Encyclopedia of Religion
and Society

William H. Swatos, Jr. Editor

Table of Contents | Cover Page  |  Editors  |  Contributors  |  Introduction  |  Web Version


The idea of religion is perhaps inseparable from a distinctive cognitive, or belief, dimension. Yet the social scientific study of religious belief has long struggled with an adequate definition of this dimension. There are two basic, long-standing schools of thought. One, the substantivist, argues that religion is fundamentally a belief in supernatural agents or forces (Goody 1961). This conceptualization has a strong intuitive appeal to social scientists, who for the most part are directly influenced by Judeo-Christian traditions, but there are drawbacks to the use of this definition in comparative studies.

In any religious culture where there exists a set of beliefs clearly demarcating a supernatural order from the natural world, little difficulty occurs in applying the term religious to beliefs about the supernatural. Christianity and Islam, objectified as they are in dogma that denigrates our present, mundane existence and looks forward to a transcendent, spiritual existence after death, yield familiar examples. But in some religious traditions, the dividing line between the natural world and the supernatural is not plainly distinguished by believers. In the traditional culture of Bali, for example, every venue of human worldly activity is tinged with spiritual understandings. Therefore, in religious cultures such as Bali's, belief in the supernatural blends with naturalistic beliefs in a smooth continuum so that the two cannot be separated without distorting the understandings of the believers.

The second school of thought, the functionalist, defines religious belief to consist of understandings about a transcendental reality, the experience of which is an ultimate concern to human beings (Luckmann 1991). This perspective may owe its origins to Plato, who insisted that we ought not be beguiled by or place our trust in the world of appearances but instead should seek that which is most true and permanent that lies hidden beyond. Although gods and spirits—seen as inhabitants of the transcendental realm—can be accommodated in this approach, the supernatural premise is not deemed essential.

This second definition of religious belief easily stretches to incorporate antidogmatic mysticisms that make a virtue of being ambiguous about the supernatural. Philosophical Buddhism, which is rather atheistic, is also easily encompassed using this perspective. The second approach defines religious belief so broadly that scientific humanism, Marxism, nationalism, and psychotherapy are all conceivably religious because they deal with deep, nonobvious realities that are an ultimate concern to at least some people.

Clearly, each school has its own strengths and weaknesses. Social scientists who advocate the supernatural premise usually sidestep the difficulty of separating the natural from the supernatural by confining their research to religious cultures where the distinction is unambiguous. When required to draw comparisons to religious cultures where the distinction is blurred, they generally do this by adopting a secularization argument. In traditional societies, religion is a "sacred canopy" (Berger 1967) that constructs and maintains an entire "world" of social experience. The natural and the supernatural meld seamlessly. As societies approach the modern, all features of social life, including religion, become more specialized and differentiated. Religious belief becomes dogma, a codified creed under the care of a professionalized corps of religious specialists. Under the influence of religious specialists, the plethora of spirits and deities is systematized into a hierarchy with a supreme god at its head. Eventually, religious professionals may declare the lesser deities and spirits to be alternative manifestations of the one deity, or they may expunge them altogether as superstitions or heterodox beliefs. Once an abstract and hierarchical conception of supernatural agents has become dogma, the separation in believers' minds between the natural world and the supernatural has been accomplished. Then, as society modernizes, even more belief in supernatural agents and forces atrophies. Increasing human control of nature as well as the spread of scientific understanding of natural processes obviates the need and desire to believe in the supernatural.

By defining religion as a belief in a transcendental reality that is of ultimate human concern and thereby diminishing the importance of the supernatural premise, advocates of the second school escape one dilemma but must cope with an embarrassment of riches in contemporary societies where traditional religions, upstart cults, civil creeds, quasi-religions, and secular "faiths" coexist in profusion while individual believers eclectically construct their personal belief systems drawing from these numerous sources. Perhaps the real strength of the second definition is that it leads to an exploration of the functional substitutes for religious belief (Luckmann 1967, 1991).

The Origin of Religious Beliefs

During the past century and a half, there have been various attempts by social scientists to explain the origin of religious beliefs. The anthropologist Edward B. Tylor (1873) represents an early attempt. To him, belief in spirit beings, or animism, constitutes the minimum definition of religion. Tylor went on to speculate that animism developed from early humanity's attempts to explain puzzling experiences. First, he finds the concept of the soul to be universal in simple societies; then he asks how people came to create the concept of the soul, and finds the answer in people's attempts to explain dreams, hallucinations, and other psychic experiences that puzzle them. From this, it is only a reasonable step that early humanity would extend the idea of the soul to other animals, and then to plants. Continuing to reason by analogy, Tylor says, early humanity would in due course come to attribute spiritual qualities to stones, water sources, weapons, food, and ornaments.

Sigmund Freud (1961) proposed an equally distinctive theory of the origin of religious belief. According to him, the religious experience was most like the experience of complete dependency that an infant feels for the parent. A belief in supernatural beings was therefore an adult projection of infantile dependency and should be interpreted as a desire to return to the security of early childhood. Freud concluded that religious beliefs are collective neuroses that save people from developing more severe individual neuroses.

…mile Durkheim (1965) argued that because all human beings are creatures of society and dependent upon it, religious beliefs should therefore be interpreted as "collective representations" that express this dependent relationship between the individual and the collectivity. When we worship the gods or other sacred symbols, we are actually worshiping our social collectivity, whether we realize this or not. Durkheim's theory entailed an important innovation because he considered religiosity to be an attribute of the collective experience rather than a product of individual intellectual speculation or wish fulfillment. As long as human beings remained social creatures, they would have need of religious beliefs and representations, but the character of these beliefs and representations would change radically under the influence of science.

A more recent effort to explain the origin of religious beliefs was made by Talcott Parsons (1972), whose explanation was grounded in certain social psychological universals of the human condition. In addition to setting forth a definition of religion as a universal feature of human society, Parsons stated that religion consisted of a more or less integrated set of beliefs concerning entities that are supernatural or sacred, and thus set apart from the ordinary objects and events that have utilitarian or instrumental importance. Further, Parsons declared that meaning in human life is fundamentally dependent on religious belief because of the universal presence of frustrating experiences that humans cannot avoid.

According to Parsons, two main types of frustration in the human situation provide the focal points for the development of religious beliefs. One of these is that people are "hit" by events that they cannot foresee, prepare for, or control, such as the occurrence of premature death. The second type is present where there is a strong emotional investment in accomplishing some goal, yet despite the greatest energy and skill brought to bear in this effort, success remains uncertain. Frustrations arising from the discrepancy between expectations and what actually occurs pose "problems of meaning," in the sense that Max Weber wrote about. That is, we can explain why an automobile accident caused a premature death, but we cannot explain why it had to happen to a person we loved. Why do the innocent suffer while "the wicked flourish like a green bay tree"? We have no practical answers to such dilemmas. Hence the significance of religious belief is that it is made up of those aspects of the life situation to which people cannot remain indifferent, which they cannot in the long run evade, but which they cannot control or adjust to with every practical means available.

The difficulty with all such attempts to explain the origins of religious beliefs is that the results are speculative and can never be adequately tested. Hence, in recent decades, social scientists have been less motivated to explain the origin of religious belief. Instead, more attention has been devoted to elucidating how religious beliefs articulate with societies and social groups. Social scientists have attempted to grapple with the social consequences of religious beliefs—to see them either as social cement or as social control (Thompson 1986). The first perspective derives from Durkheim, the second from Karl Marx and Friedrich Engels. Yet a third perspective has been to interpret religious beliefs as cultural models. This approach derives partly from Durkheim and Weber via Parsons, and partly from cultural anthropology.

Religious Beliefs as Social Cement or Social Control

The idea that religious beliefs can provide an integrating focus for relatively homogeneous collectivities and, under the right circumstances, even for complex societies is indisputably one of the main findings of twentieth-century social science. The entire corpus of cultural anthropological studies of simple societies demonstrates the first point. The second point has been the contention of the civil religion thesis. Civil religion is the expression of the cohesion of the society or nation. Sacred beliefs transcend ethnic, denominational, geographic, and class distinctions. In the United States, for example, historical figures (George Washington, Abraham Lincoln, Martin Luther King, Jr.) have been elevated to a quasi-mythical status and are venerated as champions of American ideals and values. Although there have been many debates over the civil religion thesis, it does appear to grasp a reality: Many Americans regardless of their background do say that they feel drawn together by these unifying sacred beliefs.

Noting that religious beliefs may serve as social cement is only half the picture, however. It is also useful to consider religious ideas as weapons in an ongoing struggle within society between social classes or between other groups with divergent interests (e.g., children versus adults, women versus men, immigrants versus long-established residents). Religious beliefs may be seen to serve the ends of one group or section of society attempting to control another. One kind of social control engendered by religious beliefs falls under the rubric of the dominant ideology thesis, originated by Marx and Engels (1970). In their view, the dominant faction in society not only controls the production of wealth, it also controls the production of sacred beliefs. Thus orthodox religious dogmas are supported by the dominant class because they legitimate its lifestyle and roles in society.

However, the social control function of religious beliefs is usually not so transparent. It can refer to the efforts of a dominant class or group (1) to legitimate itself in its own eyes and/or (2) to suppress—without resorting to threats and physical coercion—rebellious behavior by the subordinate class or group. In the classic Marxian formulation, (2) is accomplished because religion is "the opium of the people." On the other hand, social control can have a contrary meaning: the use of religious beliefs by the weak to constrain or modify the behavior of power holders. Thus religious belief is potentially one of the "weapons of the weak" employed to obtain concessions or favors from those with privilege and power. It is also conceivable that religious beliefs constitute an arena in which social control is attempted in both directions. The weak attempt to constrain the behavior of the strong by appealing to religious precepts, while the strong legitimate their privileged positions in public display of their conformity with these precepts. Each party has an interest in ascribing to and supporting the beliefs, but for different reasons. The situation resembles a positive-sum game (Reeves 1995).

Religious Beliefs as Cultural Models

…mile Durkheim was an early proponent of the idea that religious beliefs are cultural models. This point was explicated in a seminal article written by Talcott Parsons (1978) in which Durkheim's classic, The Elementary Forms of the Religious Life , was reassessed. According to Parsons, Durkheim was well in advance of his contemporaries when he grasped that religious beliefs embody a cultural code for social action. Operating in a manner that is analogous to the genetic code in the organic sphere and to syntax in language, the cultural code made up of religious beliefs allows variation to occur within a predetermined structure of action. Stated differently, the possibility of variation is built into the expression of the code in concrete situations, just as a language speaker is partially constrained by grammar in forming statements but is free nevertheless to express an endless variety of ideas.

Parsons's student, the anthropologist Clifford Geertz (1973), wrote the manifesto for this perspective. Geertz made an enormous impact on the social sciences when he argued that, unlike other animal species, Homo sapiens is not constitutionally whole without extragenetic, cultural models to channel perceptions, behavior, beliefs, and evaluations. The peculiar, incomplete nature of human biology necessitates the existence of a complex system of meanings represented in symbols to give human life direction and purpose. Because of their transcendence, religious beliefs objectified as symbols are often the most comprehensive and profound cultural models. At present, this perspective probably remains the dominant approach to religious beliefs in the social sciences (see Greeley 1995).

The historian Bernard Lewis's brief book, The Political Language of Islam (1988), provides an excellent example of what is possible with a cultural model perspective. Referring to Arabic, Persian, and Turkish sources, Lewis shows how political theory and practice in the Muslim countries were informed by Islamic beliefs and encapsulated in Islamic terminology. In the United States, the "civil religion" thesis also has been strongly influenced by the cultural model concept. India has long been a favorite subject for scholars interested in how religious beliefs model extreme social inequality in the caste system (Dumont 1980, Milner 1994).

Edward B. Reeves


R. N. Bellah, "Civil Religion in America," Daedalus 96 (1967):1-21

P. L. Berger, The Sacred Canopy (Garden City, N.Y.: Doubleday, 1967)

L. Dumont, Homo Hierarchicus (Chicago: University of Chicago Press, 1980)

…. Durkheim, The Elementary Forms of the Religious Life (New York: Free Press, 1965 [1912])

S. Freud, The Future of an Illusion (New York: Norton, 1961 [1927])

C. Geertz, The Interpretation of Cultures (New York: Basic Books, 1973)

J. Goody, "Religion and Ritual," British Journal of Sociology 12(1961):142-164

A. M. Greeley, Religion as Poetry (New Brunswick, N.J.: Transaction, 1995)

B. Lewis, The Political Language of Islam (Chicago: University of Chicago Press, 1988)

T. Luckmann, The Invisible Religion (New York: Macmillan, 1967)

T. Luckmann, "The New and the Old in Religion," in Social Theory for a Changing Society , ed. P. Bourdieu and J. S. Coleman (Boulder, Colo.: Westview, 1991): 167-182

K. Marx and F. Engels, The German Ideology (London: Lawrence & Wishart, 1970 [c. 1845])

M. Milner, Jr., Status and Sacredness (New York: Oxford University Press, 1994)

T. Parsons, "Religious Perspectives in Sociology and Social Psychology," in W. A. Lessa and E. Z. Vogt, Reader in Comparative Religion , 3rd ed. (New York: Harper, 1972): 88-93

T. Parsons, "Durkheim on Religion Revisited," in Action Theory and the Human Condition (New York: Free Press, 1978): 213-232

E. B. Reeves, "Power, Resistance, and the Cult of Muslim Saints in a Northern Egyptian Town," American Ethnologist 22(1995):306-323

K. Thompson, Beliefs and Ideology (London: Tavistock, 1986)

E. B. Tylor, Primitive Culture , 2nd ed. (London: Murray, 1873).

return to Encyclopedia Table of Contents

Hartford Institute for Religion Research   hirr@hartsem.edu
Hartford Seminary, 77 Sherman Street, Hartford, CT 06105  860-509-9500