The History of Religions: Essays in Methodology by Mircea Eliade and Joseph M. Kitagawa (eds.)
Mircea Eliade was born in Bucharest in 1907 and began teaching in the field the history of religions in 1946 at the Sorbonne in Paris. He was a member of the University of Chicago faculty from 1957 until his death in 1986. His many books include: Cosmos and History (1959), The Sacred and the Profane (1959), Myths, Dreams and Mysteries (1960), Images and Symbols (1969), and Myths and Reality (1963). Published by University of Chicago Press, 1959. The material was prepared for Religion Online by Ted & Winnie Brock.
Methodological Remarks on
the Study of Religion’s Symbolism by Mircea Eliade
As has frequently been noted, there has been for some time now a vogue for symbolism. (Cf. Mirdea Eliade, Images et symboles [Paris, 1952] pp. 9ff.) Several factors have contributed to the study of symbolism to give it the privileged place that it holds today. In the first place there have been the discoveries of depth psychology, especially the fact that the activity of the unconscious can be grasped through the interpretation of images, figures, and scenarios. These are not to be taken at their face value, but function as "ciphers" for situations and types which the consciousness does not want, or is not able, to recognize. (A clear exposition of the theories of Freud and Jung on the symbol is to be found in Yolande Jacobi, Komplex, Archetypus, Symbol in der Psychologie C. C. Jungs [Zurich, 1957], pp. 86 ff).
Secondly, the turn of the century witnessed the rise of abstract art and, after World War I, the poetic experiments of the surrealists, both of which served to familiarize the educated public with the non-figurative and dream worlds. But these worlds could reveal their meaning only insofar as one succeeded in deciphering their structures, which were "symbolic." A third factor served to arouse interest in the study of symbolism. This was the research of ethnologists in primitive societies and, above all, the hypotheses of Lucien Lévy-Bruhl concerning the structure and functioning of the "primitive mentality." Lévy-Bruhl considered the "primitive mentality" to be prelogical, since it would seem to he ruled by what he called "mystic participation." Before his death, however, he abandoned the hypothesis of a prelogical primitive mentality radically different from, and in opposition to, the modern mentality. (Cf. Lucien Lévy-Bruhl, Les Carnets, ed. Maurice Leenhardt [Paris, 1946]). In fact, his hypothesis had not encountered much support among ethnologists and sociologists, but it had been useful as a springboard for discussions among philosophers, sociologists, and psychologists. Furthermore, it drew the attention of the intellectual elite to the behavior of primitive man, to his psycho-mental life and his cultural creations. The present interest of philosophers in myth and symbol, especially in Europe, is due in large part to the works of Lévy-Bruhl and to the controversies that they provoked.
Finally, a considerable role in the vogue for symbolism must be attributed to the researches of certain philosophers, epistemologists, and linguists who wanted to show the symbolic character not only of language, but also of all other activities of the human spirit, from rite and myth to art and science. (Cf. Max Schlesinger, Geschichte des Symbols (Berlin, 1912); A. N. Whitehead, Symbolism: Its Meaning and Effect (New York, 1927); W. M. Urban, Language and Reality: The Philosophy of Language and the Principles of Symbolism (London and New York, 1939); F. Ernest Johnson (ed.), Religious Symbolism (New York, 1955); Symbols and Values: An Initial Study (Conference on Science, Philosophy, and Religion, No. 13 [New York, 1955]). Since man has a "symbol-forming power," all that he produces is symbolic. (It is sufficient to recall the works of Ernst Cassirer, Philosophie der symbolischen Formen (3 vols.; Berlin, 1923-29) and his Essay on Man (New Haven: Yale University Press, 1944), and Susanne K. Langer, Philosophy in a New Key: A Study in the Symbolism of Reason, Rite and Art (Cambridge: Harvard University Press, 1942).
In recalling the most recent factors that have served to generalize and popularize the interest in symbolism, we have at the same time enumerated the perspectives in which the study of symbols has been approached. These are the perspectives of depth psychology, of the plastic and poetic arts, of ethnology, of semantics, of epistemology, and of philosophy. The historian of religions can only be grateful for these researches undertaken from different points of view on a subject so important to his own field. Since the sciences of man are interdependent, each important discovery has repercussions in neighboring disciplines. What psychology or semantics teaches us concerning the function of symbols is definitely of importance for the science of religions. Fundamentally, the subject is the same: we are always dealing with the understanding of man and of his situation in the world. A fruitful study might even be undertaken on the relationships between the disciplines mentioned above and the science of religions.
This is not to say that the field of the science of religions coincides with the fields of the other disciplines. Moreover, the very procedure of the historian of religions is not identical with that of the psychologist, the linguist, or the sociologist. It is just as dissimilar to that of the theologian. The historian of religions is preoccupied uniquely with religious symbols, that is, with those that are bound up with a religious experience or a religious conception of the world.
The procedure of the historian of religions is just as different from that of the theologian. All theology implies a systematic reflection on the content of religious experience, aiming at a deeper and clearer understanding of the relationships between God-Creator and man-creature. But the historian of religions uses an empirical method of approach. He is concerned with religio-historical facts which he seeks to understand and to make intelligible to others. He is attracted to both the meaning of a religious phenomenon and to its history; he tries to do justice to both and not to sacrifice either one of them. Of course, the historian of religions also is led to systematize the results of his findings and to reflect on the structure of the religious phenomena. But then he completes his historical work as phenomenologist or philosopher of religion. In the broad sense of the term, the science of religions embraces the phenomenology as well as the philosophy of religion. But the historian of religions sensu stricto can never ignore that which is historically concrete. He applies himself to deciphering in the temporally and historically concrete the destined course of experiences that arise from an irresistible human desire to transcend time and history. All authentic religious experience implies a desperate effort to disclose the foundation of things, the ultimate reality. But all expression or conceptual formulation of such religious experience is imbedded in a historical context. Consequently, these expressions and formulations become "historical documents," comparable to all other cultural data, such as artistic creations, social and economic phenomena, and so forth. The greatest claim to merit of the history of religions is precisely its effort to decipher in a "fact," conditioned as it is by the historical moment and the cultural style of the epoch, the existential situation that made it possible.
It is equally necessary to take account of the fact that theology is preoccupied essentially with historical and revealed religions, that is, with Jewish, Christian, and Muslim monotheisms, and only secondarily with the religions of the ancient Near Eastern and Mediterranean worlds. A theological study of religious symbolism will necessarily take into account selected documents from the great monotheistic religions rather than so-called "primitive" materials. (Obviously, a theology of the history of religions will be obliged to take into consideration all these archaic and primitive religious experiences. But this theology presupposes the existence of the history of religions and depends on its results.) But the historian of religions aims to familiarize himself with the greatest possible number of religions, especially with archaic and primitive religions, where he has a chance to encounter certain religious institutions still in their elementary stages.
In brief, while the research on symbols in general and religious symbolism in particular by specialists in other disciplines deserves his consideration, the historian of religions is obliged in the final analysis, to approach the subject with his own means of investigation and in his proper perspective. There is no other perspective in which religio-historical data can better be integrated than that of the general science of religions. It is solely through timidity that historians of religions have at times accepted an integration proposed by sociologists or by anthropologists. Insofar as one can formulate general considerations on the religious behavior of man, this task rightly belongs to the historian of religions, provided, of course, that he master and integrate the results of the researches made in all the important areas of his discipline.
Unfortunately, this happens less and less frequently. (Eliade, op. cit., pp. 33 ff). There are few historians of religions who make an effort to follow the research undertaken in the domains which lie outside of their "specialty." If a historian of Greek religion at times take an interest in recent studies on Iranian or Indian religions, he is less inclined to follow the work of his specialist-colleagues, let us say, in Altaic, Bantu, or Indonesian religions. When he wishes to offer a comparison, or propose a more general explanation of phenomena of Greek or Mediterranean religions, he consults a "Manual," or pages through Frazer, or resorts to a current theory on the religion of the "primitives." In other words, he foils the very work which he is expected to do as a historian of religions: to keep himself informed about the research of his colleagues, specialists in other areas, assimilating and confronting their findings, and finally integrating them in order to better understand his Greek documents.
This hesitancy can be explained, it would seem, by two preconceived ideas. The first might be formulated in this manner: the history of religions constitutes a limitless domain which nobody can master; (This is true of all the historical disciplines. More than fifty years ago Anatole France remarked that it would take several lifetimes to read all the documents concerned uniquely with the French Revolution.) hence it is preferable to know one area well instead of wandering like a dilettante through many. The second preconception, rather more implicit than overtly recognized, is that for "general theories" about religion it is more prudent to consult a sociologist, an anthropolgoist, a psychologist, a philosopher, or a theologian. Much could be said about the inhibition of the historian of religions who faces a work of comparison and integration. For the moment, it is important to rectify the erroneous opinion that exists concerning the task of integration.
It is not a question, for the historian of religions, of substituting himself for the various specialists, that is to say, of mastering their respective philologies. Such a substitution is not only impossible; it is useless. The historian of religions whose field of investigation is, for instance, Vedic India or Classical Greece, is not required to master Chinese, Indonesian, or Bantu in order to gain access to the Taoist religious documents, the myths of the aborigines of Ceram, or the rites of Tonga for use in his research. His task is rather to inform himself of the progress made by the specialists in each of these areas. One is a historian of religions not by virtue of mastering a certain number of philologies, but because one is able to integrate religious data into a general perspective. The historian of religions does not act as a philologist, but as a hermeneutist. The mastery of his own specialty has amply taught him how to orient himself in the labyrinth of facts, where to go for the most important sources, the most appropriate translations, and such studies as are likely to guide his research. He endeavors to understand the materials that philologists and historians make available to him in his own perspective, that of the history of religions. It takes the linguist several weeks of labor to unravel the structure of a language with which he is not familiar. The historian of religions should be capable of arriving at similar results when working with religious data which are foreign to his own field of study. He is not held to duplicate the efforts of specialists, just as a historian of the nineteenth-century French novel is not expected to duplicate the labors on the manuscripts of Balzac or Flaubert, the stylistic analyses of Stendhal, or the research on the sources of Victor Hugo or Gérard de Nerval. His duty is rather to know about all these labors, to use their results, and to integrate them.
In like manner one can compare the method of the historian of religions with that of a biologist. When the latter studies, for instance, the behavior of a certain species of insect, he does not take the place of the entomologist. He expands, confronts, and integrates the investigations of the entomologist. To be certain, the biologist, too, is a specialist in one of the branches of zoology and benefits from long experience with such and such a zoological species. But his procedure is different from that of the zoologist; he is preoccupied with the general structures of animal life, and not alone with the "history" of a particular species.
The second preconceived idea held by certain historians of religions, notably that it is necessary to consult another "specialist" for the total and systematic interpretation of religious facts, is probably explained by the philosophical timidity of many scholars. Two factors above all have contributed to implant and to foster this hesitancy: on the one hand, the very structure of the discipline which serves as a sort of introduction or preparation, to the science of religions (one knows that the majority of historians of religions are former philologists, archeologists, historians, orientalists, or ethnologists); on the other hand, the inhibition created by the lamentable failure of the vast theoretical improvisations of the end of the nineteenth century and the beginning of the twentieth (mythology considered a "disease of language," astral and naturist mythologies, pan-Babylonism, animism and pre-animism, etc.). Be that as it may, the historian of religions feels more secure if he leaves to other disciplines --sociology, psychology, anthropology -- the risk of syntheses or of general theories. (Indeed, all "general theories" which have dominated the history of religions from its beginnings have been the work of linguists, anthropologists, sociologists, ethnologists, and philosophers [cf. Eliade, op. cit,, pp. 35 ff.]). But this amounts to saying that the historian of religions hesitates to complete his preparatory work as a philologist and as a historian through an effort of understanding, which, to be sure, presupposes an act of thinking.
It is not the intention of the writer to develop these observations touching on the field and methods of the science of religions. The purpose of this article is more modest. We want to show how we can envisage the study of religious symbolism in the perspective of the science of religions, and what the results of this procedure can be. In discussing this specific example, however, we shall confront methodological difficulties inherent in all research in the history of religions. To say this in another way, we shall have to discuss certain aspects of the method not in abstracto, but in such a way that these aspects may be grasped during the very process of research.
The first difficulty that faces the historian of religions is the enormous mass of documents, in our case, the considerable number of religious symbols. A problem is posed from the beginning: even supposing that one succeeds in mastering this mass of documents (which is usually not the case), has one the right to use them indiscriminately, that is, to group them, compare them, or even to manipulate them according to one’s own convenience? These religious documents are at the same time historical documents; they are an integral part of different cultural contexts. To summarize, each document has a particular meaning, part and parcel of the culture and the particular time from which it has been detached.
The difficulty is real, and we shall endeavor later on to show how one can surmount it. For the moment suffice it to say that the historian of religions is destined to encounter a similar difficulty in all his work. For, on the one hand, he wants to know all historical situations of religious behavior, and on the other hand he is obliged to abstract the structure of this behavior, such that it can be recognized in a multitude of situations. To give but one example, there exist innumerable variants of the symbolism of the Cosmic Tree. A certain number of these variants can be considered as coming from only a few centers of diffusion. One can even admit the possibility that all the variants of the Cosmic Tree come in the last analysis from one single center of diffusion. In this case, we might be permitted to hope that one day the history of the symbolism of the Cosmic Tree may be reconstructed, by pinning down the center of origin, the paths of diffusion, and the different values with which this symbol has been endowed during its migrations. Were such a historical monograph possible, it would render a great service to the science of religions. But the problem of the symbolism of the Cosmic Tree as such would not thereby be resolved. Quite another problem remains to be dealt with. What is the meaning of this symbol? What does it reveal, what does it show as a religious symbol? Each type or variety of this symbol reveals with a particular intensity or clarity certain aspects of the symbolism of the Cosmic Tree, leaving other aspects unemphasized. There are examples where the Cosmic Tree reveals itself chiefly as the imago mundi, and in other examples it presents itself as the axis mundi, as a pole that supports the Sky, binds together the three cosmic zones (Heaven, Earth, and Hell), and at the same time makes communication possible between Earth and Heaven. Still other variants emphasize the function of the periodic regeneration of the universe, or the role of the Cosmic Tree as the Center of the World or its creative potentialities, etc.
We have studied the symbolism of the Cosmic Tree in several of our previous works, (Cf. Eliade, Patterns in Comparative Religion (New York, 1958), pp. 270 ff.; Eliade, Le Chamanisme et les techniques archaïques de l’extase (Paris, 1951), pp. 244 ff.; and Images et Symboles, pp. 5S ff. and 213 ff.) and need not restate the problem here in its entirety. Suffice it to say that it is impossible to understand the meaning of the Cosmic Tree by considering only one or some of its variants. It is only by the analysis of a considerable number of examples that the structure of a symbol can be completely deciphered. Moreover, one can understand the meaning of a certain type of Cosmic Tree only after having studied the most important types and varieties. Only after an elucidation of the specific meanings of the Cosmic Tree in Mesopotamia or in ancient India can one understand the symbolism of Yggdrasil or the Cosmic Trees of Central Asia and of Siberia. In the science of religions, as elsewhere, comparisons are made in order to find both parallels and distinctions.
But there is still more. Only after taking account of all the variants do the differences of their meanings fall into relief. It is because the symbol of the Indonesian Cosmic Tree does not coincide with that of the Altaic Cosmic Tree that the first reveals all its importance for the science of religion. Thus the question is posed: Is there, in either instance, some innovation, obscuration of meaning, or a loss of the original meaning? Since we know what the Cosmic Tree means in Mesopotamia, in India, or in Siberia, the question arises: Because of what religio-historical circumstances, or by what interior reason, does the same symbol in Indonesia reveal a different meaning? Diffusion as such does not solve the problem. For even if one could demonstrate that the symbol had been diffused from a single center, one could still not give the reason why certain cultures have retained certain primary meanings, whereas others have forgotten, rejected, modified, or enriched them. One can come to understand the process of enrichment only by disengaging the structure of the symbol. It is because the Cosmic Tree symbolizes the mystery of a world in perpetual regeneration that it can symbolize, at the same time or successively, the pillar of the world and the cradle of the human race, the cosmic renovatio and the lunar rhythms, the Center of the World and the path by which one can pass from Earth to Heaven, etc. Each one of these new valorizations is possible because from the beginning the symbol of the Cosmic Tree reveals itself as a "cipher" of the world grasped as a living reality, sacred and inexhaustible. The historian of religions will have to elucidate the reasons why such a culture has retained, developed, or forgotten a symbolic aspect of the Cosmic Tree. In so doing, he will be led to penetrate more deeply into the soul of this culture, and will learn to differentiate it from others.
From a certain point of view, one could compare the situation of the historian of religions to that of the depth psychologist. One, like the other, is obliged not to lose touch with the given facts; they follow empirical methods; their goal is to understand "situations"-- personal situations in the case of the psychologist, historical situations in the case of the historian of religions. But the psychologist knows that he will not arrive at the understanding of an individual situation, and consequently cannot help his patient recover, except insofar as he can succeed in disclosing a structure behind the particular set of symptoms, that is, to the extent where he will recognize the main outlines of the history of the psyche in the peculiarities of an individual history. On the other hand, the psychologist improves his means of research and rectifies his theoretical conclusions by taking into consideration the discoveries made during the process of analysis. As we have just seen, the historian of religions proceeds no differently when, for example, he studies the symbolism of the World Tree. Whether he is led to limit himself, let us say, to Central Asia or Indonesia, or on the contrary proposes to approach this symbolism in its totality, he can accomplish his task only by taking into consideration all the important variants of the Cosmic Tree.
Since man is a homo symbolicus, and all his activities involve symbolism, it follows that all religious facts have a symbolic character. This is certainly true if we realize that every religious act and every cult object aims at a meta-empirical reality. When a tree becomes a cult object, it is not as a tree that it is venerated, but as a hierophany, that is, a manifestation of the sacred. (On hierophanies see Eliade, Patterns in Comparative Religion, pp. 3 ff. and passim.) And every religious act, by the simple fact that it is religious, is endowed with a meaning which, in the last instance, is "symbolic," since it refers to supernatural values or beings.
One could say, then, that all research undertaken on a religious subject implies the study of religious symbolism. However, in the current terminology of the science of religions, the term "symbol" is commonly reserved for religious facts whose symbolism is manifest and explicit. One speaks, for example, of the wheel as a solar symbol, of the cosmogonic egg as the symbol of the non-differentiated totality, or of the serpent as a chthonian, sexual, or funeral symbol, etc. (In like manner it is agreed that the term "symbolism" should be reserved for a structurally coherent ensemble, for example, we speak of aquatic symbolism, the structure of which cannot be deciphered except through studying a great number of religious facts which are heterogeneous in appearance, such as baptismal and lustration rites, aquatic cosmogonies, myths relative to floods or to marine catastrophes, myths featuring fecundity through contact with water, etc. [Cf. Eliade, Patterns in Comparative Religion, pp. 188 ff., and Images et symboles, pp. 164 ff. and 199 ff.]. Obviously, the current usage of the terms "symbol" and "symbolism" lack precision, but one must accommodate himself to this state of affairs. In many cases the context suffices to clarify the meaning.)
It is equally common to approach a certain religious institution --initiation, for instance -- or a religious act such as the orientatio, from the point of view of symbolism. The aim of such studies is to disregard the socio-religious contexts of the respective institutions of behavior in order to concentrate on the symbolism that they imply. Initiation is a complex phenomenon, comprising multiple rites, divergent mythologies, different social contexts, and heterogeneous ends. (Cf. Eliade, Birth and Rebirth: The Religious Meanings of Initiation in Human Culture [New York, 1958]). In the final analysis, to be sure, all of this has to do with "symbols." But the study of initiation symbolism seeks another end: to decipher the implicit symbolism in such and such a rite or initiation myth (regressus ad uterum, ritual death and resurrection, etc.), to study each one of these symbols morphologically and historically, and to elucidate the existential situation which has made the formation of these symbols possible.
The same is true of religious behavior, such as in the orientatio. There are innumerable rites of orientation and myths that justify them, all of which are ultimately derived from the experience of sacred space. To approach this problem in its totality presupposes the study of the orientation ritual, of geomancy, of the foundation rites of villages, temples, and houses, and of the symbolism of tents, huts, houses, etc. But since at the root of all this lies the experience of sacred space and a cosmological conception, we can conceive the study of the orientatio solely as a study of the symbolism of sacred space. This does not mean that one should overlook or ignore the historical and social contexts of all forms of the orientatio that one has taken the trouble to examine.
We might give many other examples of this kind of research on a particular symbolism: the "magic flight" and ascension, the night and the symbolism of darkness, solar, lunar, telluric, vegetation, and animal symbolism, the symbolism of the quest for immortality, the symbolism of the hero, and so forth. For each of these examples the procedure is essentially the same. One should strive to grasp the symbolic meaning of the religious facts in their heterogeneous, yet structurally interlocking appearances; such facts can be rites or ritual behavior as well as myths or legends or supernatural beings. Such a procedure does not imply the reduction of all meaning to a common denominator. One cannot insist strongly enough that the search for symbolic structures is not a work of reduction but of integration. We compare or contrast two expressions of a symbol not in order to reduce them to a single, pre-existent expression, but in order to discover the process whereby a structure is likely to assume enriched meanings. In studying the symbolism of flight and ascension we have given several examples of this process of enrichment; the reader who wants to verify the results obtained by such a methodological procedure is referred to that study. (Cf. Eliade, Mythes, réves et mystères (Paris, 1957), pp. 133-64 [symbolisms of ascension and "waking dreams"]).
The task of the historian of religions is not fulfilled unless he succeeds in discerning the function of religious symbolism in general. We know what the theologian, the philosopher, and the psychologist have to say on this topic. (We recall the analysis of Paul Tillich: "This is the great function of symbols: to point beyond themselves, in the power of that to which they point, to open up levels of reality which otherwise are closed, and to open up levels of the human mind of which we otherwise are not aware" (Paul Tillich, "Theology and Symbolism," in Religious Symbolism, ed. F. Ernest Johnson [New York, 1955], pp. 107-16). Let us now examine the conclusions of the historian of religions as he reflects on his own documents.
The first remark that he is led to make is that the World "speaks" or "reveals itself’’ through symbols, not, however, in a utilitarian and objective language. The symbol is not a mere reflection of objective reality. It reveals something more profound and more basic. Let us try to enumerate the different aspects of depths of this revelation.
1. Religious symbols are capable of revealing a modality of the real or a structure of the World that is not evident on the level of immediate experience. In order to illustrate in what sense the symbol aims at a modality of the real which is inaccessible to human experience, let us recall a single example, the symbolism of water, which is capable of expressing the pre-formal, the virtual, and the chaotic. This is clearly not a matter of rational knowledge; rather does the living consciousness grasp reality through the symbol, anterior to reflection. It is by such graspings that the World is constituted. Later, as these meanings expand, they will lead to the first reflections on the ultimate foundation of the World, i.e., to all cosmologies and ontologies, from the time of the Vedas to the pre-Socratics.
As to the capacity of symbols to reveal a profound structure of the World, we may recall what has been said above pertaining to the principal meanings of the Cosmic Tree. This symbol reveals the World as a living totality, periodically regenerating itself and, because of this regeneration, continually fruitful, rich, and inexhaustible. In this case also, it is not a question of a reflective knowledge, but of an immediate intuition of a "cipher" of the World. The World "speaks" through the symbol of the Cosmic Tree, and this "word" is understood directly. The World is apprehended as life, and in primitive thought, life is an aspect of being.
The religious symbols which point to the structures of life reveal a more profound, more mysterious life than that which is known through everyday experience. They unveil the miraculous, inexplicable side of life, and at the same time the sacramental dimensions of human existence. "Deciphered" in the light of religious symbols, human life itself reveals a hidden side; it comes from "another part," from far off; it is "divine" in the sense that it is the work of the gods or of supernatural beings.
2. This leads us to a second general remark: for the primitive, symbols are always religious because they point to something real or to a structure of the world. For on the archaic levels of culture, the real -- that is, the powerful, the meaningful, the living -- is equivalent to the sacred. On the other hand, the World is a creation of gods or supernatural beings; to unfold a structure of the World is equivalent to revealing a secret or a "ciphered" signification of divine workmanship. For this reason archaic religious symbols imply on ontology. It is, of course, a question of a presystematic ontology, the expression of a judgment about the world and simultaneously about human existence, a judgment that is not formulated in concepts and which rarely lends itself to conceptualization.
3. An essential characteristic of religious symbolism is its multivalence, its capacity to express simultaneously a number of meanings whose continuity is not evident on the plane of immediate experience. The symbolism of the moon, for example, reveals a connatural solidarity between the lunar rhythms, temporal becoming, water, the growth of plants, the female principle, death and resurrection, human destiny, weaving, and so forth. (Cf. Eliade, Patterns in Comparative Religion, pp. 154 ff.) In the final analysis, the symbolism of the moon reveals a correspondence of mystical order between the various levels of cosmic reality and certain modalities of human existence. Let us note that this correspondence becomes evident neither spontaneously in immediate experience nor through critical reflection. It is the result of a certain mode of "being present" in the world.
Even admitting that certain functions of the moon have been discovered through attentive observation of the lunar phases (for example, the relationship with rain and menstruation), it is hard to conceive that lunar symbolism in its totality has been constituted by a rational process. It is quite another order of knowledge that discloses, for example, the "lunar destiny" of human existence, the fact that man is "measured" by the temporal rhythms illustrated by the phases of the moon, that he is fated to die but, quite like the moon which reappears after three days of darkness, man too can begin his existence anew, that in any case he nourishes the hope in a life after death, assured or ameliorated through an initiation ritual.
4. This capacity of religious symbolism to reveal a multitude of structurally coherent meanings has an important consequence. The symbol is thus able to reveal a perspective in which heterogeneous realities are susceptible of articulation into a whole, or even of integration into a "system." In other words, the religious symbol allows man to discover a certain unity of the World and, at the same time, to disclose to himself his proper destiny as an integrating part of the World. Let us keep in mind this example of lunar symbolism. We then understand the sense in which the different meanings of lunar symbols form a sort of "system." On different levels (cosmological, anthropological, "spiritual"), the lunar rhythm reveals structures that can be homologized, that is to say, modalities of existence subject to the laws of time and of cyclic becoming, existences destined to a "life" which carries in its very structure death and rebirth. Owing to the symbolism of the moon, the World no longer appears as an arbitrary assemblage of heterogeneous and divergent realities. The diverse cosmic levels communicate with each other; they are "bound together" by the same lunar rhythm, just as human life also is "woven together" by the moon and is predestined by the "spinning" goddesses.
Another example will illustrate even better this capacity of symbols to open up a perspective through which things can be grasped and articulated into a system. The symbolism of night and darkness -- which can be discerned in the cosmogonic myths, in initiation rites, in iconographies portraying nocturnal or subterranean animals -- reveals the structural solidarity between precosmic and prenatal darkness on the one hand, and death, rebirth, and initiation on the other. (It is equally necessary to add that darkness symbolizes precosmic "chaos" as well as orgy [social confusion] and "folly" [dissolution of the personality]).
This makes possible not only the intuition of a certain mode of being, but also the understanding of the "place" of this mode of being in the constitution of the World and of the human condition. The symbolism of the Cosmic Night allows man to imagine what preceded him and preceded the World, to understand how things came into existence, and where things "existed" before they came to be. It is not an act of speculation, but a direct grasp of this mystery -- that things had a beginning, and that all that preceded and concerns this beginning has a considerable weight for human existence. One has only to recall the great importance attached to the initiation rites implying a regressus ad uteram, through which man believes he can begin a new existence, or the numerous ceremonials designed to reactualize periodically the primordial "chaos" in order to regenerate the world and human society.
5. Perhaps the most important function of religious symbolism--important above all because of the role which it will play in later philosophical speculations -- is its capacity for expressing paradoxical situations, or certain structures of ultimate reality, otherwise quite inexpressible. One example will suffice: the symbolism of the Symplegades, (Cf. Ananda K. Coomaraswamy, "Symplegades," in Homage to George Sarton, ed. M. F. Ashley Montagu (New York, 1947), pp. 463-88; cf. also Carl Hentze, Tod, Auferstehung, Weltordung (Zurich, 1955), esp. pp. 45 ff). as it may be deciphered in numerous myths, legends, and images presenting the paradoxical passage from one mode of being to another, such as transfer from this world to another world, from the Earth to Heaven or Hell, or passage from a profane mode of existence to a spiritual existence. The most frequent images are: passing between two rocks or two icebergs that bump together continuously, between two mountains in continual motion, between the jaws of a monster, or penetrating and withdrawing unhurt from a vagina dentata, or entering a mountain that has no opening. We understand what all these images point to; if there exists the possibility of a "passage," this cannot be realized except "in spirit," giving this term all the meanings that it has in archaic societies, i.e., referring to a disincarnated mode of being as well as the imaginary world and the world of ideas. One can pass through a Symplegade insofar as one is able to act "spiritually," insofar as one proves that one possesses imagination and intelligence and, consequently, is capable of detaching oneself from immediate reality. (Eliade, Birth and Rebirth, chap. iv.) No other symbol of the "difficult passage"-- not even the celebrated motif of the thin bridge like tile blade of a sword or the edge of the razor to which allusion is made in the Katha Upanishad (iii, 14) -- reveals better than the Symplegades that there is a mode of being inaccessible to immediate experience, and that one cannot attain to this mode of being except through renouncing the naïve belief in the inexpungeability of matter.
One could make similar remarks about the capacity of symbols to express the contradictory aspects of ultimate reality. Cusanus considered the coincidentia oppositorum as the most appropriate definition of the nature of God. But for many centuries this symbol had already been used to signify that which we call the "totality" or the "absolute" as well as the paradoxical coexistence in the divinity of polar and antagonistic principles. The conjunction of the Serpent (or of another symbol of the chthonian darkness and of the non-manifest) and of the Eagle (symbol of solar light and of the manifest) express, in iconography or in myths, the mystery of the totality or of cosmic unity. (Eliade, Patterns on Comparative Religion, pp. 419 ff.) Although the concepts of polarity and of the coincidentia oppositorum have been used in a systematic fashion since the beginnings of philosophical speculation, the symbols that dimly revealed them were not the result of critical reflection, but of an existential tension. In accepting his presence in the world, precisely as man found himself before the "cipher" or "word" of the world, he came to encounter the mystery of the contradictory aspects of a reality or of a "sacrality" that he was led to consider compact and homogeneous. One of the most important discoveries of the human spirit was naïvely anticipated when, through certain religious symbols, man guessed that the polarities and the antimonies could be articulated as a unity. Since then, the negative and sinister aspects of the cosmos and of the gods have not only found a justification, but have revealed themselves as an integral part of all reality or sacrality.
6. Finally, it is necessary to underline the existential value of religious symbolism, that is, the fact that a symbol always aims at a reality or a situation in which human existence is engaged. It is above all this existential dimension that marks off and distinguishes symbols from concepts. Symbols still keep their contact with the profound sources of life; they express, one might say, the "spiritual as lived" (le spirituel vécu). This is why symbols have, as it were, a "numinous aura"; they reveal that the modalities of the spirit are at the same time manifestations of life, and, consequently, they directly engage human existence. The religious symbol not only unveils a structure of reality or a dimension of existence; by the same stroke it brings a meaning into human existence. This is why even symbols aiming at the ultimate reality conjointly constitute existential revelations for the man who deciphers their message.
The religious symbol translates a human situation into cosmological terms and vice versa; more precisely, it reveals the continuity between the structures of human existence and cosmic structures. This means that man does not feel himself "isolated" in the cosmos, but that he "opens out" to a world which, thanks to a symbol, proves "familiar." On the other hand, the cosmological values of symbols enable him to leave behind the subjectivity of a situation and to recognize the objectivity of his personal experiences.
It follows that he who understands a symbol not only "opens out" to the objective world, but at the same time succeeds in emerging from his particular situation and in attaining a comprehension of the universal. This is explained by the fact that symbols have a way of causing immediate reality, as well as particular situations, to "burst." Whenever a tree incarnates the World Tree or when a spade is associated with the phallus and agricultural work with the act of generation, for example, one could say that the immediate reality of these objects or actions "bursts" or "explodes" under the irruptive force of a more profound reality. The same might be said of an individual situation, let us say, that of the neophyte locked up in the initiation hut. The symbolism "bursts" the bonds of this particular situation, making it exemplary, that is to say, indefinitely repeatable in many and varied contexts (because the initiation hut is likened to the maternal womb and at the same time to the belly of a monster and to Hell, and the darkness symbolizes the Cosmic Night, the preformal, the fetal state of the world, etc.). Consequently, because of the symbol, the individual experience is "awakened" and transmuted in a spiritual act. To "live" a symbol and to decipher its message correctly implies an opening toward the Spirit and, finally, access to the universal.
These few general remarks on religious symbolism should, of course, be elaborated and refined. Since it is impossible here to undertake a work which would require a whole book, let us confine ourselves to two more observations. The first concerns what may be called the "history" of a symbol. We have already alluded to the difficulty confronting the historian of religions when, in disengaging the structure of a symbol, he is led to study and compare documents belonging to different cultures and different historical moments. To say that a symbol has a "history" can mean two things: (a) that this symbol was constituted at a certain historical moment and that therefore it could not have existed before that moment; (b) that this symbol has been diffused, beginning from a precise cultural center, and that for this reason one must not consider it as spontaneously rediscovered in all the cultures where it is found.
That there have been symbols dependent upon precise historical situations seems indubitable in a number of cases. lt. is evident, for instance, that the horse could not have become, among other things, a symbol of death, before being domesticated. It is evident as well that the spade could not have been associated with the phallus, nor agricultural work homologized to the sexual act before the discovery of agriculture. In the same manner, the symbolism of the number 7 and, consequently, the image of the Cosmic Tree with seven branches did not appear before the discovery of the seven planets, which in Mesopotamia led to the conception of the seven planetary heavens. Moreover, there are numerous symbols that can be traced back to specific socio-political situations existing only in certain areas and taking form at precise historical moments, such as the symbols of royalty, of the matriarchate, or the systems implying the division of society into two halves at once antagonistic and complementary.
Since all this is true, it follows that the second meaning which the expression "history of a symbol" can have is equally true. The symbols depending on agriculture, royalty, the horse, and others were very probably diffused along with other elements of culture and their respective ideologies. But to recognize the historicity of certain religious symbols does not cancel out what we have said above about the function of religious symbols in general. On the one hand, it is important to note that although numerous, these symbols which are bound up with the facts of culture and thus with history, are appreciably less numerous than the symbols of cosmic structure or those related to the human condition. The majority of religious symbols reveal the World in its totality or one of its structures (night, water, heaven, stars, seasons, vegetation, temporal rhythms, animal life, etc.), or they refer to situations constitutive of all human existence, that is to say, to the fact that man is mortal, is a sexual being, and is seeking what today we call "ultimate reality." In certain cases, archaic symbols linked with death, with sexuality, or with hope for an afterlife have been modified or even replaced, by similar symbols brought in by waves of superior cultures. But these modifications, although they complicate the work of the historian of religions, do not change the central problem. To suggest a comparison with the work of the psychologist, when a European dreams of leaves of maize, the important fact is not that maize was imported into Europe only after the sixteenth century and thus became a part of the history of Europe, but that as an oneiric symbol, the maize is only one of innumerable varieties of the green leaf. The psychologist takes account of this symbolic value rather than of the historic diffusion of maize. The historian of religions finds himself in an analogous situation when he deals with archaic symbols that have been modified by cultural influences and events, for example, the World Tree, which in Central Asia and in Siberia received a new value by assimilating the Mesopotamian idea of the seven planetary heavens.
In sum, symbols linked with recent facts of culture, although they had a beginning in historic times, became religious symbols because they contributed to the "making of Worlds," in the sense that they allowed these new "Worlds" -- revealed through agriculture, through the domestication of animals, through royalty -- to "speak" and disclose themselves to men, revealing new human situations at the same time. In other words, symbols bound up with recent phases of culture are themselves constituted after the same manner as the most archaic symbols, that is, as the result of existential tensions and of ways of totally grasping the World. Whatever the history of a religious symbol may be, its function remains the same. A study of the origin and diffusion of a symbol does not release the historian of religions from the obligation of understanding it; it is for him to restore to it all the meanings it has had during the course of its history.
The second observation extends the first one in a way, since it bears on the capacity of symbols to become enriched in history. We have just seen how, under the influence of Mesopotamian ideas, the Cosmic Tree comes to symbolize, with its seven branches, the seven planetary heavens. In Christian theology and folklore the Cross is believed to rise up from the Center of the World, taking the place of the Cosmic Tree. But we have shown in a preceding study that these newly attributed meanings are conditioned by the very structure of the symbol of the Cosmic Tree. Salvation by the Cross is a new value bound to a precise historical fact -- the agony and death of Jesus -- but this new idea extends and perfects the idea of cosmic renovatio symbolized by the World Tree. (Eliade, Images et symboles, pp. 213 ff., Hugo Rahner, "The Christian Mystery and the Pagan Mysteries," The Mysteries: Papers from the Eranos Yearbooks [New York, 1955], II, 337-401, esp. 380 ff.)
All this could be formulated in another manner. Symbols are capable of being understood on more and more "elevated" planes of reference. The symbolism of darkness allows us to grasp its meaning not only in its cosmological and initiatory contexts (cosmic night, prenatal darkness, etc.), but also in the mystical experience of the "dark night of the soul" of St. John of the Cross. The case of the symbolism of the Symplegades is still more evident. As for the symbols expressing the coincidentia oppositorum, we know what role they have played in philosophical and theological speculations. But then one may ask if these "elevated" meanings were not in some manner implied in the other meanings, and if, as a consequence, they were, if not plainly understood, at least vaguely felt by men living on archaic levels of culture. This poses an important problem which unfortunately we cannot discuss here; how can one judge how far these "elevated" meanings of a symbol are fully known and realized by such and such an individual belonging to such and such a culture? (Cf. Eliade, "Centre du monde, temple, maison" in Le Symbolisme cosmique des monuments religieux [Home, 1957], pp. 57-82, esp. pp. 58 ff.) The difficulty of the problem rests in the fact that symbols address themselves not only to the awakened consciousness, but to the totality of the psychic life. Consequently, we do not have the right to conclude that the message of the symbols is confined to the meanings of which a certain number of individuals are fully conscious, even when we learn from a rigorous investigation of these individuals what they think of such and such a symbol belonging to their own tradition. Depth psychology has taught us that the symbol delivers its message and fulfills its function even when its meaning escapes awareness.
This admitted, two important consequences follow:
1. If at a certain moment in history a religious symbol has been able to express clearly a transcendent meaning, one is justified in supposing that this meaning might have been already grasped dimly at an earlier epoch.
2. In order to decipher a religious symbol, not only is it necessary to take into consideration all of its contexts, but one must above all reflect on the meanings that this symbol has had in what we might call its "maturity." Analyzing the symbolism of magic flight in a previous work, we came to the conclusion that it reveals dimly the ideas of "liberty" and of "transcendence," but that it is chiefly on the level of spiritual activity that the symbolism of flight and of ascension becomes completely intelligible. (Cf. Eliade, Mythes, réves et mystères, pp. 133 ff., esp. pp. 159 ff.) This is not to say that one must put all meanings of this symbolism on the same plane -- from the flight of shamans to the mystical ascension. However, since the "cipher" constituted by this symbolism carries with it in its structure all the values that have been progressively revealed to man in the course of time, it is necessary in deciphering them to take into account their most general meaning, that is, the one meaning which can articulate all the other, particular meanings and which alone permits us to understand how the latter have formed a structure.