You are on page 1of 16

Of Churches, Sects, and Cults: Preliminary Concepts for a Theory of Religious Movements Rodney Stark; William Sims Bainbridge

Journal for the Scientific Study of Religion, Vol. 18, No. 2. (Jun., 1979), pp. 117-131.
Stable URL: http://links.jstor.org/sici?sici=0021-8294%28197906%2918%3A2%3C117%3AOCSACP%3E2.0.CO%3B2-S Journal for the Scientific Study of Religion is currently published by Society for the Scientific Study of Religion.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/sssr.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.

http://www.jstor.org Fri Sep 14 11:37:11 2007

Of Churches, Sects, and Cults: Preliminary Concepts for a Theory of Religious Movements
RODNEY STARK*
WILLIAM SIMS BAINBRIDGE*

This paper construds a set of concepts needed for a theory of religious movements. First, the boundaries for such a theory are set by defining religion and then by translating this definition into terms more useful for theorizing. Next, we demonstrate why a faulty understanding of "ideal types" has led to conceptualizations of church and sect that prevent theorizing. We adopt Benton Johnson's solution for defining the underlying church-sect axis. In the remainder of the paper we show that the concepts of church and sect arenot adequate to encompass the full range of religious movements. The concept of cult L introduced, clarified, and then delimited to exclude cults that fall short of constituting religious movements. Throughout, we molest a variety of sacred cows.

Conceptual schemes abound in the scientific study of religion. Definitions of major concepts have been amended, reconceived,and disputed a t length. Yet, if the purpose of concepts is to serve as primary terms for theories, little progress has been made. Consider the fact that we have not yet even settled on a definition of religion; and therefore we do not even agree on the range of phenomena our theories must confront. The majority of scholars limit religion to systems of thought and activity predicated on the existence of the supernatural (Goody, 1961; Stark, 1965; Spiro, 1966; Berger, 1967).But an articulate minority demands the definition of religion be broad enough to include scientific humanism, Marxism, and other non-supernatural philosophies (Luckrnann, 1967;Bellah, 1970;Yinger, 1970).This is a critical dispute, and anyone who aspires to construct theories of religious phenomena must take a stand one way or the other. Until we resolve this issue, we will not be able to decide, for example, whether a theory of sect movements should include schismatic political movements. Similar ambiguities shroud other vital concepts. Consider the following problems: Many new religious bodies are created by schisms-they break off from other religious organizations. Such new religions commonly are called sects. But there are many other new religious bodies that do not arise through schisms. Instead, they represent religious innovation. Someone has a novel religious insight and recruits others to the faith. Often these new religions are also called sects. But a theory that explains why schismatic religious groups occur may havenothing to say about religious innovation. Is it then only a partial theory of sect formation? Or shall
*Rodney Stark is Prof. of Sociology, University of Washington. William SimsBainbridgeis Assistant Prof. in Dept. of Sociology, University of Washington.

Journal for the Scientific Study of Religion, 1979, 18 (2): 117-133

117

118

JOURNAL FOR THE SCIENTIFIC STUDY O F RELIGION

we distinguish among religious groups on the basis of their origins? If so, what names should we employ for this distinction? And what of the many "quasireligions" such asastrology, Yoga, and the like? Are they religiousmovements? How shall we identify them? These questions are not merely academic. They must be settled by anyone who wishes to theorize about religious movements. In our own work, we are immodestly attempting to construct anintegrated set of deductive theories of major religious phenomena. An important part of our effort is directed towards a theory of religious movements. We intend to present a series of deductive chains, based on common axioms, to explain the formation of religious movements, how such movements succeed or fail, how they recruit members, and so forth. Before we could proceed, however, we found it necessary to redefine a number of key concepts so they could adequately serve our theoretical needs. We also found that satisfactory treatment of these conceptual matters required that we devote a separate paper to this task. This paper consists of a conceptual preface to a series of essays on religious movements. In it we first attempt to delimit the boundaries of the phenomenon to be called religion. We do this in two steps. First, we review the case for what we have already identified a s the dominant definition in the field-a definition we support. But we then show that this definition of religion can be translated accurately into a new terminology we have developed a s part of the core theory on which all of our more narrow theories of religious phenomena rest (Stark & Bainbridge, forthcoming). The purpose of this translation is to coordinate the definition with our deductive system and render its elements fully operationalizable. That is, translation makes the definition scientifically useful both by permitting the logical deduction of many new propositions from it and by facilitating their verification through empirical research. Having thus established a definition of religion, we turn to religious movements and consider concepts available to designate different kinds of movements. We preface this discussion with a n examination of a false notion about ideal types which is widespread among social scientists and which has frustrated useful conceptions of churches and sects. We then draw upon the important work of Benton Johnson (1963) to establish a fruitful basis for the churchsect continuum. Next we show that the church-sect conceptualization is too limited to serve fully the needs of a theory of religious movements. Therefore, we demonstrate the utility of a third concept-the cult-and clarify its use. Along the way we define other key terms such as religious movement, church movement, and religious institution. In conclusion, we exhibit therelevance of these concepts for approaching the basic questions that must be answered by any adequate theory of religious movements. DEFINING "RELIGION"

An enormous amount h a s been written to define religion. Scrutiny of this literature reveals two key problems. First of all, the definition must be quite broad. As Sirnmel(1905) pointed out long ago we must achieve analytic power without losing generality, we must find:

CHURCHES, SECTS, AND CULTS

119

a definition which, without vagueness and yet with sufficient comprehensiveness, has told once for all what religion is in its essence, in that which is common alike to the religion of Christians and South Sea islander, to Buddhism and Mexican idolatry (Simmel, 1905: 349).

The second problem is that the definition must not be too broad. Many modem writers have fallen into this pit. They have created definitions which easily encompass the vast array of faiths we regard a s religions, but which apply as easily to phenomena that seem better excluded. A case in point is the concept of value orientation developed by the Kluckhohns and by Parsons and his students (cf. Parsons & Shils, 1962; Smelser, 1963).A value orientation, according to Kluckhohn:
may be defined as a generalized and organized conception. . . of nature, of man's place in it, of man's relation to man, and of thedesireableand nondesireableas they relate to man-environment and interhuman relations (Parsons & Shils, 1962: 411).

It is evident, however, that this definition applies as fully to Communism as to Catholicism, to science as readily a s to spiritualism. It seems reckless to lump together phenomena commonly thought of as religious and irreligious and argue that there are no significant differences between them. Indeed, it easily is argued that religious and secular value orientations face quite different contingencies when embodied in institutions or movements (Stark, 1965). Nevertheless, for all the problems of the many definitions offered, they contain a common core that yields a n essential element. There is consensus that the thing we are trying to grasp always h a s some primary relevance to questions about "ultimate meaning." There is nothing mysterious about what constitutes questions of ultimate meaning. These are the existential basics: Is there a purpose to existence? Why are we here? What can we hope? Is death the end? Why do we suffer? Does justice exist? etc. Answers to such questions provide ultimate meaning. But not all solutions to questions of ultimate meaning ought to be considered religion. Surely the Communist Party claims to answer many such questions, a s do various scientific philosophies. Furthermore, the problem of excluding such groups from the definition of religion is not solved by the restriction that religions employ super-empirical assumptions in their solutions. All abstractions are super-empirical, the Marxist Dialectic no less so than the Holy Trinity. It is a certain kind of superempirical assumption that identifies what we usually think of as religious solutions to questions of ultimate meaning. This is the assumption of the existence of the supernatural-forces beyond or outside nature, a n unknowable realm that can alter, suspend, or ignore physical forces. The term religion is reserved for solutions to questions of ultimate meaning which postulate the existence of a supernatural being, world, or force, and which further postulate that this force is active, that events and conditions here on earth are influenced by the supernatural. This permits us to see how religion, scientific humanism, and Marxism are both similar and different. Solutions to questions of ultimate meaning come in many forms. But only those predicated on the supernatural are religion. We would now like to approach this definitional problem in another way because a definition of religion that includes a major undefined term such as ultimate

120

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

meaning would be quite useless. Our analysis continues, then, with the observation that, in pursuit of rewards, humans seek explanations: statements about how and why rewards may be obtained and costs are incurred. It is important to realize that we do not conceive ofrewardsin a narrow andmaterialistic way. Rather, rewards are anything humans desire, and humans desire more than the things they can put in their mouths or pockets. Explanations impart meaning to the circumstances of human life. That is, they place mere facts in conceptual relationship to human desires and action. Some desired rewards are very scarce, including some that cannot be definitely shown to exist a t all. In the absence of a desired reward, explanations will often be accepted which posit attainment of the reward in the distant future or in some other non-verifiable context. This insight brings us to the key concept-compensators-on which our theory of religion turns: Compensators are postulations of reward according to explanations that are not readily susceptible to unambiguous evaluation. Nothing pejorative is meant by this term. Compensators merely refer to postulations of reward based in hope and faith rather than in knowledge. Sometimes, of course, hope is fulfilled and faith is redeemed. Compensators merely refer to postulations that the desired rewards will be obtained under conditions not readily susceptible to verification. A compensator is a n intangible promise which substitutes for the desired reward; it h a s the character of a n I.O.U., the value of which must be taken on faith. It will be apparent that some explanations take the form of compensators. They propose a strategy for gaining the desired reward in such a manner that the explanation is not readily subject to verification, and may not be subject to verification a t all. Compensators may be broken down along several dimensions. One useful distinction is between relatively specific and relatively general compensators. This follows the distinction often made in behavioral psychology between specific and general reinforcers or rewards (cf. Homans, 1974: 29). A specific compensator substitutes for a specific reward. That is, it claims to provide something of limited value and narrow scope. For example, a n antibiotic treatment for a n infection is a specific reward, while a magical potion taken in the same circumstances is a specific compensator. General compensators are substitutes for very general rewards (such a s "health") and for large collections of rewards. For example, "heaven" is a very general compensator that implies an unlimited flow of varied rewards. Two other classes of compensators can be identified. Some are utilized in lieu of rewards which exist, but which are in short supply. Transvaluational doctrines that define luxury a s sin and thus transform shortage into virtue are an example. A second class of compensators refers to desires for which no reward exists. The promise of life after death is a n example. An important aspect of general compensators-particularly, but not exclusively, those for which no reward exists-is that they serve as solutions to questions of ultimate meaning. A n answer to such questions would be a grand explanation that tells us how to achieve the most general possible rewards, tellsus why humans often bear terrible costs, and places the full scope of human desires and action in a n apparently intelligible context. Relatively general compensators come in many forms and are available from many sources. A political ideology explaining that scarce rewards will be gained

CHURCHES, SECTS, AND CULTS

121

only after the revolution (but meanwhile people must support the Party) offers compensators. I t does not give the desired rewards here and now but issues a promisory note, the value of which cannot be demonstrated a t present. A general compensator, rather than the desired reward, is also offered by a scientific philosophy advising people to have their bodies frozen in a cryogenic vault until such time a s science discovers how to overcome the aging process. Solutions to questions of ultimate meaning seem only available in the form of compensators. Consider the question: What is the goal of history? To answer is to make a statement that is beyond verification, unless perhaps in the far distant future. Furthermore, some questions of ultimate meaning can only be answered by assuming the existence of the supernatural. For example, for history to have a goal, some conscious being must have set that goal. Only a supernatural being could have the vantage point and the power to establish a goal for history. It is obvious that religion is a major source of general compensators, and that any particular religion organizes its compensators into a system. Indeed, the most general compensators entail explanations concerning ultimate meaning. It follows, therefore, that the more standard definition of religion we discussed above can be translated into the following form: A religion is a system of general compensators based on supernatural assumptions. Limited space makes it impossible to demonstrate the full merits of this definition in the present paper. Elsewhere (Stark & Bainbridge, forthcoming), we derive this definition from basic elements of our core theory. Using this translation, we derive a very large number of quite complex propositions about religious phenomena, including why and how religious movements form, and why and how they thrive or flounder. Based on a n analysis of exchange processes, and clarifying what is exchanged, this definition permits us a clearer vision of what actually is going on within religious organizations. What has been accomplished in this paper thus far is the specification of the scope of a theory of religious movements. Such a theory applies only to movements which sustain systematic solutions to questions of ultimate meaning (general compensators) and which predicate these solutions on the assumption of the existence of the supernatural. We now turn to conceptualizing the specific movements and institutions to which the theory pertains. UN-IDEAL TYPES The conceptual literature on churches and sects is dominated by typologies. Indeed, the literature commonly refers not to churches and sects, but to the "churchsect typology." Sad to say, the kind of types usually developed by sociologists are of no use in theory construction-they serve a s tautological substitutes for real theories and tend to prevent theorizing. The trouble started with Weber. It was Weber who introduced both the churchsect typology and a misunderstanding of the ideal type. In his classic work on methods, Weber (1949) advocated the construction of ideal types:

. . . by the onesided accentuation of one or more points of view and by the synthesis of a great many diffuse, discrete, more or less present and occasionally absent concrete indiurdual

122

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION phenomena, which are arranged according to these one-sidedly emphasized viewpoints into a unified analytical construct. I its conceptual purity, this mental construct cannot be found n anywhere in reality .. . They are used as conceptual instruments for comparison with and the measurement of reality (91-97).

Generations of sociologists have regarded Weber's ideal type a s similar to concepts commonly found in the physical sciences. Physics, for example, abounds with "ideal gasses" and "frictionless states." It is understood that such do not exist but are the absolute base points on a measuring continuum against which degrees of friction or the expansion of gases can be calibrated. But there is a n immense and fatal difference between these ideal types and those proposed and compounded by Weber. The ideal types of physics anchor a single continuum along which it ispossible to rank order all empirical or hypothetical cases. Comparison with the ideal is direct and unambiguous and thus permits measurement. But Weber's types prevent comparison and measurement, despite his claim that "they are indispensible for this purpose." Following Weber, sociologists often use correlates in their definitions of concepts. But it is attributes, not correlates, that belong in a definition. Consider the most minimal use of a definition: to permit clear identification of cases as belonging or not belonging to the defined class. Since correlates are not always present, and often may not be present, their use as defining features often leads to misclassification. Worse yet, when many correlates are involved (asWeber advised they should be), the result is a jumble of mixed types which cannot be ordered and thus which cannot yield measurement. The usual outcome is a proliferation of new subconcepts or types, and sometimes it seems to be that each new empirical case must become a unique type-which is to classify nothing. When attributes are the basis of definition (since they are always present in the phenomena to be classified) and when enough attributes have been utilized to limit the class in the desired fashion, no ambiguity results. For then the concept forms a n underlying unidimensional axis of variation. This kind of ideal type, then, does provide a zero point for comparison and ranking. These general remarks are, as Homans would have it, so obvious that they are highly controversial. An assessment of the church-sect typology gives them relevance. CHURCHES AND SECTS Although Weber introduced the notions of church and sect, it was his student Ernst Troeltsch (1931),who first made them important. Troeltsch used a n ideal type of church and a n ideal type of sect to categorize roughly what heregarded a s the two main varieties of religious bodies in prel9th century Christian Europe. Each type was identified by a host of characteristics which were, a t best, weak correlates of one another and of the phenomena to be classified. Subsequent attempts to utilize Troeltsch's types in other times and places caused frustration. The empirical cases just wouldn't fit well, so new users created new church-sect typologies. Indeed, it would be close to the truth to claim that each new user, or a t least each new user with

CHURCHES, SECTS, AND CULTS

123

new cases to classify, created a new typology based on different correlated features of the phenomena to be classified. And each new typology suffered the same defects as those it replaced. It wouldn't organize the data. The tragedy is that no typology so constructed will ever create the organization needed for theorizing. This problem is easy to illustrate. Suppose five correlates are used to define the ideal church, with negative values on these same five defining the ideal sect. Then, suppose we treat these criteria as dichotomies. The result is 32 logicially possible types (since the defining criteria can vary independently), of which 30 are mixed types. These mixed types cannot be ordered fully. Which is more churchlike, a grouping possessing characteristics A and B, but lacking C, D, and E, or one with D and E, but not A, B, or C? In the empirical world, mixed types have been the rule. Underlying most sociologists' interest in churches and sects is a theory about religious movements. In 1929 H. Richard Niebuhr argued that the sect is an unstable type of religious organization which, through time, tends to be transformed into a church. But, he argued, following this transformation, many members' needs that had been satisfied by the sect go unmet by the church. In time this leads to discontent, which prompts schism and the splitting off of a new sect, which then is transformed slowly into a church, thus to spawn anew sect: an endless cycle of birth, transformation, schism, and rebirth of religious movements. It is this theory that has long capitivated sociologists of religion. Troubleis that a typological conception of churches and sects prevents all theorizing. How can one theorize about the movement from sect to church when one cannot rank groups as more or less churchlike? It would humble physicists to try to theorize under such handicaps. Thus it was a n event of considerable magnitude when Johnson (1963)proposed a vast definitional rummage-sale. He discarded dozens of correlates from the various definitions of church and sect and settled on a single attribute to classify religious groups: "Achurch is a religiousgroup that accepts thesocialenvironment in which it exists. A sect is a religious group that rejects the social environment in which it exists." What Johnson did was to postulate a continuum representing the degree to which a religious group is in a state of tension with its surrounding socio-cultural environment. The ideal sect falls at one pole where the surrounding tension is so great that sect members are hunted fugitives. The ideal church anchors the other end of the continuum and virtually is the socio-cultural environment-the two are so merged it is impossible to postulate a basis for tension. Johnson'sideal types, unlike Weber's, are ideal in precisely the same way that ideal gases and frictionless states are ideal. They identify a clear axis of variation and its end points. Johnsons' reconceptualization also permits clear definition of two other important concepts: religious movement and religious institution. When we look at the low tension end of his axis we find not only churches, we find religious institutions. That is, we find a stable sector of the social structure, a cluster of roles, norms, values, and activities associated with the performance of key social functions. Social institutions are not social movements-if we define social movements as

124

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

organized groups whose primary goal i s to cause or prevent social change (Bainbridge, 1979). Institutions adapt to change. Social movements seek to alter or become institutions. Thus we can see that if religious institutions are one pole of the tension axis, as we move along the axis in the direction of greater tension we discover religious movements. That is, religious movements are social movements that wish to cause or prevent change in a system of beliefs, values, symbols, and practices concerned with providing supernaturally-based general compensators. In a very general way it can be asserted that religious movements are organized groups wishing to become religious institutions. Such groups would like to become the dominant faith in their society, although they may make little effort to achieve this end if they are convinced their chances are too remote. Johnson's axis also permits us to characterize the direction taken by religious movements. When they move towards less tension with their sociocultural environment they are church movements (although a group may remain a sect during a long period of movement in this churchlike direction). When groups move towards the high tension pole they are sect movements. In another paper (Bainbridge & Stark, forthcoming), we show that the degree of tension experienced by a religious group can be measured easily and unambiguously. We also explain that tension with -the socio-cultural environment is equivalent to subcultural deviance in which the relationship between the high tension group and the surrounding society is marked by difference, antagonism, and separation-three integrated but conceptually distinguishable aspects of deviance. Since tension defined in this way can be measured, numerous empirical studies can now readily be performed, testing any hypothesis or theory in which tension plays a part. The utility of Johnson's reconceptualization of church and sect is obvious. For example, we may liow see a t a glance that the Catholic Church in the United States is more sectlike than is the Catholic Church in Ireland. In most prior typologies this could not be seen. Furthermore, because the axis of variation is clear, variation cries out for explanation. It becomes obvious how to proceed towards theories to rectify and extend Niebuhr's work. Indeed many important variables long thought to influence the generation of sects or their transformation into churches now can be examined. In the past, these variables have been utilized in typologizing and thus were locked in tautology. Now we can ask, for example, if the arrival of a generation of members socialized into the sect a s children, rather than converted into it a s adults, plays a major role in pushing sects down the road to churchliness. In the past this variable was lost in the creation of 1)sects with convertedmembers, 2) sects with socialized members, 3) churches with converted members, and 4) churches with socialized members. These four boxes tell us nothing. A proposition that related socialization to the transformation of sects into churches could tell us much. Since the purpose of this essay is to develop the conceptual means for theorizing, it might seem that Johnson already had done the necessary. Unfortunately, he did not. There are a t least two kinds of religious movements in a high state of tension with their surrounding socio-cultural environment, and it demonstrably inhibits efficient theorizing to regard both kinds a s sects, and ignore differences between them. Therefore, wemust now add some complexity to Johnson's elegant parsimony. Neibuhr's theory exclusively concerns schismatic religious movements, which

CHURCHES, SECTS, AND CULTS

125

he identifies a s sects. He wasnot speaking of all small, deviant religious movements, but only of those whose existence began a s a n internal faction of another religious body. This is, of course, a very common kind of religious movement. However, it is not the only kind of religious movement in a high state of tension with the surrounding socio-cultural environment. Many such movements have no history of prior organizational attachment to a "parent" religion-thus they are not schismatic. Indeed they lack a close cultural continuity with (or similarity to) other religious groups in a society. Thesenon-schismatic, deviant religious groups are themselves of two types. One type represents cultural innovation. That is, along with the many familiar components of religious culture appearing in the beliefs, values, symbols, and practices of the group, there is something distinctive and new about them a s well. The second type exhibits cultural importation. Such groups represent (or claim to represent) a religious body well-established in another society. Examples are various far eastern faiths in the U.S., or Christianity in the far east. In common parlance, these deviant but non-schismatic bodies often are referred to a s cults (Eister, 1972). These preliminary remarks can be expanded to define sects and cults. Both cults and sects are deviant religious bodies-that is, they are in a state of relatively high tension with their surrounding socio-cultural environment. However, sects have a prior tie with another religious organization. To be a sect a religious movement must have been founded by persons who left another religious body for the purpose of founding the sect. The term sect, therefore, applies only to schismatic movements. Note that it is not required in this definition that a sect break off from a church, a s Niebuhr argued. To do so would land us back in the wilderness of typologies. For sects sometimes break off from other sects. Indeed, it h a s happened that churches have broken off from sects (Steinberg, 1965).Furthermore, we plan to apply elements of church-sect theory to the careers of cults. Therefore, these are matters to theorize about, not to lock into definitions. Because sects are schismatic groups they present themselves to the world a s something old. They left the parent body not to form a new faith but to reestablish the old one, from which the parent body had "drifted" (usually by becoming more churchlike). Sects claim to be the authentic, purged, refurbished version of the faith from which they split. Luther, for example, did not claim to be leading a new church but the true church cleansed of worldy encrustations. Cults, with the exception noted below, do not have a prior tie with another established religious body in the society in question. The cult may represent a n alien (external) religion, or it may have originated in the host society-but through innovation, not fission. Whether domestic or imported, the cult is something new vis-a-vis the other religious bodies ofthe society in question. If domestic-regardless of how much of the common religious culture it retains-the cult adds to that culture a new revelation or insight justifying the claim that it is different, new, more advanced. Imported cults often have little common culture with existing faiths; while they may be old in some other society, they are new and different in the importing society. Cults, then, represent a n independent religious tradition in a society. In time they may become the dominant tradition, in which case there i s no longer much

126

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

tension between them and the environment, and they become the church or churches of that society. And even long before cults become churches, they too are prone to internal schisms. Thus, within the context of cult movements, schismatic movements can form. A theory to explain sect formation can then be applied to cults to explain their schismatic tendencies. But it is vital to see that a theory of sect formation simply will not serve as a theory of cult formation. The geneses of the two are very different. To sum up: sects are breeds of a common species; cults are a different species and occur by mutation or migration. Sects, being schismatic, are embodied in religipus organizations, and their status as religious movements is clear. However, many cults do not develop into full-blown religious movements. Therefore, it is necessary to survey the range of cults to identify various forms, only some of which will fall within the scope of a theory of religious movements. CULTS Three degrees of organization (or lack of organization) characterize cults. The most diffuse and least organized kind of cult could best be identified a s a n audience. Sometimes some members of this audience actually may gather to hear a lecture. But there are virtually no aspects of formal organization to these activities, and membership remains at most a consumer activity. Indeed, cult audiences often do not gather physically but consume cult doctrines entirely through magazines, books, newspapers, radio and television. More organized than audience cults are what can be characterized a s client cults. Here the relationship between those promulgating cult doctrine and those partaking of it most closely resembles the relationship between therapist and patient, or between consultant and client. Considerable organization may be found among those offering the cult service, but clients remain little organized. Furthermore, no successful effort is made to weld the clients into a social movement. Indeed, client involvement is so partial that clients often retain a n active commitment to another religious movement or institution. Finally, there are cult movements which can be distinguished from other religious movements only in terms of the distinctions between cults and sects developed above. It is only cult movements that will be addressed in our subsequent theory. However, the less organized types currently are more common and needed to be described with some care so they would not be confused with the full-fledged cult movement. Our three "types" of cult can be distinguished in terms of the qualities of the compensators they provide. Thus, rather than being three more "unideal types," they are merely convenient words to describe particular ranges in measurement scales on dimensions we have already introduced. Audience cults are identified by the low value, and consequent low cost, of the compensators they provide. Our two other types are points along a continuum defined by the degree of generality of the compensators offered by a given cult. Audience cults offer very vague and weak compensators, often little more than a mild vicarious thrill or social entertainment. Client cults offer valued but relatively

CHURCHES, SECTS, AND CULTS

127

specific compensators. Psychoanalysis and Dianetics claim to cure neurosis, but they do not promise everlasting life. Cult movements present a much larger package of compensators, including the most general compensators offered by full-fledged sects and churches. These three levels of cults can also be described conveniently in traditional terms: audience cults provide mythology; client cults add serious magic; cult movements are true religions. Put another way, not all cults are religions, in the strict sense of the term, because not all represent systems of ultimate meaning. Audience cults are part of a diffuse occult milieu that toys with vague images of the ultimate. Also within this milieu are many unsuccessful client cults and cult movements, including many one-person operations with n o committed followers. But, in themselves, audience cults offer only very limited rewards and compensators. Client cults are magical rather than religious. This means that their main business is selling compensators rather than rewards, and the compensators are relatively specific and not embodied in a total system of ultimate meaning. Freud does not tell us how our lives meaningfully are related to the ultimate laws of the universe, but h e does provide specific meanings to attach to limited aspects of the human condition. Our distinction between magic and religion i s entirely in line with some of the best traditional descriptions of these two related expressions of human desire (e.g., Durkheim, 1915: 42). Only cult movements are genuinely religious in the fullest sense, although some of them may not accept the label of "religion." For a long time the authors of this paper debated the proper conceptual boundary between religious cults and other cult like phenomena, especially magical, service organizations. A similar debate h a s raged in the courts in recent years a s various groups (such a s Scientology, the Black Muslims, and Transcendental Meditation) have struggled to obtain or avoid legal definition as religious organizations. One of the advantages of our definition of cult is that it explains why this debatehas taken place. Because cults are culturally novel or exotic, the conventional definitions of our society do not apply to them automatically. When a sect breaks away from a church, it takes with it the label "religious," but cults are not born with the religious label attached. Under our definition of "cult" and "religion," analysis using the concept of compensators makes good sense out of the otherwise bewildering variety of cults. While all cults have certain features in common with traditional religious phenomena, only those cults that offer quite general compensators (a system of ultimate meaning) fully qualify a s religious. The extent to which a group is really religious is a n empirical question of where--on a continuum-it lies a t the moment. Our conceptualization also transforms what h a s been a swamp of unproductive debate into a fruitful field for scientific research. Recall Neibuhr's observation that sects tend to evolve into churches. Our research indicates that magical client cults tend to evolve into religious cult movements. Although examples abound, the best documented are Scientology (Wallis, 1976) and "The Power" (Bainbridge, 1978). Both groups began a s limited psychotherapy services. Because they were culturally novel and were not based on any body of verified scientific research, they were magical client cultsrather than technical medical services. As the years passed, each of these two cults began offering compensators that were more and more general and

128

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

for which no equivalent rewards existed. Their ideologies ramified into complex systems of ultimate meaning. Now, both are highly developed cult movements. Within our conceptual framework, we can also apply to cults many of the ideas originally developed in understanding sects and churches. For example, we can discuss the degree of tension that a cult experiences with the surrounding society. It would appear that cults can enjoy relatively low tension with their environment a s long a s they do not organize into religious movements. Participating in cult audiences seems to be a very low risk activity. A 1976 Gallup Poll suggests that 22 percent of Americans believe in astrology, and astrology columns and publications flourish. But very little flak is directed towards astrology. In general, the clergy of American churches seem to ignore the astrology cult. At most, persons who participate in audience cults may risk censure from thoseimmediately around them. Client cults, too, do not provoke great hostility in the surrounding socio-cultural environment. As long a s they do not run afoul of fraud statutes (by selling building lots on a fictitious planet, for example) or licensing statutes governing medical practice, they are not subject to much harrassment. Client cults seem almost never to serve a low status market, if for no other reason than they charge for their services. Consequently, client cults also seem to be somewhat protected by the high status of their clientel. For example, the spiritualists primarily have drawn on middle and upper class clientel, a s do most of today's personal adjustment cults, which seems to have lent them considerable protection from opposition. It is when cults become religious movements that their environment heats up. For example, a s Scientology evolved from a client cult to a movement seeking major commitment from members, its legal troubles grew. In similar fashion, Transcendental Meditation took little heat so long a s it concentrated on teaching clients to meditate during a few training sessions. With its transformation into a n intense religious movement-amid claims that advanced members could fly (levitate)-public reaction has grown. And it i s cult movements, not client cults or audiences, that today face opposition from irate parents who hire deprogrammers to kidnap their children from the bosom of the cult. Furthermore, among cult movements, the more a cult mobilizesits membership, the greater the opposition it engenders. Cults whose members remain in the society to pursue normal lives and occupations engender much less opposition than do cults whose members drop everything and become full-time converts. In part, this is probably because cults that function a s total institutions rupture converts' ties to conventional institutions. This generates personal grievances against the movement. It is one thing to know your son or daughter, for example, attends a weird church and has odd beliefs, but it is quite something else to lose contact with a child who takes up fulltime participation in a n alien faith. Indeed, even Catholic parents often find it painful to lose a child to a convent or monastery even though the question of deviant faith is not a t issue. Thus the rule seems to be, the more total the movement, the more total the opposition to it.

CHURCHES, SECTS, AND CULTS

CONCLUSIONS Concepts must not only facilitate theorizing; they ought to inspire it. Concepts should identify a phenomenon that arouses our interest and should present a clear enough picture of the phenomenon's variation that we are prompted to explain it. We believe the conceptual scheme developed above does encourage such theory construction. Therefore, to conclude this paper, we shall point out some basic questions thrown into relief by these particular formulations-basic questions that must be answered by any adequate theory of religious movements. The most obvious taskis to seek a set of premises from which a theory of religious schism can be deduced. Why and under what conditions do factions form in a religious group? Why and when do these result in the splitting off of a sect movement? A church movement? Under what conditions do schismatic forces produce secular rather than religious movements?This last question reminds us that we must theorize not merely about the internal workings of religious bodies, but we must also deal with their external environment. Thus, for example, we need to know not only how sects form within a parent body, but the social conditions under which religious schisms are more and less likely. It must be recognized that in posing these questions we are permitted, by our concepts, to avoid the assumption that sects split off from churches. We aredirected, instead, to the problem of faction and exodus in any kind of religious body. This is important because the historical record makes it clear that sect formation is probably more common within bodies that are themselves sects rather than churches. It also permits analysis of schismaticmovements originatingin cults, just as it lets us deal with church movements which have split off from sects or cults. Instead of converting these alternatives into un-ideal types, our conceptual scheme makes it possible to construct propositions to account for these variations. Our concepts make it clear that a theory of religious schism pertains to only some religious movements. Cults are not the result of schism (although, once founded, cults become subject to schism). Therefore, a theory of cult formation may have very few propositions in common with a theory of religious schism (a term to be preferred to sect-formation because it is more inclusive). We must explain when the sociocultural environment is conducive to cult formation, and why. We must also specify the process by which people actually form a cult, and we must explain the contingencies governing importation of cults. The next requirement is for a theory of development and transformation. Once a sect or cult is formed, what contingencies govern whether it will grow, stall, or fail? What factors operate to push it in a church-like direction (toward lower tension)? What factors push toward higher tension? Finally, we must close the circle by showing how and to what extent the factors involved in the formation of religious groups influence their development and transformation, and how these in turn are involved in the onset of schism and of cult formation and importation. This is, of course, a formidable list. However, with the conceptual clarifications outlined above, it is a t least possible to see where to begin. In work in progress we have identified some propositions which enable us to deduce a t least partial answers

130

JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

to many of these questions. With luck we shall succeed in constructing a useful beginning theory of religious movements. In conclusion we would like to suggest that such a theory is important for sociologists to pursue. One virtue of this pursuit remains a s Ellsworth Faris (1928) suggested 50 years ago: The origin of many social institutions and much human culture "is lost in mystery." We haveno timemachineto permit study ofthe past. But religious sects and cults can be studied from the start and thus can furnish a body of information "concerning the rise of institutions," and the invention of culture (Faris, 1928). Social scientists have always envied geneticists' rapidly reproducing fruit flies. We suggest that for a great many basic sociological questions the cult and sect provide a reasonable substitute. Moreover, cult and sect formation can no longer be regarded a s a dead issue. Prior to World War I1 sociologists demonstrated considerableinterest in sect and cult movements because these had always been a common feature of American society and continued to be so throughout the 1920s and 30s. However, despite the resurrection of a n active sociology of religion during the "religious revival" of the post war years, sociologists lost interest in sects and cults primarily because few seem to have been formed during the 1940s and 50s. Indeed, when the senior author went in search of a cult movement to study in 1960, such movements were hard to find. Most of those he found wereagingremnants of cults that formed and flourished during the 1930s. Even in California, a new cult movement that could be studied while still developing took more than a year of hard work to find. And it was not even a domestic innovation, but a n importation-the Korean cult of Reverend Sun M. Moon, now famous a s the "Moonies." However, when the junior author set out to study cult movements in the early 19708,no search was needed. The problem was not to find a cult movement, but to pick one to study from dozens available. Such fluctuations in the frequency of cult and sect formation are part of what we must explain. But it is clear that these days we arenot limited to historical materials about a dying phenomenon-religious innovation is alive and well. Unfortunately we have not made much theoretical progress towards explaining religious movements since 1929 when Neibuhr sketched his church-sect theory and when Ellsworth Faris (1929) published a one-page, eight-point directive listing urgent business for sociological study. His list included the following:
. . . 2. A sociological study of the origin and evolution of specific religious sects ought to be very fruitful in making us intelligent about how institutions arise and develop. The study of sects which have disappeared and of those whose birth was abortive would be highly instructive. 3. The study of social movements should include significant religiousmovements.. ..There are social, racial, and economic aspects which do not appear at all in the controversial literature.

Since Faris wrote these lines, many sects and cults have been studied ethnographically, but theoretical progress was stifled by ill-advised attempts to find a set of boxes in which to place each religious movement. Built of correlates, the sets of boxes expanded endlessly. Everything was named. Nothing was explained. We believe it is time to use this mass of rich observations for the purpose of theory construction and theory-testing. These pages have been offered a s a step in that direction.

CHURCHES, SECTS, AND CULTS REFERENCES Bainbridge, William Sims 1978 Satan's Power. Berkeley: University of California Press. 1979 "Collective behavior a n d social movements." I n Rodney Stark, Sociology. New York: Worth. Bainbridge, William Sims and Rodney Stark Forth- "Sectarian tension." coming Bellah, Robert N. 1970 "Christianity and symbolic realism." Journal for the Scientific Study of Religion 9 (2): 89-96. Berger, Peter L. 1967 The Sacred Canopy. Garden City, N.Y.: Doubleday. Durkheim, Emile 1915 The Elementary Forms of the Religious Life. London: George Allen and Unwin Ltd. Eister, Allan W. 1972 "A theory of cults." Journal for the Scientific Study of Religion 11 (4): 319333. Faris, Ellsworth 1928 "The sect and the sectarian." American Journal of Sociology (Special edition) 60: 75-89. 1929 "Some phases of religion that are susceptible of sociological study." American Journal of Sociology (Special edition) 60: 90. Goody, Jack 1961 "Religion and ritual: The definitional problem." British Journal of Sociology 12: 142-164. Homans, George Caspar 1974 Social Behavior: Its Elementary Forms. New York: Harcourt Brace. Johnson, Benton 1963 "On church a n d sect." American Sociological Review 28: 539-549. Lofland, John 1966 Doomsday Cult. Englewood Cliffs: Prentice-Hall. Lofland, John and Rodney Stark 1965 "Becoming a world-saver: A theory of conversion to a deviant perspective." American Sociological Review 30: 862875. Luckmann, Thomas 1967 The Invisible Religion. New York: Mamillan. Niebuhr, H. Richard 1929 T h e Social Sources of Denominationalism. New York: Henry Holt. Parsons, Talcott and Edward Shils ( 4 s . ) 1962 Towards a General Theory of Action. New York: Harper Torchbooks. Selznick, Philip 1960 The Organizational Weapon. Glencoe: The Free Press. Simmel, Georg 1905 "A contribution to the sociology of religion." American Journal of Sociology 11: 359-376. Smelser, Neil J . 1963 Theory of CollectiveBehavwr. New York: Free Press. Spiro, Melford E. 1966 "Religion, problems of definition and explanation." 4. 85-126 in Michael B a n t o n (ed.), Anthropological Approaches to thestudy ofReligwn. New York: Praeger. Stark, Rodney 1965 "A sociological definition of religion." Pp. 3-17 in Charles Y. Glock and Rodney Stark, Religion and Society in Tension. Chicago: Rand McNally. Stark, Rodney and William Sims Bainbridge Forth- "Towards a theory of religion: religious coming commitment." Stark, Rodney and Kevin W. Welch Forth- "IQ and religious commitment: Testing a coming theory." Steinberg, Stephen 1965 "Reform Judaism: The origin and evolution of a church movement." Journal for the Scientific Study of Religion 5: 117-129. Troeltsch, Ernst 1931 The Social Teaching of the Christian Churches. New York: Mamillan. Wallis, Roy 1976 The Road to Total Freedom. New York: Columbia University Press. Weber, Max 1949 The Methodology of the Social Sciences. Glencoe: The Fkee Press. Yinger, J. Milton 1970 The Scientific Study of Religion. New York: Mamillan.

You might also like