings; the other "closes down" possible meanings to a single answer. Is it perhaps this contrast that accounts for the chasm separating religious liberals and religious conservatives? Had we considered both sermons in full, we would have observed that the liberal sermon actually devotes more time to quoting and paraphrasing the Bible than does the conservative sermon. We would have also noted that the scriptural text speaks more directly and objectively in the liberal sermon, while the narrator himself is more intrusive in the conservative sermon. We would have observed, too, that the conservative sermon does not contain a rigid set of "thou shalt not’s" or spell out in propositional statements what it means to assert that "Jesus is the answer." In other words, the two sermons do not differ in many of the ways we might have expected them to on the basis of preconceived notions about liberalism and fundamentalism. The main contrast is in style and in the openness or restrictedness of meaning that is connoted by that style.

Broader Implications

What does this imply about religious discourse in the public sphere? Perhaps it is the style of discourse that causes it to communicate in some contexts and fail utterly to communicate in others. Perhaps clues are buried in the structure of discourse itself that say to us, "life is really too confusing and here are some simple answers," or "the answers we have are really too simple and we need to recognize the complexity of it all." As in the sentence structure of the two sermons, there may be an implicit emphasis on the priority of centripetal meaning or on the priority of centrifugal meaning. The two emphases may also be spelled out explicitly, as they were when the two ministers employed the word simple itself.

In drawing on Frye and Suleiman, I have made no attempt here to suggest a rigorous, easily codified, let alone quantifiable method of examining the structure of meaning in religious discourse. Indeed, the examples given have specifically shown the precarious bases from which inferences about how texts operate are drawn. The study of discourse is, after all, an interpretive enterprise. It depends on preconceived frameworks — which is why I introduced Frye and Suleiman before discussing the two sermons — rather than on raw induction from the empirical world.

Nevertheless, it seems to me that the study of public rhetoric can advance by following some of the clues suggested by this reading of Frye and Suleiman. We have considered the importance of examining metaphor and narrative not as straightforward thematic content but as building blocks in the construction of texts. We have considered redundancy and special kinds of narrative structures that draw the reader into the text by providing exemplary stories and by telling the exploits of apprentices.

If there is any "simple answer" to the analysis of discourse that goes to an even more elementary level than these clues suggest, it is probably to be found in the examination of distinctions and connections, contrasts and parallels. Metaphor, for example, consists of words that are clearly distinct, drawn from two different realms ("the sun is a chariot"), rather than two words that can actually be equated ("God is sacred"). And as Frye observes, metaphor depends on drawing a connection between these two contrasting words (thus, in his analysis, the importance of is). But we car broaden the idea beyond metaphor. Suppose a different kind of connection is drawn between two contrasting words. It has often been noted, for example, that Flannery O’Connor made frequent use of the phrase "as if for this purpose.¹⁴ This kind of connection muddies up our interpretations, forcing them to be more centrifugal than centripetal. When I said earlier, for instance, that it was "as if" our methods forced us to conceive of religious communities as silent worlds, what did I mean? Not that it literally did, but that there is a provocative sense in which this is the case, and giving some thought to this sense may open up the matter for us in unexpected ways.

More generally, issues of redundancy and methods of drawing an identification between readers and characters in the text also boil down to questions about distinctions and connections, contrasts and parallels — which is to say, structural features of discourse. I am suggesting that we need to pay greater attention to this. Religious discourse in the public arena is not simply talk about the gods in an otherwise secular context. It is the use of a certain rhetorical style, a style that conforms to certain rules of underlying structure but that communicates only to the extent that this structure is appropriate for the uses to which it is put.

NOTES

1. I have examined the epistemological underpinnings of this approach and alternative approaches to cultural meaning in general in my book Meaning and Moral Order: Explorations in Cultural Analysis (Berkeley and Los Angeles: University of California Press, 1987), especially chap. 2.
2. For further detail, see my book The Restructuring of American Religion: Society and Faith since World War II (Princeton: Princeton University Press, 1988).
3. Jurgen Habermas, Communication and the Evolution of Society (Boston: Beacon, 1979).
4. What I find disappointing about many studies of public rhetoric is (a) their tendency to focus only on content, rather than the form or structure of discourse, (b) their tendency to thematize this content, and (c) their tendency to regard these themes as reflections, or reinforcements, of collective values — which apparently lie hidden somewhere in the subjectivity of the collective conscience.
5. Frye, The Great Code: The Bible and Literature (New York: Harcourt Brace Jovanovich, 1982), p. 78.
6. Suleiman, Authoritarian Fictions: The Ideological Novel as a Literary Genre (New York: Columbia University Press, 1982).
7. I am aware, of course, of the more general usage of this metaphor in linguistics and literary criticism, as well as the reservations that have been expressed especially by Marxist critics of Frye’s broader theoretical framework.
8. Paul Ricoeur, The Philosophy of Paul Ricceur (Boston: Beacon Press, 1978).
9. Luhmann, "Society, Meaning, Religion — Based on Self-Reference," Sociological Analysis 46 (1985): 5-20.
10. I have had many fruitful discussions of Suleiman’s book and of the problem of meaning-restriction in texts more generally with Marsha Witten, to whom I want to give special thanks. She is currently engaged in some very promising empirical work on this issue.
11. Achtemeier, Creative Preaching: Finding the Words (Washville: Abingdon, 1980); Keck, The Bible in the Pulpit: The Renewal of Biblical Preaching (Nashville: Abingdon, 1978).
12. I am grateful to Penny Becker for collecting these and other sermons for me under a work-study program sponsored by Princeton University.
13. Marsha Witten has written a very useful paper, entitled "The Structure of Symbolic Codes and the Restriction of Meaning: Devices of Disambiguation in a Fundamentalist Christian Sermon," that examines the second of these two sermons in depth, using both quantitative and qualitative methods.
14. For instance, in the familiar opening sentence of Wise Blood, O’Connor writes, "Hazel Motes sat at a forward angle on the green plush train seat, looking one minute at the window as if he might want to jump out of it, and the next down the aisle at the other end of the car."

In this chapter I argue for the continuing importance of perspectives that emphasize the cultural dimension of religion. I also examine some of the theoretical assumptions that, in my view, have hindered the development of a cultural perspective in the sociology of religion, and I suggest some of the ways in which the so-called "new cultural sociology" may be able to make a positive contribution. New directions in the cultural analysis of religion will, I believe, be closely associated with the rediscovery and redirection of cultural analysis in the discipline more broadly.

There is nothing novel, of course, in suggesting the presence of a close connection between religion and culture. Religion, after all, is constituted by symbolism — whether in primitive amulets, totems, and rituals, the earth and sky god mythologies of ancient civilizations, crucifixes and relics of the medieval church, formalized texts and creeds of the world’s great modern religions, or even the sacred rites and markers we use to define ourselves, our relations to nature, our sense of personal identity, and our collective loyalties and destinies. If symbolism is the essence of culture, then religion surely has an important cultural dimension.

This is not to say that religion consists only of symbolism and is nothing more than culture. Like any social institution, religion consists of power and status relations, it depends on financial contributions which in turn make possible the provision of salaried professionals, and its strength or weakness in any society depends greatly on the bonds that inhere within its membership, the organizational resources it can mobilize to pursue ends in competition with other organizations, and its relations to broader patterns of wealth distribution, time allocation, communications media, and perhaps above all the state. To emphasize the cultural dimension of religion is not to deny the importance of any of these other characteristics. But such an emphasis does reflect a conviction that religion is something more than the nuts and bolts of social networks, that it is something more than the population ecology of competing organizations, and that its essential features cannot be understood entirely in the same terms one might use to understand a political party or an economic transaction. A focus on the cultural dimension of religion represents a decision to take seriously the symbolism of which religion is constituted as an object of study. Rather than merely pointing out the symbolic vistas along the expressway to an analysis of organizations, economics, motivations, and other "determinants" of religious belief, one sinks roots into the neighborhood of religious symbolism itself.

Many kinds of observers have, in fact, taken up residence within the cultural domain of religion. Anthropologists and ethnographers of all kinds come most readily to mind because of their insistence on understanding the expressions and language and mores that hold religious communities together. We have in the sociological literature a rich tradition of field work, including in recent years a large number of participant-observer studies conducted in new religious movements and an increasing number of congregational studies, many of which have paid close attention to the ways in which religious symbols (both verbal and behavioral) are patterned. We also have many fine examples of phenomenological work that has compared the symbolism of various religious traditions and sought to uncover the deep structure of religious mythology. Much of this work has been concerned first of all with problems of description, recording, induction, and the generation of greater appreciation for the richness, depth, diversity, and power of religious culture.

Sociologists of religion have generally been receptive to this broader body of work, benefitting from it, incorporating some of it into our standard accounts, and contributing to it as well. We have, however, been most deeply influenced by the theoretical orientations within our own tradition. These orientations have given us a common language with which to talk about religious culture, provided the jumping off points for empirical studies and critical discussions, and in these ways limited the kinds of problems we could address and given us the tools for addressing them.

Each of the major classical theoretical traditions from which sociologists have drawn inspiration have pointed to a close connection between religion and culture. Marx discussed religion within the more general framework of criticizing Hegel, Feuerbach, and other more popularized versions of what he and Engels described as the "alienated life elements" and "false consciousness" of bourgeois society. Though his work was always more seriously focused on the laws of oppression and expropriation internal to capitalism itself, Marx was also a brilliant analyst — and producer — of ideology, and he contributed some valuable, if sometimes neglected, insights into the workings of symbolism and discourse. Weber was, of course, far more conscientiously devoted to the systematic study of religion than was Marx. In Weber, we have a rich tradition of sociological analysis that takes seriously the cultural content of religion — its theological orientations, its understandings of evil, its soteriological teachings, the utterances of its charismatic leaders, and so on. And in Durkheim, a lifetime of work in which the changing bases of moral community lie at the center culminates in a full-scale discussion of religious myth and ritual, a discussion rich with insights about the ways in which symbolic categories are constructed and how they function. From these traditions, we have inherited not only the specific substantive emphases that distinguish each from the others but a legacy of common themes as well: (1) a theoretically grounded rationale for the importance of studying religion in any serious effort to understand the major dynamics of modern societies, (2) a view of religion that recognizes the significance of its cultural content and form, and (3) a perspective on religion that draws a strong connection between studies of religion and studies of culture more generally — specifically, studies of. ideology in Marx, studies of rationalization In ~ and studies of the symbolism of moral community in Durkheim.

The Geertzian Legacy

While the impact of these classical theories has remained strong, I would like to point to a specific contribution that, in my view, has served as a kind of watershed in our thinking about the cultural dimension of religion: Clifford Geertz’s essay "Religion as a Cultural System," published in 1966.1 Although Geertz, an anthropologist, was concerned in this essay with many issues that lay on the fringes of sociologists’ interests, his writing is clear and incisive, the essay displays exceptional erudition, and it provides not only a concise definition of religion but also a strong epistemological and philosophical defense of the importance of religion as a topic of inquiry. Few essays have been as widely reprinted in anthologies on the sociology of religion, as routinely cited in textbook discussions of religion, or (I suspect) as commonly required on the reading lists for graduate courses and general examinations in sociology of religion.

Given the clarity and power of its prose, Geertz’s essay was important in its own right; nevertheless, it entered an ongoing stream of theoretical discourse in the sociology of religion, and, like any such contribution, its significance lay not only in what it said but also in what others read into it. It shaped, I believe, the ways in which we have as a subdiscipline thought about religious culture over the past two decades, but it did so only partly because of the theoretical and methodological stance that Geertz himself adopted. Rather than simply presenting a powerful new way of thinking about religious culture, he provided a point around which a number of themes already evident in the sociology-of-religion literature crystallized.

As evidence, let me reconstruct briefly the theoretical milieu Geertz’s essay entered in 1966, showing how this milieu reinforced certain of Geertz’s themes, and then contrast this reception with the quite different way in which anthropologists themselves have depicted Geertz’s work. As a brief excursion into the sociology of knowledge, this exercise will provide a backdrop for the more critical observations I wish to raise. Of the various intellectual concerns dominating the sociological environment to which Geertz contributed in the mid1960s, three seem particularly worthy of mention.

First, the theoretical perspective of Talcott Parsons still exercised a powerful influence over the discipline in general. Geertz’s essay was by no means an example of Parsonian analysis, and Geertz’s work generally fell much farther outside the Parsonian framework than did the work of contemporaries such as BelIah, Eisenstadt, or Smelser. But Geertz had been a student of Parsons at Harvard, and his use of the phrase "cultural system" corresponded with Parsons’s systems terminology.² Moreover, some of Geertz’s language (especially his usage of terms such as mood, motivation, and orientation) was compatible with Parsonian terminology, and upon publication of the essay, these terms were quickly appropriated by Parsons and others working within the Parsonian framework at the time, such as Bellah and Roland Robertson.³ It was, therefore, impossible to escape entirely from thinking about the relation between Geertz’s essay and Parsonian theory.

Second, within sociology of religion (as, to some extent, within the larger discipline of sociology) a relatively complex battle was being waged between the various proponents of empirical positivism and their detractors. Survey research in particular, through the work of Gerhard Lenski, Joseph Fichter, Charles Glock, Rodney Stark, and others, was beginning to shape the ways in which sociologists thought about religion, on the one hand, while on the other hand Parsonian theories, speculative and comparative work in the classical tradition, and some of the newer perspectives of phenomenology posed challenges to empirical positivism. Geertz’s essay was ripe for appropriation by both sides in this struggle because it reflected the broader Parsonian and hermeneutical orientations and yet provided a concise definition of religion and a discussion of some of its empirical manifestations.

Third, sociology of religion was also caught up in a debate over the relative merits of substantive and functional definitions of religion. In the late 1 960s at Berkeley, for example, Charles Glock had become a vocal proponent of a substantive approach, arguing for the importance of distinguishing between the supernaturalism of religious worldviews and the naturalism of humanistic worldviews, while Robert Bellah was advancing a functional approach that defined religion more in terms of any system of symbolism capable of inducing a sense of personal meaning.⁴ Elsewhere, J, Milton Yinger, Peter Berger, and Thomas Luckmann were also engaged in defending various definitional approaches to religion. Geertz’s essay, which championed a functional approach, was followed in the same volume by an essay by Melford Spiro that advanced a substantive definition of religion.

In addition to heightening the overall interest in Geertz’s essay, the effect of these circumstances was, I believe, to sharpen a particular interpretation of Geertz’s argument — an interpretation that focused heavily on the subjective characteristics of religious culture. For those who interpreted it within the Parsonian framework, conceiving of religion as a cultural system entailed (1) drawing a distinction between culture and the more tangible realities of social structure and behavior of which social systems are composed; (2) conceiving of culture as a system of general values, goals, or value orientations that define the desired end-states of individual and collective action; and (3) emphasizing the ways in which religious culture reinforces cognitive, cathectic, and conative commitment to these end-states. Those who viewed Geertz as departing from the Parsonian scheme were thereby enabled to view religious culture more in terms of concrete views, moods, and motivational factors that could be studied empirically, especially with survey research methods. And for those concerned about functional and substantive definitions of religion, Geertz basically settled the issue by suggesting that symbols themselves might vary greatly in substance but could be studied usefully in terms of their subjective functions, such as creating moods and supplying meaning and motivation. In any case, I think it significant that Geertz’s essay seemed to inspire no new or distinct lines of inquiry among sociologists of religion but was cited sympathetically, if only in passing, by scholars with as diverse interests as grand Parsonian theory, survey research, ethnography, and phenomenology.

This reception might be taken to suggest either that Geertz focused mainly on the subjective functions of religion or that he had written in such general terms that nothing distinctive could be found. It is interesting, therefore, to observe that anthropologists drew a very different lesson from Geertz. As Sherry Ortner has observed in a useful survey of the literature, anthropologists mainly understood Geertz as having argued for a stronger connection between culture and practice.⁵ They considered Geertz’s critical contribution to be his departure from the Parsonian framework, especially his rejection of the Kluckhohns’ emphasis on value-orientations, which Parsons himself had appropriated. They also considered Geertz’s rejection of any Marxian or Freudian approaches that focused on the consequences of culture to be significant. And above all they believed that Geertz was trying to focus attention on the lived realities in which symbols came alive. It was important, from this perspective, to observe how people talked about themselves and their social relations, to look closely at their rituals, and to pay special attention to the actual situations in which labor and discourse took place.

I am not concerned here with trying to settle the issue of whether tie anthropologists’ reading of Geertz was more or less accurate than ‘that of the sociologists of religion. Both confronted Geertz within the constraints of problems that were of greatest interest in their own disciplines. Geertz was considered authoritative in both contexts but was read in quite different ways. At present, therefore, we find rather different views of what it means to study culture within the anthropological community than we do among sociologists of religion.

If this expedition into the historical backwaters helps to relativize our understanding of Geertz, it should also suggest the value now, some twenty years later, of looking critically at some of the assumptions in our own tradition. I suggested briefly that our reading of Geertz mainly helped to reinforce a "subjective" perspective on religion. Now we would do well to consider the specific ways it has done so, and on what basis.

Problematizing the Modern Conception of Religion

By subjective, I mean an understanding of religion that privileges its connection to the individual, its location within the inner part of the individual (beliefs, outlooks, felt needs, emotions, inner experiences), and its role within the individual’s personality (providing meaning, wholeness, comfort, psychological compensations, even a sense of attachment or belonging). As I have suggested elsewhere, conceiving of religion in this way tends to underemphasize its collective features or leads to conceptions in which its collective features are depicted in ad hominem terms, such as "collective conscience," "soul of the nation," "group spirit," and the like.⁶ Thinking of religious culture in this way also reinforces a view of the world in which social structure is equated with behavior or with the way things "really are," while culture becomes nonbehavior — mood, attitude, feeling — and only a perception or interpretation of the way reality actually is (an "aura of facticity," in Geertz’s words). It is easy, of course, once a conceptual dualism of this kind has been established, to argue that culture cannot be understood sociologically unless it is "explained" in terms of social structure — unless the "sources" or "causes" of religious beliefs are located within such obdurate features of the social world as class interests, power relations, social networks, family backgrounds, and the like.

A further concern that may arise from this understanding of religious culture is that it is difficult to observe or measure it. Some argue that religious culture cannot be observed adequately at all. How, they ask, can one really explore the life of faith, the richness of the spirit, the complexity of inner conviction? Others are more confident of their ability to create adequate measures, but these measures tend to entail many assumptions about the observer’s ability to probe the inner psyche of the individual. Attitude scales, projective tests, indexes of belief, and measures of consistency are the result.

I am not in any way denying the importance of the psychological or social-psychological dimension of religion. In a modern society such as ours, in which religious pluralism has resulted in a great deal of private or individualized religion, the connection between the inner life of the person and the subjective aspects of religious belief and conviction are especially significant. However difficult, partial, and inadequate they may be, studies of individual religious commitment remain vital to our understanding of modern religion. But given the many aspects of religion, few would be willing to argue that the social-psychological approach alone is adequate. Studies of churches, sects, cults, and denominations — all conceived of as concrete organizational entities — abound. So do studies of the more behavioral dimensions of individual religiosity: attendance at religious services, giving to religious organizations, participation in religious movements, support of religious lobbies, viewing religious programs on television, and so on. We have in recent years witnessed a number of new theoretical schemes — or attempts to revive old schemes in which collective, behavioral, observable variables predominate: ecological theories, economistic models, market metaphors, notions of moral order and moral economy, and cybernetic and behaviorist approaches, to name a few. We have not, in my view, sufficiently come to terms with the concept of culture. It is one thing to recognize that religious organizations exist independent of the subjective characteristics of individual religiosity and quite another to speak of religious culture itself without falling into terminology such as "moods and motivation," "worldview," "meaning," and "value orientation."

Furthermore, the attractiveness of the main alternatives — various "poststructural" or "decentered" approaches — as ways of thinking about religious culture remains limited. To take the individual entirely out of the picture, focusing only on the formal characteristics of language, may be appropriate in studies of linguistics. To consider only the ways in which genre is put together may be desirable in literary criticism. To emphasize the rationality of validity claims, warrants, and backings may be useful for understanding jurisprudence and political discourse. But religion is different: it is more personal, less rational, often deliberately expressive, often not deliberate at all.

If we want to move toward a more adequate concept of religious culture, then, it would seem best to work with (and move through) our present understandings of religion, subjective as they may be, rather than opting for some entirely different perspective on culture. In my view, we need a view of religious culture that is broad enough to subsume and problematize the subjective approach to religion itself; that is to say, we need to think about the assumptions of this approach as a product of cultural construction itself.

We need to view the individual and the internal dimensions of the individual as problematic cultural constructs rather than simply taking them for granted as the starting point for studies of religion. If we ask someone, either in a survey interview or in real life, what he or she believes about religion, we are making certain presuppositions: we are assuming that the individual is a meaningful unit of reality, that he or she has a certain degree of self-awareness regarding his or her role as a unit of reality, that this unit is or can be conceived of apart from some feature of external reality about which one can have beliefs, and that belief (or commitment in general) is an appropriate way of relating to this external reality. Assumptions like these are scarcely remarkable: in modern society we operate on the basis of such assumptions most of the time. Were we to engage in broader historical and comparative research, however, they would immediately become problematic. We would learn that they have not always been present and, for this reason, it has not always been possible to conceive of religion as a problem of personal commitment. Recognizing these assumptions as cultural constructions in our own context should also render them problematic. We could then ask empirical questions about the bases upon which these constructions are produced and disseminated.

These considerations point, of course, to the relevance of work in the social sciences that has sought to problematize the modern concept of individuality. Two lines of inquiry seem particularly interesting. One concerns the ways in which social institutions provide cues and markers that help us to think of ourselves as bounded units of reality. The state’s propensity to assign individuals identities through voter registration lists and social security numbers or more generally to reinforce conceptions of individual rights serves as an example; the roles of educational systems (through individualized test scores) and professional careers (organized around cumulative skills attached to the individual’s biography) provide further examples.⁷ This work is important because it shows the dependence of self-constructs on markers in the culture at large: the self is understood not only in terms of internal development but also as a product of external reinforcement. Also of interest are inquiries that problematize the relation between private selves and public roles. These studies suggest that the modern individual has to an important degree been the creation of a more sharply defined public sphere from which the private realm can be more clearly differentiated.⁸ Questions of interior space then become more important, as do the relations between these inner realities and those that constitute the public or external realms.

In addition, beliefs and commitments should be understood not simply as the individual’s sense of attachment to something external but as expressions that depend on language. It is not the case that we have no observable evidence to substantiate assertions that modern religion is distinctly a matter of belief or that it depends on a particular understanding of the individual. Clues about the ways in which reality is categorized can be obtained from the language of religious discourse itself. When a person tries to convert a colleague by saying, "Let me share something that I’ve found to be true in my own life," we learn something vital about the location of religion in relation to that person’s sense of identity. Or when a pastor tells a congregation, "Each of you must find Jesus in a way that is meaningful to you," we also learn something vital about the nature of contemporary religion.

We can, of course, jump immediately to the obvious conclusion — namely, that contemporary religion has become acutely individualistic — and we can decry the loss of community and predict the eventual demise of religious tradition as we know it. But we can also credit religious discourse with a more active, powerful role than that and, in keeping with that, pay closer attention to what is actually said. Religious discourse lies at the intersection of the individual and the community. It individuates conviction, as these examples suggest, but it also reinforces a sense of collectivity at the same time. The very act of attempting to convert one’s colleague may be an act of solidarity, a way of creating a mutual bond, and the pastor who preaches in individualistic terms may still invoke a sense of community by calling — in the collective setting of the morning worship service — for unity amid diversity.

Once discourse becomes part of the picture, greater attention must also be paid to its content, as opposed to its context. It is one thing to say, as Peter Berger does, that conversation about religion supplies a plausibility structure, a social context in which shared conviction strengthens the commitment of all by making subjective belief an intersubjective reality.⁹ It is another thing to say that the content of this conversation itself must become an object of serious investigation. The one approach says merely that social interaction with like-minded believers is an important phenomenon and that we should know more about how often, with whom, and in what settings this sort of social interaction takes place. The other approach says that such interaction is merely the setting in which an important cultural drama takes place. We would not attempt to understand what goes on in the halls of Congress by merely counting the number of times meetings are held; we would want to know what was said. Similarly, we should not seek to analyze religious belief merely as a function of its plausibility structure; we should also seek to know what is said.

A sermon that carries conviction and authority does so not only because it is presented to a large audience in a large auditorium by a pastor with a fine reputation. It carries conviction and authority because of its content. The pastor does not simply preach the sermon; he or she also constitutes authority within the sermon through a strategic choice of examples and self-disclosures. Stories in the newspapers about religious leaders do not offend or persuade us simply because we bring to them certain educational levels and political inclinations; they offend or persuade because something in the stories signals to us that the reported event violates certain standards of common decency or that it can be understood within some familiar framework.

More emphasis also needs to be placed on discourse as practice. The examples I have just given should also underscore this point. A long tradition of theological study has focused on the content of religious discourse — in scriptures, creeds, sermons, and theological statements. In many cases, only a given text is studied. A sociological approach, in contrast, is inevitably more interested in the text’s relation to the social world. This relation should not be construed in a way that gives priority to the social world, or even in a way that creates a sharp distinction between text and context; rather, it should emphasize the active —and interactive — exchange between the behavior of speaking and the behavior of doing. Speaking itself is, as J. L. Austin has shown, a way of doing.¹⁰ But whether it accomplishes something through the act of speaking itself or only characterizes something that must be done separately, speaking occurs in time, involves an expenditure of resources, and imposes constraints on the sequence of its own unfolding.

A sermon develops over time: it can be analyzed in terms of the resources that constitute it (biblical references, the work of biblical commentaries, the homiletic instruction the speaker received as a student, the authority to speak that is bestowed on the office of pastor), and it can also be examined in terms of the effects it has on its audience (their ability to remember key arguments, whether they agree or disagree with these arguments, whether phrases from it are incorporated into their own discourse). Beyond this, though, a sermon rextualizes certain features of its own context (e.g., it offers comments about the weather, refers in passing to the previous day’s anniversary celebration, reflects on the week’s political events). These references are not incorporated into the text simply as themes that can be understood through standard content analysis procedures. The sermon itself conrextualizes them and thus alters their meaning by placing them polemically in relationship to other arguments, by selecting some of their features at the expense of others, by incorporating them into narratives, and by presenting them in a way that evokes a certain response or identification from the listener. In short, the sermon contains its own rhythms, its own redundancies, its own image of time and space, and it paints a social canvas of its own, complete with locations for the speaker, a differentiated audience of listeners, and various bystanders who serve as apprentices, tutors, protagonists, and opponents. And as it unfolds, the sermon becomes self-referencing, requiring a certain degree of coherence, incorporating a level of redundancy, and finally arriving at certain conclusions and not others by virtue of what has already been said and how it was said. Some of these constraints are present because of the social context in which the sermon is produced, others depend on the way the sermon textualizes these elements, and still others occur as features of the internal structure of the sermon itself. All of them are likely to follow more or less familiar or established social conventions.

This suggests that we may need to rethink what it means to say that religion is institutionalized. It is institutionalized not only in the organizational realities we point to as churches, sects, cults, or denominations. And it is institutionalized not only in the subjective expectations individuals hold about what things to believe in, how strongly to believe in them, and how to act upon these beliefs. It is also institutionalized in customary manners of speaking. Religion is a kind of code — a code evident in creeds and scriptures but also in the informal ways in which discourse about these creeds and scriptures is organized. In one setting it may be necessary for discourse to include signals of uniformity or consensus — gestures of agreement such as smiles and nods of the head, more formal statements such as the recitation of common creeds, or even informal modes of demonstrating agreement such as reading from a common translation, focusing on childhood experiences likely to have been common to everyone, or incorporating code words such as amen and blessing. In another setting, it may be necessary for discourse to include signals of diversity and experimentation — sentences that begin with "I think" or "in my experience," discussions of diverse denominational backgrounds or different theological traditions, rituals of inclusion that represent symbolic variations in race, age, gender, or wealth, and ways of telling stories that attach distinct personality characteristics and needs to individual characters. These are all features of religious culture that become institutionalized as part of the discourse in which religious communities engage.

And if institutionalization becomes problematic, then the ways in which social forces shape individual religious commitments also become problematic. Differences of education, age, or denominational background may still supply clues to the social forces that shape individual religious commitment. But language games mediate between these broad features of one’s background and the level of one’s religious commitment. A way of talking about religion learned within a certain educational stratum or denominational tradition may come to define the very meaning of religious commitment. To have "really struggled" with one’s faith may be the hallmark of commitment in one setting; to have "kept the faith of our fathers" may be the proper expectation in another. Behavioral differences may result, but the language used to talk about behavior also provides cues for inclusion or exclusion in a particular religious community.

Some of this will of course already sound familiar because it is consistent with arguments put forward by Clifford Geertz in his 1966 essay, or with arguments that have surfaced more prominently in Geertz’s recent work as well as in the work of Peter Berger, Robert Bellah, and others. Whether or not something represents a "new direction" is, then, not really essential. As our discussion of Geerrz has already illustrated, the theoretical tradition is rich — but various themes selected from it resonate with the dominant concerns of particular time periods. If questions about the subjective character and functioning of religion were emphasized a quarter century ago, questions of discourse and practice are gaining increasing attention now.

Religious Culture and the New Cultural Sociology

What some have called the "new cultural sociology" is an area in which questions of discourse and practice have most clearly risen to the surface.¹¹ Work in this area has attempted to make the study of culture a more empirically grounded discipline. It has differentiated itself from grand macrotheoretical formulations concerned with value orientations, the comparison of whole civilizations, or global assertions about dominant themes in modernity such as universalism, pragmatism, and the like. Rather than thinking of culture as something implicit or taken for granted — as something about latent normative patterns that can be inferred only from observing regularities in social behavior — the new cultural sociology regards culture as something tangible, explicit, and overtly produced. It consists of texts, discourse, language, music, and the symbolic-expressive dimensions of interpersonal behavior, organizations, economic transactions, and so on. Proponents of the view hold that any specific cultural artifact should be examined in terms of questions about its production, its relations with other cultural elements, its internal structure, and the resources that determine how well it becomes institutionalized.

What are the issues, strategies, and hypotheses that may be useful to derive from the new cultural sociology for studies of religion?

One important set of issues focuses on the internal structure of culture itself. Rather than asking what symbolism does for the individual or where it is located within the individual’s consciousness, we ask how symbols themselves are arranged and related to one another. We might say that the study of religious culture shifts from a study of the meaning of symbolism to a study of the symbolism of meaning. That is to say, we take Bellah even more seriously than Bellah took himself when he argued for an approach in sociology of religion called "symbolic realism." We regard symbolism as a reality, important in its own right, worthy of systematic investigation. Consequently, we attempt to look directly at symbolism — in the case of religion, at the symbolism of meaning — rather than looking through it to see how it functions for the individual or even to give an interpretation of what sort of meaning it conveys. To be sure, we still engage in a hermeneutic process when we attempt this kind of analysis: we ultimately give our own interpretation of how symbols are put together and, in this sense, say why they convey meaning and perhaps even reveal what they mean to us. This is to say that we do not succeed in turning symbols into objects susceptible to a purely positivist form of analysis. But we recognize that the true object of analysis is symbolism itself, not the supposed meanings that subjects attach to it.

An example of how structural studies of this type might differ from more conventional studies might clarify the discussion at this point. Fundamentalism might serve, because it has been a topic of much interest in critical studies of contemporary religion. Within the more traditional framework, fundamentalism has been described as a "world-view," a rather tight (or narrow, or simplistic) view of the world — an orientation that is perhaps hierarchically organized around the ultimate value of otherworldly salvation, an orientation that supplies totally encompassing normative expectations for how people should behave, a set of beliefs and assumptions that are deeply meaningful to the people who hold them and that give meaning to these people’s lives. It has been called part of an authoritarian or paranoid cultural style. It seems to require a high level of commitment to some group or charismatic leader. It is a kind of elaborate dogmatism involving unquestioning acceptance of a vast set of scripturally legitimated injunctions and theological tenets. It is backward looking, a kind of longing quest to rediscover tradition, to make life simple, and (for this reason) it probably shares an "elective affinity" with dogmatism, bigotry, and political conservatism of other kinds.

If that is recognizably a caricature of fundamentalism, it is at least recognizable — and it is the way, I’m sure, that many fundamentalists and nonfundamentalists alike would describe it. Certainly it is a description that fits the dualism of theories that draw a sharp contrast between culture and social structure. Note how many of the words are subjective and intersubjective: worldview, orientation, value, normative expectations, beliefs, meaning, commitment, longing. It is also true that these ways of characterizing fundamentalism are at least one step removed from empirical observation. They are abstractions: they represent a thematization of fundamentalist culture, an extraction from observations, a way of summarizing what fundamentalism is supposedly all about. They represent inferences, mostly about things that cannot be observed very well directly.

Suppose, though, that we take a more structural approach. Suppose we examine carefully the texts of fundamentalists’ tracts and creeds, the texts of fundamentalist sermons, and the texts of discourse in which fundamentalists engage. Doing so — and this has been done— gives a rather different impression of fundamentalism. For instance, the structure of its discourse is not at all a tightly integrated dogmatic system; in fact, it is a hodgepodge of rather loosely coupled or even discrete statements or tenets. Often the main tenets are merely listed and then related to one another mostly in seriatim fashion rather than in terms of logical dependency. Consequently, they can be combined quite flexibly with other tenets (for example, some fundamentalists emphasize the so-called gifts of the spirit while others don’t, some expect the imminent end of the world while others don’t, etc.). Moreover, these tenets by no means encompass all of life in a kind of normatively determined worldview. They may contain admonitions about gambling and alcohol but have nothing to say about dieting, nuclear weapons, or brands of automobiles. And one could go on and on with this kind of analysis, showing that fundamentalists have a fairly predictable — and common — structure in their moral discourse, that their discourse contains certain distinctions about self, knowledge, agency, authority, community, and the like.

What is interesting about this kind of analysis is that it focuses on observable cultural materials: texts, sermons, discourse. It seldom strays into the subjective consciousness of the individual fundamentalist orinvO~ the collective unconscious of fundamentalists as a group. Rather than fundamentalist culture being a kind of implicit Zeitgeist, it consists of explicit, manifest, observable products. Moreover, the kind of analysis to which these products are subjected is not simply a process of extracting their content. The focus is not to thematize the underlying beliefs of fundamentalists from identifying common assumptions in their literature. The focus is on patterns among discursive elements, structural relations among these elements — their arrangement, the boundaries that separate them, the connections drawn between them, their variety, the degree of repetition evident, formal properties, the structure that gives them internal coherence. In short, the methodological cues are drawn from such writers as Mary Douglas, Levi-Strauss, Foucault, Bakhtin, Todorov, Bourdieu, and Derrida. We stand to learn a lot about the ways in which religious discourse is put together from this kind of analysis.

A second general set of issues concerns the ways in which social relations are dramatized symbolically and ritually. As sociologists, we are probably also interested in something more specific about the social dimension of this discourse. Thus far we have taken it largely out of context, frozen it into a static image; now we wish to let the projector roll again, we want to examine culture, as Ann Swidler has put it, "in action."¹² We want to take account of the fact that culture is not simply text, product, and object but that it is an act, that it is "enacted." Note that it is still important for us to have examined the internal structure of the discourse itself. We need to know how the discourse is arranged. But now we ask not only about its structure but also about its function. We conceive of it as that thing which it is principally intended to be: as communication. We see it as "expressive behavior," and particularly (given our interest in the social) as expressive behavior that communicates something about social relations.

Consider, for example, two fundamentalist sermons. Both reach their didactic climax by asserting that "Jesus is the answer." We note from our structural analysis that this statement is extremely brief, that the preachers do not spell out "how" Jesus is the answer or what that answer actually is. And we note that before reaching this conclusion, both ruled out a number of alternatives such as success, money, fame, love, education, and a whole variety of other things that might have been thought to be the answer. But now we put each sermon in its real-world context. We recognize that one is being preached in a local congregation of a hundred people, the other on television to a national audience. We also note that the first preacher goes on to say: "If you don’t know Jesus, if you haven’t found that answer, turn to the person next to you before you leave and find out. We want to help you. Join with us and see what it can mean to know Jesus." At the same point, the other preacher says, "Ask Jesus into your heart, right now,’ there in your living room, and you can discover the abundant life, you can experience victory over your problems, you can be happy and know real peace in your life."

Obviously, we have learned something very important about the social relations that are dramatized or expressed in these two sermons. In the one, "Jesus is the answer" is a kind of "restricted code" (to use Bernstein’s term), a phrase that is defined in interaction with a community, that is intended to draw people into a community.¹³ In the other, the same phrase is connected with individualizing words: "your living room," "you," "your life." In other words, we see that fundamentalist discourse functions as a kind of ritual — indeed, as part of a ritual — and that it dramatizes something about social relations. It tells people how to behave in relation to one another, thereby creating a kind of moral order among them. And so looking at this discourse helps us to understand those social relations — and to understand the kinds of social relations in which this discourse may be especially meaningful. In other words, we would expect certain kinds of discourse to be present in some social situations and not in other situations. All of this points to the importance of linking culture with practice or, more precisely, of viewing culture as practice, rather than conceiving of it as something static or embodied solely in texts.

Finally, any adequate sociology of religious culture must also be concerned with questions about institutions and institutionalization. We’ve focused on the dramaturgic or communicative functions of culture as action, and this emphasis may suggest something about their connections between culture and its social context. But we can also push this search for connections to a somewhat more complex level. We can also raise the obvious, but important, fact that cultural acts don’t just happen. They are produced, and like any other outcome, it takes resources to produce them. We don’t quit looking at their internal structure or their dramaturgic functions, but we now ask about the resources that affect cultural production. We make explicit what has been only implicit thus far.

In our fundamentalist example, we make explicit the fact that the discourse is being produced by a real speaker, a preacher; that it takes resources to train, house, clothe, and feed this preacher; that these resources have to be extracted from the social environment (by getting people to put money in the collection plate and by donating some of their time); and that all of this takes place not just in a kind of ad hoc way but in a highly institutionalized setting. There are fundamentalist denominations, fundamentalist seminaries, fundamentalist writers, fundamentalist broadcasting organizations, and, more generally, well-institutionalized relationships between churches and the state and among churches themselves — all of which bear importantly on the production of fundamentalist discourse.

And when we examine these institutional contexts in which cultural production takes place, we are likely to be interested in the relationships between these contexts and the content of what is produced. In other words, we probably will want to go beyond a mere organizational analysis of the numbers of books, sermons, or theater productions that are produced. We will want to give culture itself a more interactive role in this process. Certain kinds of discourse may help the organization solicit resources; at the same time, discourse may commit the organization to certain paths of action that delimit what it can do in the future. So we are not simply ignoring culture and looking at cultural organizations, as some have charged, but are having to pay attention to the complex relationships existing among producers and consumers, the discourse and other kinds of cultural products, and the larger institutional settings in which this interaction takes place.

The Uniqueness of Religious Discourse

Does any of this presuppose something unique about religion as religion? Or does it assume that religion is merely an instance of discourse more generally? That, it seems to me, should be left open as an empirical question. Modern religion is to a very considerable extent a discursive practice. It encompasses ritual, and it mandates behavior and feeling, but it gives a privileged place to discourse: it grounds itself in the Word, whether that means a formally codified text or a broader conception of the divine spirit; it thrives on professional and popular interpretations of the Word; it requires the construction and maintenance of community through communication of shared convictions and experiences; and it mandates verbal expressions of sincerity, emotion, and commitment. But it remains — along with law, science, education, philosophy, and ordinary conversation — only one of the domains of modern society in which discourse is privileged.

Nevertheless, religious discourse speaks to a distinctive realm of existence, and we may wish to look for clues about the special ways in which an analysis of religious culture may prove useful to our understanding of the religious endeavor. Geertz was correct, I believe, in suggesting that religion speaks to our desire for personal meaning and that it does so by clothing concepts of a general order of existence with a powerful and enduring sense of facticity. That is a form of discourse that differs importantly from the talk we engage in about yesterday’s ballgame or tomorrow’s excursion to the beach. It requires a level of seriousness and sincerity but also a substantive scope that exceeds most other forms of discourse. We must, as Durkheim recognized, be able to set it apart from the profane speech (and practice) of everyday life. And we must be able to do so in a way that not only speaks about but also speaks (enacts) the paradoxical nature of life itself. Religious discourse totalizes its content and yet leaves room for individual expressions of faith That is its basic dilemma — a dilemma that we can perhaps understand better as we pay closer attention to the practice of religious culture.

Endnotes

1. Geertz, "Religion as a Cultural System," in Anthropological Approaches to the Study of Religion, ed. Michael Banton (London: Tavistock, 1966), pp. 1-46; reprinted in Geertz, The Interpretation of Cultures (New York: Basic Books, 1973), pp. 87-125.
2. Parsons divided the world into four systems — physical, personality, social, and cultural, which he linked with the disciplines of natural science, psychology, sociology, and anthropology, respectively. See Parsons, The Social System (New York: Free Press, 1951); and Toward a General Theory of Action, ed. Talcott Parsons and Edward A. Shils (New York: Harper & Row, 1951).
3. See Bellah, Beyond Belief (New York: Harper & Row, 1970); and Robertson, The Sociological Interpretation of Religion (New York: Schocken, 1970).
4. See Glock and Rodney Stark, Religion and Society in Tension (Chicago: Rand McNally, 1965).
5. Ortner, "Theory in Anthropology since the Sixties," Comparative Studies in Sociology and History 26 (1984): 126-66. Religion, on the one hand, while on the other hand Parsonian theories, speculative and comparative work in the classical tradition, and some of the newer perspectives of phenomenology posed challenges to empirical positivism. Geertz’s essay was ripe for appropriation by both sides in this struggle because it reflected the broader Parsonian and hermeneutical orientations and yet provided a concise definition of religion and a discussion of some of its empirical manifestations.
6. See Robert Wuthnow, Meaning and Moral Order: Explorations in Cultural Analysis (Berkeley and Los Angeles: University of California Press, 1987).
7. See, e.g., George M. Thomas, John W. Meyer, Francisco 0. Ramirez, and John Boli, Institutional Structure: Constituting State, Society, and the Individual (Beverly Hills: Sage, 1987).
8. See, e.g., Richard Sennett, The Fall of Public Man: On the Social Psychology of Capitalism (New York: Random House, 1976).
9. Berger, The Sacred Canopy: Elements of a Sociological Theory of Religion (Garden City, N.Y.: Doubleday, 1967).
10. Austin, How to Do Things with Words (Cambridge: Harvard University Press, 1975).
11. For an overview of the new cultural sociology, see Robert Wuthnow and Marsha Witten, "New Directions in the Study of Culture," Annual Review of Sociology 14 (1988): 111-33.
12. Swidler, "Culture in Action: Symbols and Strategies," American Sociological Review 51(1986): 273-86.
13. Basil Bernstein, Class, Codes and Control (New York: Schocken, 1975).