410 - Anthropology and Comparative Religion Turner

Anthropology and Comparative Religion
By Mary Douglas and Edmund F. Perry

"Originally, comparative religion was freely used to buttress theologians in their nineteenth century debate against encroachments of science and Enlightenment on their authority…. One of our methodological questions is whether motif-research as [Anders] Nygren conceived it, that is, as a method for theological inquiry, could ever be used for comparative religion."

THE MAGNITUDE of information about religion is exponentially increased when we combine the accumulated resources of anthropology and comparative religion. In our thinking about the multifaceted character of religion and the perplexities which beset all those who want to study it objectively, we start by recognizing a basic distinction. On the one hand, there are fieldworkers, and, on the other, there are textualists. Each discipline fosters separate interests.

The textualists examine written hymns, sermons, and mythology. These records indicate to them what kinds of questions comparative religion should address, and this emphasis further restricts what kinds of written texts they study and what kinds they reject. The fieldworkers, often seeming to be bereft of the materials preferred by the textualists, can actually hear myths being recounted, hymns being chanted, and sermons being preached, and in addition they see ceremonies enacted. They also have access to the background for interpreting religious texts.

Mary Douglas, a social antropologist, is the Avalon Professor of the Humanities at Northwestern University. She has published Purity and Danger: An Analysis of Concepts of Pollution and Taboo (London: Routledge and Kegan Paul, 1966), Natural Symbols: Explorations in Cosmology (New York: Pantheon, 1982), and recently co-edited Religion and America: Spirituality in a Secular Age, with Stephen M. Tipton (Boston: Beacon, 1982).
Edmund F. Perry is Professor of the Comparative Study of ReIigions at Northwestern University. He recently completed with Ross Reat a book length essay, The Phenomenon Called God. For the past five years be has been researching the religious formation of contemporary Buddhist monks in Sri Lanka.

The appointment in 1981 of Mary Douglas to the Avalon Chair of the Humanities at Northwestern was part of that institution's intent to connect the social sciences to the humanities. One part of the agenda was to engage anthropology and comparative religion in professional conversation. This paper was originally presented by Douglas, and Perry at the American Academy of Religion Annual Meeting, Dallas, Texas, December 19, 1983.

411 - Anthropology and Comparative Religion Turner

This background would normally be available to the textualists in the form of legal documents, wills, and administrative edicts.

Such a difference in methods and materials ought to be complementary, but in practice the textualists tend to overvalue the religious texts and the fieldworkers to undervalue them, and consequently they pursue different questions. To reconcile these divergent emphases, we have thought to go back to three remarkable Scandinavian religious scholars. Born a century ago, their work rises above these current differences since they clearly belong with the textualists but adopted the fieldworkers' favored questions. Wilhelm Gronbech (b. 1873), Johannes Pedersen (b. 1883), and Anders Nygren (b. 1890) were committed to the assumption that the full social context is necessary to understanding the written text.

COMPARATIVE RELIGION:
ONE SCANDINAVIAN TRADITION

Pedersen's four-volume Israel-Its Life and Culture¹ is an ethnography presented almost as if be had come back from a sojourn among contemporary exotic people as a fieldworker impressed with the systematic structure of their values. He is able to give an account of honor and shame, purity, justice and holiness, and also, of course, of sacrifice, covenant and God, only by giving at the same time, an account of the institutions of the family, the law, the city, war, chiefs, and kings. The narratives of the Old Testament are submitted to attentive philological criticism. He did not naively suppose that anything in the narrative could mean the same to us as it meant to its writers. Accordingly, be went to pains to expound their central concepts, such as soul, life, blessing, and sin. That this was already far too simple an approach to the interpretation of mythology we know now, thanks to the fieldworkers who have turned their attention to the Bible as well as to scholars specialized in Near Eastern texts.

But where did Pedersen get his sociological interests and empirical openness? Some of it certainly from Wilhelm Gronbech, author of The Culture of the Teutons,² whom Pedersen acknowledged as colleague and friend. Both trained in philology, both at the University of Copenhagen, both well placed to hear the intellectual debates of the turn of the century, each presented his chosen culture as a coherent, intelligible way of life.

Gronbech taught that a philosophical idea exists as part of a system of beliefs. Writing at the time of the maximum prestige of Gestalt psychology, he assumed that understanding of religion required comprehension of the whole system. As one interpreter of Gronbech observed, "He depicted the essence of a system of philosophical belief through an

¹ Johannes Pedersen, Israel: Its Life and Culture, Vol. I (Copenhagen, 1920), Vol. II (Copenhagen, 1940).
² Wilhelm Gronbech, Culture of the Teutons, Vol, I (London: Oxford University Press, 1931).

412 - Anthropology and Comparative Religion Turner

exhaustive inquiry into the meaning, interrelationship, and metaphysical connotations of basic assumptions incorporated in certain words and concepts. He posited the scientific understanding of spiritual phenomena as resting upon the comprehension of the whole, of which those phenomena are only a part."³ But he thought that the assumptions and beliefs formed a patterned whole because the believers were living their daily lives in a society that functioned as an organism.

So here we have it, both Gronbech and Pedersen were committed to the contemporary search for essences and organic wholeness. This is the same background in which social anthropology emerged in Britain in the 1920s. No wonder these textualists writing about vanished peoples strike a sympathetic chord for contemporary fieldworkers who trace their own origins to Radcliffe Brown and Bronislaw Malinowski. It was Gronbech, but it could just as easily have been the anthropologists, who taught that religion could be understood only through an exhaustive inquiry into the interrelationships of basic assumptions incorporated in certain key words and concepts.

The real originality of these textualists shows by comparison with another treatment of myths which they opposed. Many have mocked Max Müller, but few so cuttingly as Gronbech.⁴ He reproached the -comparative religion of his day for a double dose of romance. On the one hand, the folklorists had found the Teutonic heroes in the shape and style of Classical Greece. (His essay on the political implications of Jacob Grimm's work is echoed today by Michael Herzfeld's study of the political uses of the folklore movement in Greece itself.)⁵ On the other hand, mythologists of his day "clad the old myths in their own profundities; a little pantheism, and some natural science, a little shallow poetry and some tenacious moralizing … dependent on traditional thoughts and forms but lacking their force, longing for something new but impotent to create." Above all, Gronbech was protesting the super-imposing of meanings upon dead forms, and advocating research that scrupulously sought to find the forms' original life. His own account of the Norsemen uncovered a distinctive way of living and thinking. Since the issue of meanings super-imposed upon alien religious forms is always a problem for comparative religion, Gronbech provides a more appropriate starting point for this essay.

NYGREN'S PHILOSOPHY OF RELIGION AND
THEOLOGY

Philosopher, theologian, historian of religion, and Bishop of Lund, Anders Nygren, stands in the straight path of this tradition. He shared Pedersen's and Gronbech's conviction that a historic culture has a unity

³ P. M. Mitchel, "Introduction" to W. Gronbech, Religious Currents in the Nineteenth Century, trans. W.D. Paden and P.M. Mitchell (Lawrence: University of Kansas Press, 1964), p. 9.
⁴ See his heavy irony in the chapter on "The New Germany" in Gronbech, Religious Currents, pp. 116-119.
⁵ Michael Herzfeld, Ours Once More: Folklore, Ideology, and the Making of Modern Greece (Austin: University of Texas Press, 1982).

413 - Anthropology and Comparative Religion Turner

and logical coherence of its own, and he evidently got the idea from the same sources. (He actually wrote a study of the German physiologist and psychologist, Wilhelm Wundt.⁶) More exposed to philosophical currents of thought than the Danish historians of religion, he meant to go much deeper into the search for a particular structure supporting a particular cultural system. As a Christian minister, he was concerned to protect the status of religious knowledge from the attacks on metaphysics that extended from Hume through Kant to the Austrian linguistic philosophers.

(1) Nygren's Program. Nygren shared the general fascination of his period with the idea of the a priori. It, was fashionable to show that Kant had not said the last word about the fundamental preconditions of experience. Henri Bergson argued that consciousness of time had been neglected in philosophy by comparison with spatial categories; Durkheim had added a new fundamental category, society.

Nygren tried to demonstrate that religion was the necessary precondition of all discourse. By this one dexterous stroke he hoped to restore the secure foundation of religious knowledge. He aimed to establish a critical philosophy based "on experience and nothing but experience,"⁷ and not on questionable metaphysical speculation. His philosophy of religion would not be searching for an "ultimate reality" beyond space and time. He attacked metaphysics as severely as any rationalist might attack religion, confident that he had created a logical and scientific basis for religious philosophy. He was not offering a science of experience, but a science of the analysis of the presuppositions of experience.

Such a science proceeds as a critical examination of language and presuppositions. It engages, first, in a "critical analysis of concepts" so that philosophers can be quite clear about what they are thinking and talking about and can eliminate both ambiguity and confusion from their communication. When the critical analysis of language has been made, the philosopher proceeds, second, to a "logical analysis of presuppositions" which leads, third, to "a general theory of experience," or to what is really the same for Nygren, a "general theory of validity." Logical analysis of presuppositions identifies the basic concepts which constitute the fundamental categories of experience. These are the concepts which, though we may not be conscious of them, are logically presupposed in all meaningful discourse. The validity of each of these presuppositions is demonstrated by showing that it must be assumed necessarily if we are to entertain the idea of any validity at all.

By this method Nygren concludes that there are four main conceptual frameworks of experience, "contexts of meaning," or "universes of discourse"⁸ The four which have predominated throughout human

⁶ Anders Nygren, "Wilhelm Wundt," in Kirkeleksikon for Norden 1926.
⁷ Anders Nygren, Filosofisk och kristen etik (Lund: C.W.K. Gleerup, 1923), p. 107.
⁸ Nygren's method does not depend on the strict repetition of certain technical words and phrases but on the faithful translation of concepts. Bernard Erling correctly points out that Nygren consciously related his philosophical terminology to the changes occurring in

414 - Anthropology and Comparative Religion Turner

history are science, morality, art and religion. As Philip Watson has observed,⁹ these four universes of discourse are not four different languages of discourse; they are rather four different uses of language. They are not mutually exclusive. Any person may experience all of them at the same time. Yet each has its own question: for science, the question of truth; for morality, the question of good and evil with the closely related question of right and wrong; for art, the aesthetic question of beauty and the sublime. Finally, Nygren adds religion with the question of the eternal. Erling remarks that "these questions represent inescapable presuppositions derived from critical analysis of given areas of experience without which these areas of experience cannot be properly understood."¹⁰ Furthermore, Nygren insists that there is no overlap: no one of these inquiries can do duty for the other.

(2) Religion and "the Eternal." The force of this strong partitioning establishes religion as an independent form of experience with its own context of meaning or universe of discourse. Religious experience is distinguished from the other domains of experience by its reference to the eternal.11 As the "fundamental presupposition" of the "category of religion," the question of the eternal cannot provide religious content, and should not. "It allows every religion to-give its own characteristic answer. The category itself prescribes nothing as to what the answer must be. Knowledge of the answer can be obtained only by going to the actual religions." 12 Religious content can derive only from experience itself, not from philosophical analysis of presuppositions. As the logical presupposition of religious experience, the eternal is a conceptual referent indicating something which transcends our sensible, finite, and temporal existence.

By giving to religion the domain of the eternal, and by placing the eternal thus far back in the hierarchy of inescapable presuppositions,

philosophical, discussion in the period between 1921 and 1970 (Erling, "Motif- Research and Historical Method," in Charles W. Kegley, ed., The Philosophy and Theology of Anders Nygren. Carbondale: Southern Illinois University Press, 1970, p. 102).
In 1921, Nygren published a major work entitled Religiost a priori: dess filosofisk forutsatiningar och teologiska konsekvenser (Lund: Gleerupska Univertsitetsbokhandeln, 1921), but by 1958, he discussed "The Religious Realm of Meaning," (The Christian Century, July 16, 1958) without recourse to the term a priori. In his last major work, Meaning and Method: Prolegomena to a Scientific Philosophy of Religion and a Scientific Theology (London: Epworth Press, 1972), he refers to this term only in its historical significance. In this latter work, Nygren comments on his terminology and points out that in his earlier writings he used the term, "areas of experience," but finding it so easily misinterpreted, he abandoned it "in the late twenties … in favor of the term "contexts of meaning," which is both more adequate and less liable to be misconstrued." (p. 269 n.5)
⁹ Philip S. Watson, "Systematic Theology and Motif-Research," in Kegley, p. 226.
¹⁰ Erling, p. 109.
¹¹ CL Nygren's discussion of "the question of the eternal" in Essence of Christianity (London, 1960), pp. 25-48, and "the category of the eternal" in both Religiost a priori and Meaning and Method. (passim).
¹² Meaning and Method, p. 343.

415 - Anthropology and Comparative Religion Turner

Nygren thought to have played a trump card in the epistemological game, as he understood the rules. All other domains of discourse use a temporal framework. This implies that all these other domains are dependent on a deeper, more abstract structure of presuppositions in which time can be known because of the possibility of a non-temporal order. The idea of temporality depends upon some notion of the eternal. Instead of being rejected, residual, with no secure philosophical standing, religion has now been defined as the universe of discourse from which morality, science, and aesthetics take their start.

It is important not to make a clumsy caricature of Nygren's exposition. Superficially, his argument seems to follow the form of the old scholastic arguments for the existence of God, which he repudiated. But he is not using logic to prove that anything exists; he is trying to prove the logical necessity of a kind of discourse, about the eternal, and to establish its necessary logical priority before all other kinds of discourse. Paul Holmer is rather unfair when he says that Nygren "was bewitched by the thought of a logic, a more ideal structure, open to a philosophical discernment." ¹³ It is true that he really believed in the possibility of establishing a universal structure of discourse which could be uncovered once and for all and whose logical necessity would be irresistible. At the time that he was writing his Ph.D. dissertation on the religious a priori, it was a fashionable quest. Many thinkers were looking for some kind of hard-wiring between logical structures and physiology, or for permanent, invincible structures in logic itself. ¹⁴ But Nygren cannot be accused of a form of Platonic idealism that attributes more reality to ideas than to things. Nor was he a skeletalist, to use Ernest Gellner's felicitous phrase: "The central idea in skeletalism is that we somehow have access to the basic form of thought or being, and through this can see the general form of our world."¹⁵ Nygren expressly denied that reality must conform to the logical form of thought.

The career of Nygren is characterized by a concern for the use of scientific methods in the philosophical, historical and theological study of religion. This follows from his rejection of all metaphysics. He held with Kant that there can be no science of the transcendent. All science anchors itself firmly to the actualities of space and time and seeks to gain a clear and accurate understanding of those actualities. In Nygren's view every genuine science has: (a) a distinctive subject matter to investigate, (b) a distinctive interest in that specific subject matter so that inquiry does not duplicate some other science, and (c) a method of inquiry appropriate both to its own purpose and subject matter. There is no one method for all inquiries and all subjects to be investigated.

¹³ Paul Holmer, "Nygren and Linguistic Analysis: Language and Meanings," in Kegley, P. 91.
¹⁴ S. Jay Gould, The Mismeasure of Man (New York: W.W. Norton, 1981), p. 291.
¹⁵ E. Gellner, The Legitimation of Belief (Cambridge: Cambridge University Press, 1974), pp. 66-67.

416 - Anthropology and Comparative Religion Turner

"There is nothing which can be called 'the scientific method,' but several different methods, adapted to different objects which the various sciences investigate." ¹⁶

We have noted the intellectual strategy which led Nygren the critical philosopher to find the concept of the eternal as the fixed point with which everything that can be called religion is connected. With the thought of the eternal as his lodestar, Nygren the historian of religion turned to the empirical record of humanity's religious history. In all of the bewildering variety-historical, geographical, and systemic-the question of the eternal emerged everywhere in four more or less constant forms. Wherever religion appears, (a) it claims to unveil the eternal, (b) it explains the disquiet and tension which arise from the disclosure of the eternal, (c) it provides a means of overcoming the estrangement created by the disclosure of the eternal, and (d) it provides a way of life or fellowship or union with the eternal. These four elements or characteristics constitute for Nygren the essential structure of religion, the human experience of and relation to the eternal. The content of each characteristic derives from the actualities of each particular religion which forms its own complete constellation of meaning.

The focus on the eternal is at least as good an indication of the subject matter -of comparative religion as -the definitions currently in favor. Many define religion by experience of the transcendental. ¹⁷ But the word eternal, while covering the same ground, has the merit of not pre-empting any position between religious doctrines of immanence and transcendence. Others, including most anthropologists, define religion as belief in supernatural or superhuman beings, ¹⁸ but this includes all the spooks and goblins imaginable without including any specifically religious feelings or values. Others, again, prefer not to define at all, but indicate what religion is by its functions: what religions do for the emotions, for thought, for the culture.¹⁹ Here again, religion as commonly

¹⁶ Anders Nygren, "Reply to Interpreters and Critics," in Kegley, p. 353. It was Nygren's professional obsession to develop techniques of inquiry appropriate to the distinctive character of religious phenomena and meanings. Wilfred CantwelI Smith has commended the same objective in his proposal of comparative religion as a "humane science" in which human persons study other human persons in their relation to transcendence. See Smith, Towards a World Theology (Philadelphia: Westminster Press, 1981), pp. 56-104.
¹⁷ Most notably, among contemporary scholars of religion, Wilfred Cantwell Smith has advocated the usefulness of "transcendence" in the attempt to arrive at any understanding of religion in some generic sense. "By 'transcendence' I mean formally a reality that transcends the immediate mundane" (Smith, Faith and Belief, Princeton: Princeton University Press, 1979, p. 161). This term and concept figure in his books beginning with The Meaning and End of Religion (New York: The Macmillan Co., 1962), and continuing to his recent Towards a World Theology.
¹⁸ Jack Goody, "Religion and Ritual: the DefinitionaI Problem," in British Journal of Sociology 12 (1961): 142-164; Melford Spiro, "Religion: Problems of Definitions and Explanation," in Anthropological Approaches to the Study of Religion, ed. Michael Banton (London: Tavistock Publications, 1966), pp. 85-126; and Ake HuIkrantz, "The Concept of the Supernatural in Primal Religions," History of Religions 22 (1983): 3.
¹⁹ S. F. Nadel, Nupe Religion (London: RoutIedge and Kegan Paul Ltd., 1954), and Clifford C. Geertz, "Religion as a Cultural System," in Banton, pp. 1-40.

417 - Anthropology and Comparative Religion Turner

understood tends to slip between the meshes: on the one hand, many other beliefs perform these functions;²⁰ on the other hand, this functional approach tends to leave out religion in modern industrial society, where the processes of secularization have eroded all the culturally unifying functions.

So there is a lot to be said for Nygren's definition. It works just as well for textualists as for fieldworkers. Eternal refers to laws which operate on other principles than our time-space-framed reality. It would include any class of beings endowed with bilocation, levitation, invisibility, extra-sensory knowledge of the hidden past and future.

(3) Theology and Motif-Research. While Nygren's program has the advantage for our conversation of making comparative religion and anthropology complementary, it separates them from theology, which, again, is just as well. Nygren locates the work of the theologian in a single religion. Restricted in this way, theology can fulfill Nygren's conditions for a "scientific inquiry." It has its own subject matter to investigate, namely, the particular religious tradition. Like every other science it has a distinctive interest in its subject matter which is peculiarly its own. It does not duplicate the work of history or comparative religion. It centers its interest on the question of what a religious tradition is in all of its manifestations: its fundamental character, its distinctive features, and its unity.

Nygren proposed that theology can meet the third requirement of his idea of a science by employing motif-research as a method of inquiry appropriate to its purpose. Motif-research aims to discover the defining core or growth center of a religion (or any given constellation of meaning). It examines the social and historical evidence in hand for the understanding of a religion; it forms an hypothesis as to "the fundamental motif" (which may be something more or other than a single concept) amid all of the diverse expositions and complex manifestations of this religion; and, finally, it tests this hypothesis by checking it against the evidence. What we conclude to be the basic motif must not be a matter of subjective and arbitrary judgment; it must be open to objective examination and verification.

In Agape and Eros, ²¹ we have Nygren's application of motif-research to Christianity. With the question of the eternal as the categorical question of religion, Nygren turns to the historically given realities of Christianity to identify the fundamental motif which specifically characterizes the Christian relation to the eternal. In Christianity, relationship with the eternal occurs as fellowship or communion with God. From

²⁰ Robin Horton, "A Definition of Religion and its Uses," Journal of the Royal Anthropological Institute (1960): 90.
²¹ Anders Nygren, Agape and Eros: A Study of the Christian Idea of Love, Part I, trans. A.G. Hebert (London: SPCK, 1932); Agape and Eros: The History of the Christian Idea of Love, Part II, trans. Philip S. Watson (London: SPCK, 1939).

418 - Anthropology and Comparative Religion Turner

its beginning to the present, Christianity appears as the gospel of God's gift of himself in Christ for fellowship with humanity.

Nygren asks of his historical material, "What is the nature of this fellowship and what is its general structure?" The answer to this question will be the fundamental Christian motif which gives definitive and distinctive' character to Christian fellowship. Throughout its varied cultural and historical expressions Christianity manifests the permanent element of fellowship with God on the basis of Jesus Christ. In the New Testament accounts of Christian origins, Nygren finds that Jesus Christ (a) reveals the eternal God, (b) constitutes the norm of judgment by which humans discern their unworthiness for fellowship with God, (c) is himself the enabling power by which unfit humanity is atoned with God, and (d) is the real presence of God with whom Christians have communion. This structure means that God establishes Christian fellowship. Nothing in humanity, including sin, can be regarded as motivating God to enter into fellowship. It is the nature of God to institute fellowship: God is agape, "love." The fundamental Christian motif is the agape-love motif: unmotivated and spontaneous because God is love; impartial and unconditional and, therefore, addressed alike to the rich and the poor, the proud and the complacent; creative, giving infinite value to its recipients who otherwise would be deprived of value; and initiating, God making the first and definitive move to establish fellowship with sinful humanity.

This agape motif contrasts strikingly with the nomos motif in the Jewish environment in which Christianity originated and the eros motif in the Hellenistic world within which Christianity spread and developed. The nomos motif provides fellowship with God on the basis of righteousness according to the law and the eros motif represents humanity's desire for God and its effort, starting with itself and impelled by its own resources, to satisfy its own need for a divine life. The history of Christianity unfolds for Nygren as a history of the struggle of the agape motif with these other fundamental motifs.

The nomos motif, for example, recurs early on in the writings of the Apostolic Fathers and Tertullian, and the eros motif reasserts itself in the Gnostic and Alexandrian theologies, while Marcion and Irenaeus exemplify the agape motif. Augustine attempted a synthesis of eros and agape in his Latin term caritas. Nygren regards Augustine's effort to be pivotal but unsuccessful. Augustine's caritas Dei, says Nygren, is not God's love for humanity, a love which the Holy Spirit has "shed abroad in human hearts" (Romans 5:5), but human love for God which is eros, acquisitive and self-serving love. The Augustinian synthesis prevailed in Christian history for a thousand years until it was challenged by the Renaissance to return to the unadulterated eros motif, and by Luther to return to the biblical agape motif which constitutes a divine-human fellowship in consequence of God's extravagant good will to humanity in Christ. In this way Nygren shows the agape motif to be the internal,

419 - Anthropology and Comparative Religion Turner

governing logic of Christianity in its varied, intellectual, and historical manifestations.

Nygren certainly tried to demarcate the kind of subject matter and the kinds of questions that theology addresses from the subject matter and questions proper to his empirical science of comparative religion. This is not so easy since the two disciplines have been closely intertwined from the early days of the latter. Originally, comparative religion was freely used to buttress theologians in their nineteenth century debate against the encroachments of science and Enlightenment on their authority. However, the concept of a primitive human society deeply imbued with religious feeling works either way, as the religious ideal from which secular moderns have lapsed, or as the flawed intellectual gropings over which scientific knowledge has triumphed.

One of our methodological questions is whether motif-research as Nygren conceived it, that is, as a method for theological inquiry, could ever be used for comparative religion.²² Why not? Nygren used it comparatively. But could the conceptual tool ever resist the weight of prior commitments to a particular view of the eternal? How can motif research be conducted without imposing meanings on its subject matter? We recall Gronbech's inveighing against the imported ideas of romantic mythologists.

Nygren never thought that motif-research would provide a method for settling the conflicting truth-claims of the several religions. Whether we accept one religion as true or not is a matter for personal judgment. We can properly make comparative evaluations of religious world-views but not of religions. Religions must be measured by the religious standard: experience of the eternal. When we do this, according to Nygren, "many forms of religion which stand lower down in the scale when regarded as world-views take a higher place when we regard them from the autonomous point of view of religion and the question of the eternal; for they can show a more vigorous experience of the eternal than many a 'higher' religion."²³

ANTHROPOLOGY AND COMPARATIVE
RELIGION

We have provided a much too brief survey of the thought of three Scandinavian historians of religion. It serves as a point of departure for reviewing some of the differences about method that concern workers in comparative religion and anthropology. One of this survey's advantages is to break the line of demarcation based on text and fieldwork, for on each of the main dividing issues these textualists are to be found allied with fieldworkers.

²² Its use in historical, literary and textual research has been successful. Note the contributions in Kegley.
²³ Anders Nygren, Dogmatikens vetenskapliga grundlaggning (Lund: C.W.K. Gleerup, 1922), p. 42.

420 - Anthropology and Comparative Religion Turner

We note the guiding conviction in Gronbech, Pedersen, and Nygren that a culture has to be treated as a whole, and as a logical structure. A method that obliterates the differences between distinct religious traditions was a wrong method in their eyes. Their own work sought to uncover the distinctive logical patterns of Islam,²⁴ Israel, the Norsemen, Classical Greece ²⁵ and Roman Christianity. They disapproved of the attempts of comparative religion in their day to sink the differences for the sake of disengaging large, universal patterns of thought.²⁶ This methodological strictness is increased by Nygren in his campaign to drive metaphysical speculations out of the discipline.

(1) Eliade and Recent Folkloric Tradition. We note that to some extent Nygren wrote in vain, since Mircea Eliade and his colleagues maintain the folkloric tradition. They work on a world scale, not transporting favored themes from one region to another, but seeking the eternal realities in the human consciousness. Their world view is just as sympathetic to our times as it was congenial to German folklorists to find classical Greek themes embodied in the Norse sagas. For it is completely congenial to the post-romantic culture of the West to believe in some distant time when human being, God, and nature formed an organic unity, to Which the myths of the world bear fragmentary witness and from which moment of unity the rest of the story has been a long, downhill disintegration.²⁷

Ninian Smart has remarked that in Eliade's hands myths serve as "the vehicle of a certain world view, and a means of giving life to much of man's archaic religious symbolism."²⁸ Admittedly, comparative religion would lose a lot of its attractiveness if it did not seem to present a theory of humanity and the universe under guise of analyzing other cultures' theories of the same. But the tempting opportunity for oblique metaphysics would vanish if mythologists adopted, the methodological counsels of Nygren.

Edmund Leach in a review of two books by Eliade charged him with inadequate and old fashioned scholarship.²⁹ Using mythology to launch a hidden metaphysics is as dubious an enterprise now as when Nygren wrote. Another way of not being with the times is falling into logical traps which used to be hidden and which have subsequently been, well sign-posted. For example, Leach convicts Eliade of confusing structure with content, a distinction which greatly interested Nygren. His idea of

²⁴ Johannes Pedersen, "Der Islam," in J. E. Lehmann, ed., Textbuch zur Religionsgeschichte (n.p., n.d.), pp. 135-170.
²⁵ W. Gronbech, Der Hellenism (Gottingen: n.p., 1953).
²⁶ This is the burden of Edmund Leach's reproach against James Frazer and Mircea
Eliade.
²⁷ Frank Kermode, The Romantic Image (London: Routledge and Kegan Paul, 1957).
²⁸ Ninian Smart, "Beyond Eliade," Numen VoI. 25, No. 2 (1978): 171-183.
²⁹ Edmund R. Leach, "Sermons by a Man on a Ladder," review of Mircea Eliade, Mephistopheles and the Androgyne (New York: Sheed and Ward) and The Two and the One (New York: Havrill Press) in The New York Review of Books, Vol. VII, No. 6, October 20, 1966.

421 - Anthropology and Comparative Religion Turner

motif-research depended upon distinguishing a religion's logical structure from its manifold content. Leach and Nygren come interestingly close together on this issue.

(2) Leach's View of Religious Discourse. Leach is interested in the focus of religious discourse upon mediating categories, beings that lie betwixt-and-between the everyday classifications of reality, anomalies which are treated as sacred or taboo.³⁰ He writes as if the content of religions is invented as a kind of practical remedy or cover-up of intolerable logical paradoxes or antinomies. Anomaly occupies in his scheme for identifying religious discourse somewhat the same place as the eternal in Nygren's scheme. Fair enough, for (though Nygren himself would deny it) religion is explicitly a metaphysical exercise and as such must deal with the fundamental relations of language and reason. ³¹ Paradox must therefore emerge at a very general level of structure. Moving down to particular religions and particular myths we are at a lower logical level at which culturally selected structures are used for organizing culturally selected contents. At that level what is going to count as a paradox or as a mediating category depends on the historical development of the particular tradition.

At the general level of structure Leach is right about antinomy and paradox at the foundations of religious thought. At the level of any particular religion or myth he is wrong if he thinks that the members of the culture will agree with him in identifying paradoxes. One person's anomaly is another's normal case. For example, at the general level there is anomaly in the Christian definition of a human as part spirit, part matter, but it is not one that leads Christians to regard other humans as sacred or taboo. Paradoxes are not necessarily obvious, nor do they necessarily constitute cognitive discomfort which seeks relief by creating sacred mediating elements.³² It takes a special kind of training to spot an antinomy and being a theologian is not enough.

One of the very greatest of Christian thinkers enters the logic books wearing a dunce's cap. "There is the ancient paradox of Epimenides the Cretan, who said that all Cretans were liars. If he spoke the truth, he was a liar. It seems that this paradox may have reached the ears of St. Paul and that he missed the point of it, He warned, in his epistle to Titus: 'one of themselves, even a prophet of their own, said the Cretans are always liars.'³³ There is really no problem at all about failing to see contradiction. Most people do it all the time. But if they do perceive a logical error, they have several recourses, not one. W.V.O. Quine says: "Actually the paradox of Epimenides is untidy; there are loopholes.

³⁰ Edmund R. Leach, Genesis as Myth and Other Essays (London: Cape, 1969).
³¹ See Bernard Williams, "Tertullian's Paradox," in New Essays in Philosophical Theology, eds. Anthony Flew and Alasdair McIntyre (New York: Macmillan, 1955).
³² Edmund R. Leach and Alan Aycock, Strueturalist Interpretations of Biblical Myth (Cambridge: Cambridge University Press, 1983).
³³ W.V.O. Quine, The Ways of Paradox and Other Essays, Revised and Enlarged Edition (Cambridge, Mass.: Harvard University Press, 1976), p. 47.

422 - Anthropology and Comparative Religion Turner

Perhaps some Cretans were liars, notably Epimenides and others were not; perhaps Epimenides was a liar who occasionally told the truth; either way, it turns out that the contradiction vanishes."³⁴

Leach often seems to make the same formal mistake as Nygren in attributing to logic an independent power to force recognition of fixed structures. As Nygren thought that the category of the eternal was self-evident once it had been pointed out, so Leach writes as if the structure of anomaly is self-evident across cultures. He is surely right to advocate the tracing of recognized anomalies within a cultural tradition as a key to its logical structure; but not if he tries to use this key in the absence of accumulated historical knowledge.

(3) Motif-Research: Anthropological Applications. We are led back to the. question of whether motif research can ever be applied across several cultural traditions. Nygren, using it to identify key concepts such as Agape, Eros, Nomos, seems to have been satisfied with static comparisons. He says this tradition is different from that one and that from the other; at that point be stops. It is the point at which fieldworkers feel safe in stopping too. At the limit, they can take in a whole cultural region and show how the same themes emerge and are combined-into very different transformations of the same underlying logical system.

There is nothing methodologically suspect, for example, in tracing how the idea of Divine Kingship is transformed in different West African states, how the Temple in the Balinese state is transformed through Indonesia, or ancestral control in West Africa. But such extensions of motif-research can never bring us beyond the regional and historic compartmentalization of which Jacob Neusner complains in his response to this essay.³⁵ Restriction to a historic region endorses a cultural relativism fatal to further conversation between anthropology and comparative religion and leads us straight back to the frustrations of our starting point.

We can make a different kind of return to Nygren's concern for the logical structures that maintain cultural unity. There are many questions we can ask following Leach's placement of logical contradiction as the distinctive focus of religious discourse. And there need to be scholarly restraints that will help us avoid the pitfalls. The major methodological problem comes from pretending to leave our cultural presuppositions behind when we examine those of another culture-an impossible feat and unnecessary. Rather than claim to be saying something that is valid cross-culturally, there is quite enough challenge in giving an account that is valid within our own cultural terms. It is

³⁴ Ibid.
³⁵ Neusner's most important criticism of anthropologists centered on the difficulty of dredging from a morass of detail anything that could count as a general prop08ition or hypothesis or guideline useful to comparative religion. [Editor's Note: Neusner's remarks may be found printed in the pages immediately following this article.]

423 - Anthropology and Comparative Religion Turner

quite a lot to demand that comparative religion meet the normal standards of precision and completeness that are applied by ourselves interpreting secular discourse.

With those modest and safer constraints, there is still an enormous program for comparative religion. If politicians say that they will abolish taxes, there are established criteria for deciding that they are not likely to do so, and why they say what they say. If preachers say that hell fire will burn their opponents and others say that God is all merciful, there are similar contextual interpretative rules that we can apply to understand why they say what they say. The part of Nygren's program which he enunciated clearly enough about contextualizing religious statements is the part which he exemplified least well in his own research on motifs.

Mythologists are tempted to rough-handle myths, partly because literary texts on their own offer little resistance. Institutions are also pliant to violent misinterpretation, but at least institutions have a spatio-temporal existence: they organize flows of human traffic, they muster and sustain collaboration, their physical effects can be seen in revenues actually collected or battles won. Comparative religion is better protected from sheer fantasy when it concentrates on relating religious ideas to succession to office, family inheritance, marriage practices, and protocols for war and peacemaking. The evidence can be scrutinized: a person cannot be in two places at once, nor can an army. Controversy can move to some agreement about technological levels, size of population, climate and resources, degrees of centralization, and devolution of authority. Historians have always paid attention to these matters.

If anthropology has a particular methodological idea, it is to look for systematic connections between changes in the social sphere and changes in the sphere of religion. We know well that the same doctrines can be put to very different interpretations by two interspersed populations. The origin of even slight shifts in the form of words or in the meanings assigned to the same words lies in the different commitment to institutions. Religion legitimizes a set of social norms. In doing so, it delegitimizes a rival set. In the conflicting dialogues about how social ideals will be actualized, we find the surest clues to the interpretation of religions.

Illustrating this method, we will review a program of religious studies that implicitly was influenced by Nygren; that is the corpus put in hand by two friends, Evans-Pritchard and Meyer Fortes, in the post war years and carried on by their students. Evans-Pritchard refers several times to Nygren and to his idea that religions could be compared by identifying their fundamental motifs.³⁶ The same conception of a dominant set of

³⁶ E.E. Evans- Pritchard, Theories of Primitive Religion (Oxford: Clarendon Press, 1965), p. 14; Evans- Pritchard, Essays in Social Anthropology (London: Faber and Faber, 1962), p. 54; Evans-Pritchard, Nuer Religion (Oxford: Clarendon Press, 1956), p. 317.

424 - Anthropology and Comparative Religion Turner

ideas and consistent patterning among the rest is a strong tradition in American cultural anthropology.

Unfortunately, the problem of identifying a central motif is exactly the same as the one that arises in the case of mythology: how to control the subjective input of the analyst, how to prevent the analytic method from imposing its categories and values-how to defend objectivity. The new methods of studying African religions leaned heavily on detailed record of the institutions of the worshippers. This is the only protection objectivity can ever have and the only guarantee of fidelity to a translated text. The exegesis of religious ideas depended on the actual religious discourse of the believers. The first originality in this approach was to take local African religions seriously as religions. The second was the sophistication of a critical ongoing dialogue among field workers.

At first, African religions were examined in the light of religious ideas in the Judeo-Graeco-Roman tradition. Later, they were examined in the light each study shed on the others. Inevitably, the terms of the research imparted a bias, created principles of selection, and provided a strongly colored terminology. To counteract this bias, the investigators expressly exposed their own theological backgrounds. In consequence, several African religions are now represented by a rich and sensitive literature. Because of the disciplined interchange, this record contributes as much to methodology as to the knowledge of African religions. Each volume that made a signal advance in method showed at the same time new flaws; the next study would seek to correct the balance, but tended to tip over in another direction.

Let us start with two books by Meyer Fortes on the Tallensi of the Upper Volta of Ghana.³⁷ They were not ostensibly about religion, but he had found it necessary in writing about social forms to analyze Tallensi theories of the universe, and especially the power attributed to. their ancestors. This was the first detailed study of the integration of an ancestor cult with the system of marriage and filiation. Later, Fortes elaborated upon more common religious themes, such as the explanation of evil, individual accountability to God, personal identity.

Observe that this material is not myth. It is a series of mutually implicated precepts about how to behave. It is a series of rules about burial, eating, speaking, about formal respect between a man and his wife, a man and his in-laws, a man and his first-born son, coherently anchored in social institutions. In these essays, Fortes wrote as one reared in the Jewish faith, explicitly comparing Tallensi and Old Testament religion. The weaknesses here, mainly selective bias, have to be offset in our appraisal by the enriched appreciation. What Fortes took to be the central motifs of Tallensi religion were remarkably like certain ones of Judaism: piety, respect for age, and kinship. From this base,

³⁷ Meyer Fortes, The Dynamics of Clanship Among the Tallensi (London: Oxford University Press, 1945); and Fortes, The Web of Kinship Among the Tallensi (London: Oxford University Press, 1949).

425 - Anthropology and Comparative Religion Turner

Robin Horton developed a very important comparison of conscience, and ideas of destiny in other West African cultures.³⁸

Evans-Pritchard's study of The Nuer Religion, 1956, shows the same bias, but he not only saw the Nuer sacrifice through the lens of the Old Testament, he also used the terms from Christian medieval philosophy to translate concepts for which modern, secular English has no equivalent. This has of course been criticized but the problem of translation goes deeper. In our view, Evans-Pritchard's use of the terms "God," "sin," "soul," is no worse than using "Virgin Birth" to translate a heterogenous collection of ideas. It is a possible source of confusion, but terms internal to our own religious history may be better than outsider terms, such as "deity" for God, "taboo" for sin, "fetish" for sacrament. Evans-Pritchard's more serious blunder was to suppose that Christian and Nuer religious discourse could be matched, concept for concept. Though these two studies of Tallensi and Nuer religion were probably the fullest and deepest source at that date on religions to which the foreign writers did not convert, they laid bare problems of objectivity which would hardly have been guessed before.

For the next illustration in this exploratory series, we would name Divinity and Experience, the Religion of the Dinka, by R.G. Lienhardt.³⁹ The Dinka are neighbors of the Nuer in the Nilotic Sudan. Both Nuer and Dinka religions distinguish themselves from Tallensi worship by the immediacy of the presence of God: they pay no cult to powerful mediating ancestors. Why this difference? What kinds of religions suppose their God to be intimately present among creatures and what kinds suppose God to be very distant and relatively ineffectual? Asked in the light of a comparison of institutional supports for one or another kind of cult, those questions can be answered quite precisely.

Lienhardt's work takes up the challenge of objectivity at two new points. An intention to be objective in judgment is not enough. Improving on the account of Nuer religion, he showed that objectivity has to be sought in extreme linguistic subtlety as well. It is too simple to say that Dinka and Nuer address their God directly. The word God should be justified. Fortes had usually preferred to evade the issue by sticking with the native terms. Evans-Pritchard had no trouble about using the word God. Lienhardt makes a careful study of grammatical and syntactic restrictions on the equivalent Dinka term and finds that he can use divinity with a capital D and divinity with a small d to correspond to the flexibilities of Dinka usage. With similar linguistic finesse, be translated complex reflections on the relation of spirit to matter. His main contribution to methodology in this book is a thorough scouting out of the concept of ritual. In the context of old assumptions about African

³⁸ Robin Horton, "Destiny and the Unconscious in West Africa," Africa (January 1961): 110-116.
³⁹ R.G. Lienhardt, Divinity and Experience: the Religion of the Dinka (Oxford: Clarendon Press, 1961).

426 - Anthropology and Comparative Religion Turner

magic, this was the first sophisticated analysis of cognitive processes fixed by publicly enacted forms.

It may be the very intensity of the internal debate among Africanists in the postwar decades that kept Polynesian, Melanesian and American Indian area specialists from sharing very much in it. In the same spirit comparative religionists may hesitate to appropriate the research for their purposes. Certainly, the complete bibliography would be formidably extensive, but it would be a pity to allow the sheer quality and quantity of the work to prevent assimilation into the program of comparative religion.

We will be content to indicate two general directions followed by subsequent work. The first is in semantic analysis. Of all the attempts to find a central knot of ideas around which ritual symbols are articulated, Victor Turner's exegesis of Ndembu philosophy most brilliantly fulfills Nygren's original program.⁴⁰ The second effort has been to document the way that religious discourse mobilizes political support. The first three studies gave a static view of the relation between religion and society. Already, the charting of dynamic interaction had begun in that part of Africanist research that focused upon witchcraft ideas.⁴¹ The path-breaker in religious studies was John Middleton who, in his Lugbara Religion,⁴² showed the phases in an ancestor cult were part of the phases in the growth and decline of the lineage whose members constituted a congregation. The delicate confrontation of religious leaders and their followers henceforth became the question that had to be examined at a new level of insight. From these beginnings, an important record of the interaction of African cultures with Islam and with Christianity can only be noted here. We have merely tried to convey an impression of rich resources for comparative religion which the Africanist anthropologists have prepared.

We can sum up by returning to Nygren's idea of the motif. It becomes clear that the apparent logical coherence of a set of religious ideas in a particular civilization is not due to the simple application of logical rules of noncontradiction. They seem logically coherent, first because they rest upon consistent institutional forms and second because a set of

⁴⁰ Victor Turner, Schism and Continuity in an African Society (Manchester: Manchester University Press for the Rhodes- Livingstone Institute, 1957); Turner, The Forest of Symbols: Aspects of Ndembu Ritual (Ithaca: Cornell University Press, 1967); Turner, Drums of Affliction (Oxford: Clarendon Press, 1968); and Turner, The Ritual Process: Structure and Anti-Structure (Chicago: Aldine Publishing Co., 1969).
⁴¹ The path-breaking monograph was by Clyde Mitchell, The Yao Village (Manchester: Manchester University Press, 1956), but it was inspired by a book that was published much later by I. Schapera, Rain-making Rites of the Tswana Tribes, African Social Research Documents (Cambridge: African Studies Center, 1971). Mitchell related alternating cycles of belief in ancestors and witches to the cyclic renewal of the kinship structure.
^{42 J}ohn Middleton, Lugbara Religion: Ritual and Authority Among East African People (London: Oxford University Press, 1960). It should be mentioned that the first dynamic analysis of cycles of religious changes to cyclic changes in the political regime is Leach's now classic, Political Systems of Highland Burma (London: Athlone Press, 1970).

427 - Anthropology and Comparative Religion Turner

analogies has been constructed upon them. These analogies create coherence by extending words and logical operations from one context to another. Overlapping and repeating, such analogies impose a complex ordering on experience where confusion would otherwise reign.

Some of the processes have been traced by which new ideas have to be reformulated in familiar terms before they can be accepted. For example, Evans-Pritchard thought that the Nuer rejected Dinka "magic" because they found such ideas incompatible with their basically "manist" conception of the universe as dominated by spirit-an explanation in terms of motif-coherence. Jean Buxton, encountering among the neighboring Mandari similar opposition to foreign Dinka magic, suggests that political opposition to Dinka culture was more important than logical incompatibility with the central philosophical Motif.⁴³ A recognizable traditional conceptual scheme and awareness of historical difference are mutually reinforcing.

Investigating carefully the shared construction of grand public concepts and their institutional supports brings us closer to understanding religious stability-the first necessary step to understanding religious change. We need to turn our eyes away from the exotic and the bizarre. As Durkheim said: "Religious conceptions have as their object, before everything else, to express and explain, not that which is exceptional and abnormal in things, but, on the contrary, that which is constant and regular. Very frequently, the gods serve less to account for the monstrosities, fantasies and anomalies than for the regular march of the universe, for the movement of the stars, the rhythm of the seasons, the annual growth of vegetation, the perpetuation of the species."⁴⁴

⁴³ Jean Buxton, Religion and Healing Among the Mandari (Oxford: Clarendon Press, 1973).
⁴⁴ Emil Durkheim, The Elementary Forms of the Religious Life (London: Allen and Unwin, 1915).