HISTORY OF RELIGIONS

HISTORY OF RELIGIONS . This article presents an overview of the history of religions as a scientific discipline. It is not intended to provide a comprehensive survey of the specific data that lie within the province of the historian of religions, nor does it attempt to survey the broader history of the discipline. The purpose here is rather to provide a brief description of the nature of the history of religions and to discuss its methods of research. The first part presents a theoretical examination of the dialectical method proper to the discipline. In the second part, attention will be given to the actual field of comparative research through the presentation of a brief historical typology.

Theoretical Overview

The discipline of the history of religions is characterized by the dialectical relationship that exists between its object of study and its methods of research. It is, of course, the concept of religion that best defines the discipline's object. This concept, however, though a necessary precondition for research, is never allowed to function as an a priori category, which would predetermine the direction of the historian's inquiries. Instead, it is held in dialectical tension with the ongoing progress of research. The methods employed in this research are in turn adapted to the deeply historical nature of their subject matter. Such methods are essentially inductive, intended to grasp religion in its concreteness, in its historical creativity, and in its meaningfulness for the cultural, social, and individual lives with which it is interwoven. The dialectic that emerges from this interaction of the concept of religion with specific, ongoing historical investigations may be taken as a distinctive feature of the discipline.

The nature of this dialectical foundation of the history of religions may be clarified by contrasting it with several alternative approaches to religion. First of all, it must be distinguished from the hermeneutical approach, which fixes upon a single interpretative key to unlock the mysteries of the phenomenon under investigation. Hans Jonas's intelligent application of modern existentialist categories, such as the feeling of Geworfenheit ("thrownness") in his study of Gnosticism, or Rudolf Otto's use of the category of the holy, may be taken as examples of such an approach. In either case, the task of accurate historical description and the construction of a complete and articulate typology of the phenomenon under study are in danger of being neglected. As a result the interpretative key is insufficiently tested against the facts. In addition, the crucial problems of continuity and change, which are unavoidable in the comparative study of religions, tend to be overlooked.

The dialectical character of the history of religions discipline may also be contrasted with the so-called phenomenological method. This method tends to focus only on the synchronic elements of religion, describing and classifying religious forms without reference to particular historical contexts. It aims at capturing the meaning of religious phenomena without committing itself to an analysis of the historical, cultural, social, and psychological settings of those phenomena. It thereby neglects the study of the diachronic, formative processes that give a religious phenomenon its depth and endow it with the colors of real life. In the end, the special sensitivity of the individual phenomenologist is called upon to fill this gap. No doubt there is something valuable in the phenomenologist's capacity for empathy, or Einführung. But the historian of religions must attend first and foremost to the fact that the object of his study is a historical object. The dialectical processes of creation and change, development, and even revolution escape the phenomenologist's ahistorical gaze.

It would be a mistake to conclude from the inadequacies of the phenomenologist's method that the proper alternative lies in historicism. On the contrary, historicism, whether in its idealistic or materialistic form, must also be distinguished from the approach of a historian of religions. While religion is a decidedly historical phenomenon, it must not be reduced to history. Historicism makes religion a mere moment in a dialectic that essentially transcends it. This is as true for Hegel as it is for Marx. In addition to these historical forms of reductionism, there are sociological and psychological approaches to religion that are reductive as well. Whatever the particular form, however, all forms of reductionism have in common an appeal to a univocal conception of religion located within a preconceived frame of reference.

It is no solution to counter historical reductionism with an appeal to the irreducible character of religion as perceived by the subjective, experiential sensitivity of the phenomenologist. In both cases, that of the phenomenologist as well as that of the historicist, there is an illegitimate appeal to an a priori, preconceived conception of religion. It is this a priori character of their respective conceptions of religion that is incompatible with the positive, inductive, comparative-historical approach proper to the historian of religions and to the dialectic that the historian must preserve between his tentative interpretative categories and the ongoing progress of his research.

Theory is equally illegitimate in the history of religions when it is applied in an a priori, undialectical manner. Nevertheless, it is not to be rejected out of hand. On the contrary, theory is indispensable when it functions as hypothesis open to verification, revision, or rejection on the basis of empirical research. The distinguished anthropologist E. E. Evans-Pritchard, for example, goes too far in his distrust of theory. He would limit the study of religion to the description of the social and cultural functions of specific religions in order to escape the distortions of general theories of religion. This is overly cautious. In fact, even though theoretical biases are unavoidable, arbitrarily reductive positions can be tested on the basis of positive research and their shortcomings exposed. An excellent example of this is Andrew Lang's criticism of the theory of animism. Using concrete evidence, Lang showed that this theory was reductive in its claim to represent the most archaic form of religion. In fact, the theory was unable to account for the widespread belief in a high god among primitive peoples. This use of empirical, historical evidence to test proposed theories or hypotheses provides ample insurance against the dangers of overextended or reductive generalizations.

So far I have argued that preconceived or a priori theories of religion that are not held accountable to the findings of historical research remain outside the dialectic that governs the work of the historian of religions. At most, such theories can function as hypotheses awaiting verification. It may be helpful now to examine the dialectic itself more closely by focusing on the particular problem of the definition of religion. How, if at all, are we to arrive at a clear concept of religion? And how does this concept function in actual research? In raising these questions one is immediately confronted with what at first appears to be a methodological impasse. On the one hand, if such a concept is to escape the problems of an a priori characterization of religion, it must itself arise out of actual comparative historical study. On the other hand, it is hard to see how such comparative study is to proceed until this concept is in hand.

The concept of religion

The appearance of an impasse is created by the way in which the problem is posed. If we assume that a concept of religion is at the same time a prerequisite for and a result of comparison, then we are indeed faced with an insurmountable paradox. But this is not the case if we resort to a dialectic that unites the notion with its employment in such a way as to make them mutually dependent. In this case the concept of religion is inductive in origin, and its clarification goes hand in hand with progress in empirical research.

At this point it is clear that an adequate notion of religion is not to be formed through a mere a priori selection of data to which research should be extended. The primary problem is not to extend a conception of religion over the widest possible range of material but rather to discover a conception that is adequate to specific historical contexts. The search for a universally adequate definition of religion can lead quickly to a minimal notion of religion, a kind of lowest common denominator with no practical usefulness. Such a univocal definition, which would seek to rank the different religions as so many species under a single genus, is clearly inadequate for a fundamentally empirical discipline. Such a discipline requires conceptual categories that are continuously being created and are always open to further revision in the light of the development of comparative studies. Far from being a univocal concept, the notion of religion that emerges from the continued comparison of new and varied historical materials is an analogical notion.

Things are described analogically when they correspond to each other in certain important respects but differ from one another in other, equally important respects. Analogical notions thus contrast sharply with univocal notions. When we apply the term vertebrate, for instance, to a man, a dog, and a crocodile, we are applying it univocally, since the notion of vertebrate is integrally and equally realized in each of these cases. The same cannot be said of analogical notions. When we describe both a feather pillow and our summer's reading as "light," we are speaking analogically, since the quality of lightness applies very differently in each case. To say, then, that the historian of religions must employ the term religion analogically rather than univocally means that it cannot be assumed that all the phenomena studied under the rubric of religion are all "religious" in the same respect. This is especially true when studies extend beyond the context of closely related cultural milieus.

Now, it is exactly the analogical value of the notion of religion that relieves it of the impossible task of achieving universal extension while remaining responsive to the specifics of history. If the notion of religion is analogical, the aim of research will not be the progressive extension of a univocal concept but the documentation of sets of partial affinities between different systems of belief and practice, which are the segments of the polychromatic network that constitutes the variegated world of religion. It is clear that this procedure is much more in keeping with the comparative-historical aims proper to the history of religions than any merely phenomenological approach. The approach I am describing is meant to remain constantly in touch with the concreteness of its object; facts and sets of facts (historical contexts and processes) are compared directly, without being submitted to an intermediate process of abstract categorization.

If the historian of religions cannot rely on a univocal concept of religion or its deductive use, it follows that he will not be dependent upon philosophy or some metascience for his categories. On the contrary, it is to be expected that comparative-historical research will seek its categories precisely in the historical continuities and discontinuities it studies. The historian of religions is not concerned with facts isolated from their historical contexts and processes but rather with these contexts and processes themselves. The historian of religions aims at discovering the degree and quality of the affinity that exists between religions and that warrants characterizing them as such.

Given the empirical nature of the history of religions and its interpretative categories, and in particular its analogical conception of religion, it becomes apparent that its findings can be neither verified nor falsified through appeals to either a priori reasoning or personal convictions. The sole criterion of adequacy becomes empirical. What must be shown is the adequacy of a particular account to the facts it claims to describe, both the facts of a given historical situation and the relationship between one such situation and others. Ultimately, in the comparative context proper to the discipline, this will lead to a principle of holism as the final criterion of adequacy. Accounts will be more adequate as they encompass broader ranges of concrete data.

At this point it may be asked why the concept of religion should be retained at all, even in an analogical sense. Strictly speaking, one could simply refer to various sets of continuities by the use of arbitrary symbols (x, y, z, and so forth). But this analytical procedure would not fit the historical character of the discipline, which is rooted in a cultural milieu and motivated by a specific intellectual interest. It is not by chance that the problem I am analyzing is termed religion and not way, law, or ethos, although the latter terms would be equally legitimate labels for comparative study. It is for historical reasons that this particular field is called comparative religion. This does not mean that the validity of the enterprise is merely relative or Eurocentric. It means only that in positing this notion one is conscious that the comparative study of religions began in a particular historical and cultural milieu. It was in this milieu that the term religion came to be used in analyzing the problem of continuity and change within systems of belief and ritual behavior. It was in this setting that "true" religion began to be contrasted with the "other" religions. Out of this context arose the history of religions, which rejected the pretense of evaluating different religions normatively (a task, on the other hand, legitimate and unavoidable for philosophy and theology) and instead studied them phenomenologically. On this level it could speak both of the universality of religion, since systems of belief and ritual practice more or less comparable with those in the West can be found everywhere, and of religions in the plural, given the clear differences that distinguish these systems.

These considerations lead us back to the dialectic that is the basis of the comparative-historical methodology proper to the history of religions. The emergence of the notion of religion among Western students of religion was the result of just such a dialectic. This notion was not constructed a priori but was already a part of an elaborate cultural and historical milieu that had a Jewish and Christian background. Increased contact with non-Western forms of religious belief and practice initiated the dialectical development just described. This dialectical study continues today as research is progressively extended to new materials, in harmony with the progressively broadened experience of the historian of religions.

This open, dialectical character of comparative historical study will be best served if the actual modalities of this research reflect a consciousness of the analogical character of its categories, especially the category or notion of religion. In this way the distortive effects of the forced employment of univocal terms will be avoided. Dialectical, historical comparison will be free to investigate a wide variety of continuities between different systems of belief and practice without aiming at the reduction of these to a single, comprehensive, univocal referent. The result will be a complex, multidimensional map of religion, a map readable in all directions, on which a given feature or set of features will appear now central, now peripheral. Buddhism, for instance, may appear peripheral when the map is read "theistically" and yet occupy a central place when viewed in terms of its monastic institutions. Similarly, it may appear central when viewed in terms of the widely held belief in transmigration, but more peripheral in its denial of the reality of the transmigrating soul.

Holistic approach

In addition to restricting itself to an analogical use of its interpretative categories in creating such a map of the religious universe, the history of religions must also attempt to be holistic in its approach to its materials. It must study religious beliefs and practices within the specific contexts that give them their full meaning. Within these contexts, functional interpretations will often be of positive value. This is particularly true in the case of simple societies where the functions of religion are relatively undifferentiated from the rest of social life. But the fundamentally analogical character of religion prevents the use of such functional models to explain religion as a whole. As a whole, it remains a concrete whole, and the holisms characteristic of the history of religions must be realized not at the level of univocal theory but at the level of contextualized historical description. This description will aim at the construction of an exhaustive historical typology, a multidimensional map of the actual religious terrain.

The achievement of such an historical typology of religions will result from modalities of historical comparison that must be further specified. In the first place, comparative-historical research in the history of religions must be distinguished from what we may call "idiographic" research, namely, research concentrated on religion in a particular cultural context. Such specific, noncomparative studies are necessary but not sufficient. Nor can the comparative-historical research envisioned here be identified with a systematic, purely formal typology, nor with a phenomenology that neglects issues of origin, growth, and change. The goal is rather the establishment of specific sets of synchronic and diachronic continuities and discontinuities that apply to more than one religion and perhaps to an entire cultural area. In any given case these patterns may be explained either on the basis of cultural diffusion or as the result of independent but parallel developments. In the latter case, parallelism need not signal a unilinear evolution in the history of religions but may rather point to analogies between specific historical and cultural circumstances.

It will often be difficult to decide whether a given pattern is to be explained on the basis of diffusion or parallelism. What is important is to avoid an a priori theoretical option in favor of either. This was the mistake of those who, in reaction to the theories of unilinear evolution typical of the nineteenth century, adopted an equally monolithic hypothesis of universal diffusionism. The fact is that one of the major gains in the field of comparative-historical research has been the discovery of partly similar cultural achievements in the field of religion and culture that are not due to phenomena of diffusion, not even stimulus diffusion. A typical example is the birth and diffusion of polytheism. The connection of polytheistic cults and their typical features (such as theogony, theomachy, specialized sanctuaries and priesthoods, anthropomorphic and hypermorphic gods, etc.) with what are known as archaic high cultures is too evident to be overlooked. It is impossible at the present day to explain all such connections solely on the basis of diffusion, even though a diffusionist hypothesis remains obligatory in some well-defined cases (as, for instance, in the case of the widespread cultural and religious diffusion of cuneiform literature).

Historical Typology of Religion

Having dwelt on some methodological issues concerning the history of religions as a comparative-historical discipline, I shall turn now to a brief overview of the subject matter of this discipline, that is, to the presentation of a concise and inevitably selective historical typology of religion. But first a few preliminary remarks on earlier attempts to map the most general features of the religions may be in order.

In such earlier attempts we meet with notions that owe much to the different cultural epochs in which they were conceived and to the different theories of the origin and development of religion current at the time. Many of these notions are a legacy of philosophical and theological discussion. The best example here is the concept of monotheism, especially as contrasted with polytheism, deism, and pantheism. Many more concepts, including animism, preanimism, animatism, manism, and—more ancient than these—fetishism, owe their origin to the positivistic evolutionary theories typical of the nineteenth century. All of these notions were conceived as expressing primitive or archaic stages of religious thought. As such they presupposed a view of religion as primarily concerned with the mental representation of reality. The cultic and ethical aspects of religion were underestimated, as well as the structural and historical complexities of so-called primitive forms of religion (and it was particularly on primitive religions that the new history of religions discipline focused). This rather exclusive emphasis on the conceptual aspect of religion was typical of nineteenth-century anthropology and should be distinguished from what we may call its rational aspect, which had been emphasized both by Scholastic theology and by the Deism of the eighteenth century. These one-sided excesses of the past should serve as a caveat for today's historian of religions. This does not mean, however, that attention to the conceptual dimensions of religion should be lessened or arbitrarily reduced in turn.

A very different but equally problematical approach to the study of religion, one mentioned above, was the descriptive and interpretative tendency associated with the familiar concepts of "the holy" and "the numinous," a tendency already in existence before Rudolf Otto's famous book The Idea of the Holy (1917, first German edition) gave it a permanent place in religious studies. This new tendency was the opposite of the analytical and conceptual approach just mentioned, and this difference was clearly expressed by Otto's refusal to conceptualize the experience of the holy or to trace it to a specific source, such as God, the gods, spirits, or anything else. Otto's analysis of the essence of religion as the experience of the holy was focused on terms such as immortality, freedom, or the absolute. According to him, man is made intuitively aware of the transcendent reality to which these terms refer through his experience of their opposites on earth. Apart from the similarities of this kind of argument with the ontological argument for the existence of God found in some medieval and modern philosophical systems, the historian of religions must question whether such negative characterizations of the holy are as universal as Otto believed. In fact, they appear to be intertwined, from the point of view of both history and metaphysics, with peculiar notions concerning God, the divine, and religious values, both on the individual level and, as Walter Baetke aptly pointed out, on the social level.

A historical typology of religion should avoid the danger of employing a merely analytical and classificatory conceptualization, particularly if this is intended to reduce the issue of religious ideas and experiences to an issue of mental representations, or to the so-called imaginaire. It should also avoid reliance on selective intuitions in the guise of a scientific phenomenology of religion. Both attitudes ultimately prove to be reductive and ethnocentric. As was made clear in the previous section, a historical typology of religion will serve to map specific sets of analogically related affinities that are neither merely conceptual nor phenomenological but in the best sense historical.

Viewed historically, then, religions can be divided into two broad groups, and this divison will provide the general framework for a genuinely historical typology. This initial division is between those religions that are described as ethnic and those that are founded.

Ethnic religions

Ethnic religions are a part of the culture of a people and do not owe their origin to a historical founder. No single individual has endowed these religious traditions with the unmistakable impress of his personality. Rather, ethnic religions are historical formations that are originally indistinguishable from the formation processes of the cultures and populations to which they belong. Ethnic religions are not restricted to tribal or nonliterate cultures. They may also be found in highly developed literate cultures. Their exact character, which may thus range from unitary to syncretistic, depends upon specific cultural and historical circumstances.

The historical character of ethnic religions, including those of nonliterate cultures, requires that they be studied holistically. Their actual contents and functions in the epochs and contexts for which there is documentary evidence must be closely examined. This must be done without losing sight of the changes and adaptations that these traditions undergo in the course of time, due to internal, developmental tendencies or to influences and stimuli coming from outside.

The fact that even nonliterate ethnic religions are parts of a larger world means that the purely descriptive or "idiographic" study of these traditions cannot do without a comparative outlook. Comparison becomes inevitable when the scholar's attention is drawn to the cross-cultural affinities or "family resemblances" that link the cultural and religious expressions of different countries and areas.

Approaches to comparison

Two main approaches to comparison may be distinguished in the history of religions. The first developed in the nineteenth century, and the other took shape at the end of that century and at the beginning of the twentieth. Each of these warrants a brief discussion before I continue.

The first approach, practiced in the nineteenth century, was inseparably linked to evolutionary thought and to the elaboration of general theories concerning the origins and the growth of religion on a world scale. It was on this basis that notions such as animism and animatism were introduced into what was understood to be scientific research (hence the German name of this field, still with us, of Religionswissenschaft ). These notions were intended to apply cross-culturally, indeed universally. There is no need to repeat what I have already said about the negative reductive tendencies implicit in such an approach. It was of course highly reductive and arbitrary, insofar as its unilinear evolutionist presuppositions were effective. But it also had some positive effects, inasmuch as it made possible the identification of conceptual, ritual, behavioral, and ethical affinities or continuities that transcended such older categories as idolatry, paganism, and superstition, which were no longer suitable for a descriptive approach.

The rise of a descriptive phenomenology of religion contributed to a further refinement of these new patterns. Take, for instance, the case of the notion of shamanism. Once shamanism was differentiated from the generic notion of animism and considered not only as a peculiar element of religious behavior but also as an element of a structure implying a cosmology and a worldview, it could contribute to the transition from an evolutionistic outlook too fond of concepts and representations to a cultural-historical study based on the discovery of cultural wholes and cultural areas. In other words, the elaboration of a more rigorous, multidimensional, and descriptive phenomenology of religion allowed the history of religions (Religionswissenschaft ) to survive the inevitable crisis of evolutionism and its universal and unilinear stages and to enter a more rewarding phase characterized by a new form of comparison, namely cultural-historical comparison.

This new form of comparison proved effective when, at the beginning of the twentieth century, cultural-historical research was able to concentrate on area studies, as in the pioneering works of Bernhard Ankermann and Leo Frobenius (on Africa), Fritz Graebner (on Oceania), and Franz Boas (on the North Pacific). On this basis it was possible to connect phenomenological and historical research in the study of a particular group of ethnic religions belonging to nonliterate cultures. In this way the originally negative term nonliterate acquired positive content. Nonliterate ethnic traditions became definable on the basis of undeniable categorical continuities (e.g., shamanism) that were at the same time specific to particular areas. As a result, the so-called primitive religions ceased to be studied in terms of a procrustean bed of an alleged, indiscriminate primitivism. Such attributions now had to be demonstrated on the basis of cultural-historical inquiry. In this way the study of nonliterate religions entered with full rights into the field of religio-historical research proper. At the same time, the historian of religions was not obliged to renounce his fundamental interest in comparison and (where appropriate) study of historical development, two aspects that had been neglected by the heirs of the older anthropological methods and the practitioners of an exclusively functional social anthropological research.

High cultures

I have already noted that ethnic religions are found among literate as well as nonliterate cultures. In particular, the high cultures of antiquity produced ethnic religious traditions that make special demands on the historian. Study of them requires a philological competence that clearly excludes any simplistic or superficially phenomenological approach. It remains true, however, that even the historian of religions lacking such specialized knowledge can still contribute to a fuller understanding of these religions on the basis of his typological-historical experience. Take, as an example, the discovery of some of the classical characteristics of the demiurgic trickster in such diverse figures as the Greek Prometheus, the ancient Egyptian Seth, and Yurugu, also named Ogo, of the Dogon of West Africa. This discovery and the accompanying insight into the dualistic cosmology that provides the backdrop for such figures would have been impossible for the classicist or Egyptologist working only within his own speciality. It resulted rather from the comparative method of the historian of religions. This comparative approach can be particularly successful in the comparison of the mythologies of cultures that belong to the same subcontinent but have had different histories. It could, for instance, be expected to shed much light on the difficult question of the relationships between ancient Egyptian culture and other African cultures, both in the sense of an African substratum of Egyptian culture and of an Egyptian influence on sub-Saharan Africa.

Another fascinating problem for comparative research concerns the continuities and discontinuities between the nonliterate cultures and religions on the one hand and the high cultures and their religious systems on the other. Given the differences that exist between them as well as the differences that exist within the respective traditions themselves, the question of their mutual interrelationships becomes quite complex. Particularly important is the question of the partial continuities that exist between the high gods of some nonliterate cultures and the heavenly deities who head the pantheons of some of the typical polytheistic formations in the high cultures of antiquity. This kind of partial continuity, which sometimes extends to founded and universalistic religions as well (as in the case of the God of heaven among the Israelites and the high divinity of pre-Islamic Arabia), is one of the most interesting phenomena in the history of religions.

Monotheism and polytheism

Closely connected to the question of continuities between cultures is the question of the exact nature of monotheism. This question has been the subject of a long debate within the history of religions. Some have claimed to find monotheism in the religions of nonliterate cultures, particularly among hunters and gatherers. Others, in contrast, have viewed it as a late phenomenon in the process of evolution, or even, as Raffaele Pettazzoni put it, as a revolutionary stance against a preceding form of polytheism. Actually these alternatives cannot be formulated so rigidly. The explicit and polemical formulations of monotheism that we find in the Bible and the Qurʾān have little in common with the high gods of contemporaneous nonliterate cultures. Although the latter are not necessarily dei otiosi, severed from the cult and from the world of men, they are nevertheless far from the intense dynamism of the God of the Bible and the Qurʾān. It must be admitted that not all religious complexes extraneous to polytheism are ipso facto monotheistic. The most developed polytheistic systems did not "evolve" in the direction of monotheism, nor did they express a revolutionary movement in that direction. Rather, they tended to become progressively monistic, elaborating the notion of a deity who is "polyonymous" (as is Isis in the aretalogies of the Hellenistic period) or is pantheos, a god containing all the gods. In its most mature form, such a deity was conceived as "theopantistic," that is, as identical with the cosmos and at the same time transcending it. Such is the picture of Zeus presented in an Orphic poem: "Zeus is the beginning (or head), Zeus is the middle, from Zeus all things came into existence," or in a fragment of Aeschylus: "Zeus is all these things, and what is beyond them." This assessment of Zeus bears comparison with the theopantistic speculation concerning the figure of Purusa in Vedic India.

It is important to realize that polytheism, as a historical type of religion, is much more specific than a merely formal notion of a plurality of gods. In fact, polytheism is not found in all types of culture, but is specifically linked to the high cultures of antiquity (and also to some in modern times, particularly in East and South Asia). These cultures characteristically possess an advanced form of cereal agriculture and show a degree of social stratification, with an attendant differentiation of classes, professions (including scribes), a priesthood and nobility, established sanctuaries, and the like. The cosmos itself is represented as having a departmental organization. Such cultures are found in the ancient Mediterranean area, and in West, South, and East Asia. Further instances can be found in the Nordic countries of Europe, in Central America and Peru, in the medieval kingdoms of the Sudan (owing to Mediterranean influences), and to an extent in Polynesia. The social and historical specificity of polytheism as a religious type would seem to require that opposition between it and monotheism become acute only in particular historical and religious situations. Such situations include the historical emergence of the Hebrew people and their religion in the Near East, the vogue of the cult of Ahura Mazda in Iran, the rise and diffusion of Christianity in the Mediterranean world, and the preaching of Muhammad against the religion of the pre-Islamic Arabs. Clearly, the problem of polytheism and its relation to monotheism is not to be solved on the basis of a general phenomenological "stratigraphy" of religion and its main forms, any more than it was solved by a unilinear evolutionism. It is not a question of relative anteriority between polytheism and monotheism, homogeneous in themselves, but of specific, noninterchangeable historical formations that can interact with different types of religious and cultural organization. Monotheistic formations may present themselves either as immemorial tradition, as a novelty, or as the message of the one God and his triumph over false deities. Other formations could have evolved into pantheons that inspired and also reflected the complex organization of the high cultures of antiquity. Roman religion, for instance, is the product of many different cultural influences, having been almost completely reshaped in the course of time on the basis of long-lasting Greek influence (particularly in its pantheon). Nevertheless, it never lost its continuity with its most primitive expressions. Among these was the notion of a heavenly god, Jupiter, who had an Indo-European heritage. This god was to dominate Roman religion under the name of Optimus Maximus, yet this did not prevent him from being assimilated to the supreme god of the Greek pantheon, Zeus, and to all types of religious, ethical, philosophical, and symbolic notions.

I have already had occasion to note that the general type that I have identified as ethnic religions and contrasted with founded religions admits of a certain amount of internal diversity. Thus ethnic religions need not be those of nonliterate tribal societies but may also be found in more complex literate societies. I shall now introduce two further subtypes of ethnic religions, those that are scriptural and those that are national.

Scriptural and national religions

Some ethnic religions are in fact characterized by the existence of sacred scriptures, organized on the basis of a "canon." Acceptance of these is considered an essential aspect of religious affiliation. The outstanding example is the function of the Vedas in Hinduism. Ethnic religions possessing scriptures exist in a situation midway between the nonscriptural religions of tribal societies, where religious affiliation is indistinguishable from the simple fact of social life, and at the other extreme, the universal religions, where the individual as such becomes a convert to the "good news" of a prophetic message written down in a book.

Another important subtype among ethnic religions is that of the national religions, those cults that promote a national and political consciousness. This is the case with state Shinto as practiced in some periods of Japanese history, with some forms of Hinduism, with Zoroastrianism in the Sasanid empire, and with the official cults in ancient Rome, such as the cult of Capitoline Jupiter and the cult of the emperor. It must be added that some of the founded religions may initially embrace a national outlook. Islam, for instance, may have been conceived originally as a prophetic message addressed to the Arab nation, although, to be sure, it is considered to be the final form of the historical revelation of God. The situation is similar in Judaism, whose universalism was mediated through the entire course of Hebrew history. There were also particular occasions when Christianity, though possessed of a universalist message, became bound up with specific national, cultural, and political interests, as, for instance, those of the Byzantine empire.

Founded religions

I shall now look more closely at the founded religions, which include, in addition to Judaism, Christianity, and Islam, the religions of Zarathushtra (Zoroaster) and the Buddha, the religion of the Sikhs, and, with certain qualifications, the plethora of prophetic-nativistic cults and the "new religions." These traditions, which all trace their origin to a specific historical founder, raise fundamental historical questions. When a religion is founded, this obviously takes place in a cultural milieu already characterized by specific religious notions and institutions. The relation of the prophet or founder and his original followers to this milieu may vary considerably, depending on whether the preceding religious environment had been predominantly of the ethnic type or had already known prophets and forerunners. In general there is a vivid sense of novelty in the foundation of these religions, sometimes even a revolutionary novelty. The founder nevertheless often conceives his mission as in part a restoration of primordial values long since lost. In some cases this question of partial continuity is decisive and helps illuminate the religious quality of the founder's message. This can be seen in the case of Buddhism. Many of the notions fundamental to Buddhist doctrine, such as dharma, karman, saṃsāra, and mokṣa, already played an important religious role in India at the time of the Buddha. An understanding of the Buddha's appropriation and modification of these terms is thus of great importance for a typological assessment of his message as a "religious" message, bound up with the religious traditions that preceded him.

Among the founded religions the universal religions stand out as a definite subtype. These religions are based on a universal message of salvation, not limited to any particular group, ethnic or otherwise. They are characterized both by eschatological and otherworldly perspectives and by strong this-worldly ethical and social commitments. Their message is addressed to the individual and demands conversion and adherence to a religious community that, in sociological terms, may be described as a church. This community typically undergoes a rapid initial expansion, sometimes suffers persecution, and actively engages in missionary activities aimed at making new converts. These characteristics clearly distinguish the universal religions from the ethnic and national religions, even though, as noted above, universal religions may occasionally become closely associated with specific cultures.

The emphasis on a universal message and personal conversion in response to it, an emphasis that transcends all racial and social barriers, differentiates the universalistic religions from the cults of antiquity that were sometimes their rivals. These cults, such as the mystery religions, would be better described as cosmopolitan rather than universal. Unlike commitment to any of the universal religions, participation in such cults could coexist with whatever other religious commitments an individual might have had, such as to the gods of the tribe, city, or state; the universal religions, however, demanded the individual's total allegiance.

This contrast becomes less clear in the case of the syncretistic tendencies found at the popular levels of the universal religions. Christianity and Islam, for instance, have been influenced, in some of their popular or ethnic manifestations in South America and Africa respectively, by local traditions of animism or "spiritism." The same lack of a clear differentiation between the universal and the local is found in the prophetic-nativistic cults that reinterpret the message of the great universal religions in strictly local terms, although not infrequently these same cults are inimical to preexisting forms of local magic and sorcery.

The case of Buddhism is somewhat unique. Although it displays many of the features that were attributed above to the universal religions, it nevertheless resembles the mystery cults of antiquity in its ability to coexist, in the belief system of a single individual and in a single cultural milieu, with other forms of locally preexistent religious belief and practice. This is especially evident in contemporary Japan, where a single individual may have a double allegiance to Buddhism and Shinto, according to the circumstances. The rather peculiar status of the universalism of Buddhism is linked to the equally peculiar status of Buddhism as a "religion."

Finally, another historical type of religion is comprised of those mysteriosophic (i.e., Orphic) and Gnostic movements of antiquity and the Middle Ages that drew heavily on the universal religions of Christianity, Islam, and Zoroastrianism, borrowing many of their basic terms but totally reshaping them to suit their own needs. This procedure is found as well in the scientistic theosophy of some contemporary Gnostics, whose reinterpretations of the basic tenets of different religions are for the most part superficial.

Bibliography

For further discussion of the questions raised in this entry, the reader is referred to Problems and Methods of the History of Religions, edited by Alessandro Bausani, C. Jouco Bleeker, and myself (Leiden, 1972), to my book The History of Religions, (Leiden, 1975), and to The History of Religions: Retrospect and Prospect, edited by Joseph M. Kitagawa (New York, 1985). As general references on the history of religions, the following works are also recommended.

Baaren, Th. P. van, and H. J. M. Drijvers, eds. Religion, Culture and Methodology. The Hague, 1973.

Baird, Robert D. Category Formation and the History of Religions. The Hague, 1971.

Banton, Michael, ed. Anthropological Approaches to the Study of Religion. London, 1966.

Bianchi, Ugo. Probleme der Religionsgeschichte. Göttingen, 1964.

Eliade, Mircea. The Quest: History and Meaning in Religion. Chicago, 1969.

Eliade, Mircea, and Joseph M. Kitagawa, eds. The History of Religions: Essays in Methodology. Chicago, 1959.

Evans-Pritchard, E. E. Theories of Primitive Religion. Oxford, 1965.

Graebner, Fritz. Die Methode der Ethnologie (1911). Reprint, Oosterhout, 1966.

Honko, Lauri, ed. Science of Religion: Studies in Methodology. The Hague, 1979.

Kitagawa, Joseph M., ed. The History of Religions: Essays on the Problem of Understanding. Chicago, 1967.

Lanczkowski, Günter. Religionswissenschaft als Problem und Aufgabe. Tübingen, 1965.

Lanczkowski, Günter, comp. Selbstverständnis und Wesen der Religionswissenschaft. Darmstadt, 1974.

Lang, Andrew. The Making of Religion. 3d ed. New York, 1909.

Pettazzoni, Raffaele. L'essere supremo nelle religioni primitive. Turin, 1957.

Pinard de la Boullaye, Henri. L'étude comparée des religions. 4th ed. 3 vols. Paris, 1929–1931.

Puech, Henri-Charles, ed. Histoire des religions, vol. 1. Paris, 1970.

Pye, Michael. Comparative Religion. New York, 1972.

Rudolph, Kurt. Die Religionsgeschichte an der Leipziger Universität und die Entwicklung der Religionswissenschaft. Berlin, 1962.

Rupp, Alfred. Religion, Phänomen und Geschichte: Prolegomena zur Methodologie der Religionsgeschichte. Saarbrücken, 1978.

Schlette, Heinz Robert. Einführung in das Studium der Religionen. Freiburg, 1971.

Schmidt, Wilhelm. Der Ursprung der Gottesidee, vol. 1. 2d ed. Münster, 1926.

Sharpe, Eric J. Comparative Religion: A History. London, 1975.

Smart, Ninian. The Science of Religion and the Sociology of Knowledge: Some Methodological Questions. Princeton, 1973.

Waardenburg, Jacques. Classical Approaches to the Study of Religion. 2 vols. The Hague, 1973–1974.

Widengren, Geo. Religionsphänomenologie. Berlin, 1969.

Ugo Bianchi (1987)