(Bibliographic refererence: Nelson, John K. 1990. The Anthropology of Religion. A Field Statement for the Department of Anthropology, University of California, Berkeley)

Introduction
Almost as old as the field of anthropology itself are anthropological attempts to analyze what has been thought to be a genuine cultural universal: religion. Coming in the wake of theological interpretations (as revealing certain "truths" about the nature of the cosmos) and nearly simultaneous with sociological explanations (focusing on social structure and how various religious institutions have social functions), anthropology has taken as its task to detail the customs, belief systems, myths, and power relations for particular peoples and societies as they both structure and are structured around religious orientations to the world. Ioan Lewis (1982) has compared the roles anthropologists have played in their own societies with the religious specialists they have studied: like the shaman, anthropologists have mediated between their own groups and those exotic other worlds in an effort to bring back "wisdom".

This field statement follows in that tradition to a certain degree, attempting to compile the efforts of a multitude of researchers covering nearly a full century and bring, to one place, a summation of their approaches. And yet it breaks with the tendency to categorize and separate the religious traditions of societies under investigation into neat compartments of analysis called "magic", "witchcraft", "possession", "ritual" or "belief" (see Lehmann and Myers 1989, Wallace 1966, or Lesser and Vogt for examples, as well as Morris 1987:2-3 for the effects of this tendency). Because of theoretical developments in anthropology as well as in philosophy, history, and other disciplines of the humanities during the last thirty years, what were once seen as boundaries between distinct or isolated elements of a socio-religious system (magic vs. religion, for example) have been rendered instead as linkages between sets of meaning that dynamically interact with each other. Where categories and concepts were once used "to formulate an underlying uniformity behind diverse phenomena" (Geertz 1968:24) interpretive anthropology / sociology, hermeneutics, deconstruction theory, semiotics, neo-Marxist critique, and even later developments in psychological and psychoanalytical analysis have thickened the plot considerably. Anthropologists involved in cross-cultural comparisons have slowly learned to exercise caution when employing concepts whose roots lie in the Western tradition to make sense of practices originating outside that tradition. Criticising anthropological studies of witchcraft, for example, Crick suggests that "great violence must be done to the conceptual structures of another culture in speaking about witchcraft if it lacks the environing categories which defined it in our own (1976: 112, cited in Caplan 1983).

Perhaps the most striking effect of these theoretical developments is on what was once a distinct field of study in anthropology. As reflected by the title of this field statement, anthropological studies of religion can not be separated from the study of power, ideology, or semiotics any more than heart surgery can proceed without a knowledge of the interrelation of the body's circulatory, nervous, and muscular structures. Thanks in part to the education I have received here at Berkeley from numerous books, articles, professors, and colleagues, the topical areas into which this field statement is divided reflects both a selective history of the field as well as a subjective approach towards achieving an orientation to highly complex and intricately interwoven areas of social life that will be called, for the sake of convenience, "religion."

Thematic Contents
I will first briefly sketch the historical development of general and comparative theories of religion then move on to the ritual process. By using the word "process" and changing "ritual" from an all-inclusive noun to an adjective, I hope to broaden what might be considered ritual without emptying the term of its semantic usefulness. Again beginning with historical and contemporary theories, the discussion here will attempt to integrate a sampling of semiotic theories and studies as to why and on which levels these events are meaningful. A survey of the literature on ideology, empowerment, and social change via secularization follows as an extension of the discussion on ritual processes and the symbolism they employ. A final objective is to state a working model for interpreting not only religious attitudes towards experience (ethos and world view) but the sorts of social institutions which support those attitudes.

Honoring the Ancestors: Founding Fathers of the Anthropology of Religion

Depending on who one reads (and the theoretical and political inclinations of these authors), the anthropological study of religion begins in either early 18th century Germany or France. Morris (1987), for example, holds that to understand later theorists such as Tylor, Durkheim, or Malinowski one must begin with the intellectual tradition started by Hegel who, among his other tasks, attempted in the 1790's to "explore and integrate the irrational into an expanded reason" and thus argued for a tolerance of all religious traditions (Morris 1987: 5). Ludwig Feuerbach, one of Hegel's disciples, tried to limit the theology of his period by what we would now call an "anthropological" grounding in specific social contexts and thus had a great deal of influence on the next member of the lineage, Karl Marx. (Attention will be given to Marx's ideas as well as his critics in Part III of this paper). The last member of this lineage is, of course, Max Weber. Influenced more by Kantian non-materialistic philosophy than by Hegel, he wrote as much in reaction to his predecessors as he did in praise of them. (For excellent discussions of Weber and his ideas see Andreski 1984, Wrong 1970, Marshall 1982, as well as Morris 1987.)

The other moiety of early anthropologists might claim as their founder the French writer Montesquieu. He initiated what Evans-Pritchard calls a "pragmatic" way of regarding religion (1965: 49) by writing in 1750 that even though a religion may be thought to be false, it can serve a useful social function as long as it conforms to the type of government with which it is associated. Another influence on later theorists such as Durkheim and Spencer was Fustel de Coulanges, a French historian who insisted in his 1864 work The Ancient City that religious ideas were not only the cause of social changes but the essence of all social phenomena (cited in Morris 1987: 112).

Moving to England, we first encounter Herbert Spencer, (writing about the same time as Charles Darwin) who systematically posited the idea of society as an "organism" in works such as The Principles of Sociology (1876). Yet perhaps the honor of the first real anthropologist to write about religion and society should go to Edward Tylor for actually travelling to the places he would later write about (Mexico and North Africa). He is primarily remembered for coining the term animism , for distinguishing between "spirit" and "soul", and for his theory of magic, science, and religion (Primitive Culture, 1871) which, if we are to believe Evans-Pritchard (1965: 29), Sir James Frazer tried to systematize in his massive work The Golden Bough (1922). Robertson-Smith brought forth the notion of totemism in his study of the Semitic societies of ancient Arabia (1889) and asserted (at the cost of his professorship at Oxford) that rituals had primacy over beliefs because of the way they bound together members of a community.

All of these writers provided the intellectual heritage for the seminal theories of Emile Durkheim to emerge. It would be difficult to summarize the enduring contribution and importance his The Elementary Forms of the Religious Life (1915) has had in stimulating research on the interrelations between society and its collectively held symbols--so I will mention some of the more poignant critiques of this work as a way of addressing theoretical advances in the field.

Goldenweiser was one of the first to respond in 1917 to Durkheim's notions of the sacred and profane, the totemic principle, and the bases of religion for a society (Goldenweiser 1964 [1917]). Complaining that sacred objects can not be interpreted through one general principle called "the sacred" and that it is shaky to rest a definition of religion on this principle, Goldenweiser also pointed out that there is more than one context shaping Durkheim's "totemic principle", namely, that the classifications are based on human social divisions, not natural ones as claimed. Indeed, to follow Evans-Pritchard, that Australian totemism is the original form of religion seems to arbitrarily rest on the assumption that people with the simplest culture and social organization must have the simplest religion (Evans-Pritchard 1967: 66). Levi-Strauss also used this distinction between natural and social classifications in his theories of totemism (1963), seeing these it as an aspect of a mode of thought that mediates between the two (see Morris 1987: 275-288).

Durkheim's functionalism--that religious rituals are the mechanisms for expressing and reinforcing the solidarity of the group so that "religion is society worshipping itself" (Durkheim 1963: 226)--has been generally accepted for several decades, but even here problems emerge in the way he uses society as a "homogeneous entity" (Morris 1987:122) without recognizing internal divisions according to sex, race, class, or ethnic identity. Giddens (1978: 130) has mentioned another aspect of this generalizing tendency, in that Durkheim never addressed the possibility of religious beliefs frequently having an ideological function. Far more than a generalized reflection of homogeneous society, religion is all too often a tool of dominant or privileged classes to legitimate their power and specific interests. Bernard Lacroix (1979) defends Durkheim against these criticisms by trying to read behind the lines and find an implicit theory regarding constraint ("an effect of the pre-existence of symbolic systems" [89]), power ("it constantly adapts to the milieu it governs"), and social differentiation, but he merely draws upon recent French structuralist theory in his labors without conclusive evidence from Durkheim's text itself.

Despite these objections, to talk about "civil religion", "group solidarity," or concepts of "the sacred" one simply cannot avoid the substantial contributions of Durkheim to anthropology in general and the study of religion in particular. One can agree with Durkheim's emphasis on the fact that social solidarity structures all religious experiences, and yet differ with the way he treats religion as a function without reference to meaning (or the way that meaning is purposefully constructed).

Similar discussions, (necessarily truncated) regarding Weber, Malinowski, Radcliffe-Brown, Lowie, Levy-Bruhl, Radin, or Levi-Strauss could occupy the next twenty pages of this paper. A more efficient use of time and space, however, is to refer to those works which perform the same task in a more readable and complete manner. In addition to Morris (1987) and Evans-Pritchard (1965) already mentioned, succinct summaries of the early theorists can be found in Giddens (1978), Pickering (1984), Schoffeleers (1978), and Bellah (1973) on Durkheim's sociology of religion, Wrong (1970), Marshall (1982), Nelson (1973), and Leuthy (1970) on Weber's protestant ethic, Parsons (1963), Lesssa and Vogt (1965), Wallace (1966), Bryan Turner (1983), Bottomore and Mulkay (1984) on other theorists to name but a few of the more useful explanatory texts.

One of the enduring conundrums scholars feel obliged to grapple with is to define what makes a "religion." But, in spite of the challenge posed by this noble goal, two problems are apparent. First, if you create a universal, functional definition, (such as Geertz' classic 1973 attempt positing "moods and motivations") then every society has a religion. The possibility of a non-religious society or individual is precluded because, even if secular, there are still systems of meaning operating with common symbols, rituals, and rationales that can be considered "ultimate" (people can't help but carry them around in their heads) and that foster social consensus and solidarity. The second kind of definition, that of a more substantive nature, focuses on religion as having an official claim repeated on a regular basis as a system or as a body of rights giving substance, structure, and release to the specific "spiritual" concerns of its adherents. Bernard McGrane, in his book Beyond Anthropology (1989), pinpoints this tradition as having begun in the Enlightenment, where "religion" became a concept detached from Christianity and was applicable to any number of pagan practices (McGrane 1989: 56), all of which were to be discriminated and guarded against lest there be incursion into Christian spiritual and physical territories.

And yet, if we have learned anything from the earlier theories, we can see that religion is not something that exists in itself but relates more to certain overlapping processes such as the legitimation of authority, the degree which resistance to that authority is tolerated (or exercised), and individual, class, or ethnic interests. Though what follows is an oversimplified example, it is not important that each member of a society "believe" a substantive doctrine because, regardless of their active or conscious participation, there is still one hegemonic symbol system to which they position not only their fundamental notions about the world but whatever challenges they make to this old order (see Laitin 1987). In the United States, for example, a majority of the population thinks within a conceptual framework originating in Christian-informed constitutional law, and yet few people actively participate in sustaining the symbol systems of that hegemony.

However, an understanding of religion based upon the second kind of definition, one common to most Christians, would dismiss a majority of the "religions" of mankind throughout history. Gustavo Benavides (1988: 6) writes in the introduction to the volume Religion and Power that religions are neither mirrors of reality nor static models of it, but ideological fields composed of multiple models and mirrors in constant competition. The opinion is echoed by Huston Smith--"religion is about power quite as much as is politics"--who then goes on to address the issue of how "religious" belief sustains communal as well as political arrangements in the Soviet Union, China, South America, India and the United States (Smith 1987) despite claims that religion has become "irrelevant."

The same impetus driving many of the early scholars--namely, what are the origins of religion?--and fueling their theoretical pursuits for an all-inclusive definition is still a major part of the literature on religious activity and attest to the remarkable continuity of the debate. Anthropologists have been comfortable in relying on economic, material, class-interest, or ideological explanations to answer this question as it pertains to specific societies but for scholars in the fields of religious studies, folklore, or the classics, the question has a biting persistence.

One of the most articulate of these scholars is the late Joseph Campbell. He believed, for instance (1989), that not only is human ritual activity at least 200,000 years old,[1] these paleolithic practices--in the form of the hunt, initiation, and the guidance of these activities by shamans--extended in an arc reaching from Europe to North America and existed well into the 20th century on the American Great Plains. Campbell sees North America as a zone of diffusion to which myths and rituals were slowly transferred from Eurasia. Although he has as his agenda a global networking of mythological themes and often seems to select only those examples corroborating his "Great Arc" hypothesis (much like Levi-Strauss did in a structural manner), the themes he invokes (Woman as Goddess, Oedipal conflict, defense and surrender, shamanic authority, the syncretistic assimilation of myths) and questions he raises (Has western civilization assimilated the old myths into its political and religious institutions? Have human beings resolved their ambivalence towards Death through ritual practices? How have shamans been transmuted into less-recognizable social roles?) can no longer be dismissed in the manner of Evans-Pritchard who wrote, regarding "intellectualist" theories, "these situations could have arisen in the way described, but there is no evidence to prove that they did" (Evans-Pritchard 1965: 25).

Anthropologists can address these recurring concerns by examining the sets of circumstances of those events responsible for "locking in" the narrative of an established myth (Cohen 1969: 350). One need only look at the rise of religious cults, the "New Age" spirituality movement, fundamentalism, and the millions of people actively involved in these phenomena to ascertain the importance of explaining their processes (see Part III). What is socially and psychologically significant is still finding its way into these systems of belief, and, as Cohen points out for myths, both deal simultaneously with what is perceived and available and linking it up to a deeper, more primordial sense of reality (Cohen 1969: 349, see also Gordon 1981, Sahlins 1981). As we will see later in discussing ideology, the myths a people believe about themselves simultaneously provide and block off explanations about social practices, cognitive schemes, and legitimate sources of authority and power. It makes a great deal of difference, according to Kenneth Burke (cited in Geertz 1968: 3), whether you believe life to be a dream, a pilgrimage, a labyrinth, or a carnival. But while historians of religion may focus on these metaphors alone, the anthropologist must look deeper into the social institutions associated with supporting those ideas as well as the way in which they are rendered accessible to those who use them (see Girardot [1983] on "chaos" as a Tibetan mythological construct, Pilgrim [1986] on the Japanese concept of "interval", Parker [1983] on myths of purity and pollution for the Greeks or, on the same themes in Japan, Earhart [1984] and Ross [1965]).

The Ritual Process
In her stimulating book Sherpas Through Their Rituals , Sherry Ortner believes that while the standard categories of kinship, economy, politics, or religion may help in one's approach to ethnography, they do not provide an experience of the interconnections between these categories (Ortner 1978: 1). Only in performances--selected by the people themselves (and not the observer) as embodiments of their culture--can a researcher begin to experience how the fundamental assumptions of a people are constructed and reestablished.

The variety and breadth of the literature on ritual is staggering, even if one attempts to confine one's reading to anthropological studies. References extend to other disciplines and theories occasionally interweave at dizzying speeds. However, if I may be somewhat reductionistic about my reading thus far, it appears that there are three major approaches which appear in the literature regarding the ritual process: psychoanalytic theory, transition-rite theory, and structural-functional theory. I'll take a moment to briefly discuss what might fall into each category (building on the discussion by Paige and Paige [1980]).

The first, based primarily on Freud (1913) and subsequent interpreters of his theories, sees ritual practice as one of the many expressions of conflicts within the psychosexual personalities of the members of a society. A successful attempt at convincingly utilizing this perspective is Stanley Brandes' 1980 study of the metaphors employed by Andalusian farmers in Spain, whose folklore, colloquial speech, and sacred and secular rituals show the insecurity men feel regarding their place in the social hierarchy as well as having their masculinity undermined by women (see especially Brandes 1980: 99-114.) For other scholarly works in this vein, the helpful bibliography in Paige and Paige (1980) provides useful direction.

The transition-rite theory, based on the model by Arnold Van Gennep (1965 [1909]), holds ritual as the process that reinforces a society's age and sex role structure by dramatizing individuals' transitions to these roles. The literature abounds with hundreds of studies on this theme, such as Audrey Richards' classic Chisungu (1965) about the initiation of Bemba young women in tribal Zambia, Beattie's studies on the Nyoro (1969, 1971), Lewis' study of circumcision in New Guinea (1980) or Raphael's 1988 book about "unstated" rituals of male initiation in the United States. Even though Van Gennep's model does not suggest why initiation rites dominate some societies and not others, it has provided a pragmatic place to begin inquiry into many ritual practices.

The final category has articulate origins in the works of Durkheim (1915), where, to oversimplify again, ritual reinforces a collective view of society and increases group solidarity. While a majority of studies are content to parrot this conclusion after long ethnographic accounts, especially important here is Durkheim's notion (and to some extent Weber's as well) that ritual offsets the structural strains in society caused by the competing interests of clan or lineage groups'. Richards (1965) was among the first to identify in her analysis the explicit political implications these antinomies have for group cohesiveness and that, in the case of the Bemba, ritual was not an alternative to politics but a continuation of it by other means (Paige and Paige 1980: 43). More recently, Kertzer (1988) has established the linkages between political action and the ritualized but secular, symbolic acts it largely consists of in modern societies. Rephrasing Durkheim, Kertzer emphasizes the ability of ritual to express social dependence without necessarily dealing with supernatural beings (p. 9).

Definitions of Ritual

A field statement on the anthropology of religion and ritual could hardly avoid mentioning an evolutionary progression in the definitions of ritual. Briefly, here are a few of the most frequently cited definitions--although, as mentioned earlier in the discussion on general ideas about religion, we must remember that we are only talking about a category of analysis, not an object to be observed. A further cautionary reminder is to guard against saying what ritual practices or symbols a people "believe" in, because often this simple concept tends to arbitrarily "bracket off" ideas that people hold about the world from the way they actually think about the world itself (see Ruel's excellent 1982 article for a full discussion). [2]

Some of the ideas about ritual posited by Durkheim, Mauss, and Radcliffe-Brown, held it to be an essentially non-rational, mystical and non-utilitarian activity by which the relationship of "means" to "ends" is always symbolic (Leach 1968a:5). It could be an effective way of ensuring social solidarity and creating attitudes of respect for authoritative or communal symbolic values further enhancing the group's cohesiveness, but the people themselves didn't really know what was being communicated and expressed (Goody 1961: 155). Contemporary researchers still flirt with ideas about the abyss separating the participant from his actions, calling ritual practices "verbally irretrievable", not a representation of the mundane world, and so clogged by the simultaneous clustering of symbols that the ignorant participant mimics actions and formulas without understanding their content (Aijmer 1987:13; see also Staal's contentious "The Meaningless of Ritual" [1980] as well as Penner's 1983 rebuke).

Malinowski, however, tried to avoid this assumption by distinguishing between rituals that were "magical" (having practical value) and "religious" (having expressive value). Yet his assigning an opposition between expressive and empirical ritual ends seems an imposition from ethnocentric western models on societies that are likely to see no separation at all, let alone conflict (Nadel, cited in Goody 1961: 155). In Japanese society for example, acts appearing to be merely technical, such as swordsmithing, pottery making, or weaving, are frequently loaded with culture-specific, symbolic, or non-rational values. Ritual does appear to mark itself off from the customary performance of technical acts by a preponderance of symbolic action, but what is arrived at is not an opposition to practical activity but rather "a continuum of action stretching from the purely technical to the purely symbolic" (La Fontaine 1972: 161). Ritual processes are, after all, a way of interweaving a variety of social institutions with the surrounding natural world, making them seem facts of life, similar to the seasons (La Fontaine 1985: 33) and, to those who follow them, just as inevitable.

While popular definitions include any repetitive action as "ritual", it should be thought of here as separate from merely conventional behavior. Instead, ritual is social action following a pattern deemed morally correct, requiring a leader to direct the organized cooperation of individuals in order for the ritual to achieve its purpose (La Fontaine 1985:11). The relationship between means and ends may not be intrinsic, but it is through the transmission of information, not matter and energy, that ritual achieves real effects for its participants (Rappaport 1979:187). This approach holds for situations secular in nature (such as the "patriotic" tone of a 4th of July speech, 21-gun salute, and fireworks) as well as those more apparently religious in scope, but it should be remembered that germane to both is a concern with intentionality. Sometimes expressive and other times instrumental, a ritual cannot be seen only for its symbolic objects and overtones but must constantly be factoring in a wide range of purposes that are both symbolic and mundane, sacred and secular, as well as religious, social, and, as will be developed below, concerned with power at a multitude of levels (see Ahern 1979, Bloch 1974, 1987, Firth 1973,1981, and Sperber 1975). As such, it is as much a cumulative process as it is a discrete event.

For Radcliffe-Brown (1945), the primary effects of ritual processes are to maintain the continuity of social structures. And yet, as Leach points out (1968b), positing such a logical consistency for all ritual processes in all societies leads to a misconception about the way human conceptual processes work, as in the case of honoring a sitting King by those in attendance standing, even though cultural categories of "high" and "low" might lead to a supposition that those of higher authority will logically be placed in superior physical positions. Leach brings to the discussion a more dynamic outlook, counting ritual practices as equivalent to cultural categories by which "messages" are communicated (1954, 1979) and through which the "values and structures of a contradictory world may be addressed and manipulated" (Comaroff 1985: 196). Even though the concordant character of the natural world is held to be fundamental for most small-scale societies, there is still the obvious recognition in ritual practices of purity and impurity, good fortune and bad, fertility or sterility, weakness and its transformation into the power to fulfill obligations and achieve goals--all valid concerns of day-to-day human social interaction. Purity, for example, becomes contaminated with impurity in the most casual and accidental ways (Leach, 1968a)--such as the chance witnessing or contact with blood, disease, deformity, disasters, omens or death. But like some kind of alchemy, ritual practices frequently serve as operations for changing the potency from one side of an equation to the other by precisely defined conduct.

Victor Turner follows Leach's basic premise concerning communication, then tries to establish that the experience of an individual in ritual actually transcends cultural categories and social structures (1969), becoming "anti-structure" through which structural conflicts are muted and resolved. Turner has provided one of the most useful of all categorizations of the ritual process in seeing intimate correspondences to seasonal matters (such as planting and harvest), contingency (life crises and either the avoidance or exorcism of affliction), divination (as in determining when to plant crops), initiations (separate from life crises because a secret knowledge of a select group is imparted), defense (of the body, of property, of communal or national well-being), and acknowledgement (of ancestral spirits, of good fortune in general or that of a petition made to the deities). What this means for the participants of a ritual is not a fixed expectation about what must happen because of the nature of the event. Instead, through repetition, variation, and the contrast of symbols and themes, they learn in what cultural settings and with what degree of intensity the predominant themes should apply (Turner 1973:1104).

And yet, for all of the scholars mentioned, the ritual process is regarded as "representational behavior" which is identifiable by an observer in every culture as if it is only a kind of "coded action" standing for something else (Asad 1988: 77). Abner Cohen was one of the first to identify the fact that religious symbols "articulate an endless array of informal political groupings whose operation is a fundamental part of the total political structure of a society" (1969: 218). In the same article, attempting to open up a field for "political anthropology", he proposed as a central theoretical interest the study of the involvement of symbols with relationships of power, recognizing that this would create variance from classifications provided by the cultural traditions of which the symbols are part (op. cit.)[3]

Twenty years later, Talal Asad's "Towards a Geneology of the Concept of Ritual" (1988) echoes the concerns of Cohen in a somewhat novel approach. Asad contends that anthropological notions of ritual are quite new, focusing, as they do, on the symbolic, sociological, and psychological functions of the process. He argues that in seeing a festival, funeral, or exorcism as representing something other than what is happening, anthropologists take on roles similar to theologians in that both embody an authority to interpret the meanings of culturally-generated "texts", especially when they are deemed ethnographically inadequate or incomplete (77), as well as to identify and classify symbolic behavior (aided by indigenous informants).

What is missing from this kind of anthropology is the original meaning of "ritual", (at least as Asad believes the term was used in pre-modern Christianity) where it meant learning how to do something rather than the symbolic meaning of what is done (Asad 1988:79). Ritual practices, in other words, draw upon the "prestige of the body" (Zuesse 1979: 248) in which the deepest form of knowing is through doing. Just as concepts of sacred space create a linkage between local and heavenly geographies, so does the body become the vehicle through which decidedly insignificant individuals become sanctified participants in a divine order (especially in the Christian tradition). In Shinto ritual practices in Japan, for example, the body's physical condition, actions, and emotional responses serve to mobilize not only messages (in petitioning the deities) but the renewal of physical and spiritual energies in contexts ranging from private ceremonies enacted at a particular family's request to raucous street festivals that border on anarchy (see the volume on Matsuri [1988])--all for the purpose of influencing "preternatural entities on behalf of the actor's goals and interests" (Turner 1973: 1100).

In this display of "proper" or "sanctioned" behavior (near-anarchy included, as Gluckman's classic 1963 book on "rituals of rebellion" or Brandes' studies of "festivals of inversion" in Spain [1980] and Mexico [1988] show), what matters most are the tactics employed for realizing correct action; a disciplining of bodily activity that qualifies, in Foucault's eyes, as an imposition of a "regime of truth" (Foucault 1980). But, he would add, this system of ordered procedures is linked in a circular fashion with the system of power that produces and sustains the event, as well as to the effects of power which both induce and extend the hold of these rituals upon their participants' lives (131-133). As Asad also demonstrates, it is never possible to discuss the effectiveness of ritual behavior without considering the strategies involved in the formation of "correct" rules or models to follow. In this sense, (and I feel it is crucial to follow his terminology) , he asserts, like Foucault and Cohen, that what is often called "symbolic" behavior is actually ideologically motivated through the hidden exercise of strategic power (Asad 1988: 83).

This would appear to radically shift the focus of interpretation in many of the ethnographic studies of ritual practices as well as symbolic constructions of religion and culture. Geertz' work comes immediately to mind, with his well-known phrase that religious symbols give meaning to existence by providing a model of the world as it is and a model for the world as it ought to be, models which he says are "social events like any other" (1973:93). While stressing the primacy of meaning--or "how people look at the world" (1968:90)--in his comparative study of Islam in Morocco and Indonesia, nowhere can one find mention that the culturally constructed "webs of meaning" are actually being spun by a select few, while most people are simply caught up in them. There is hardly any mention of colonial violence, exploitation and domination. As Scholte rightly points out, it is of paramount importance to discover "who symbolizes and defines significance, and on whose behalf, or at whose expense" (Scholte 1986:10). This kind of analysis appears to stem from a Marxist tradition, providing a sense of the politics of cultural representations and how these are produced and maintained. Continue