Webster's Third New International Dictionary gives the following definition of myth: "a story that is usually of unknown origin and at least partially traditional, that ostensibly relates historical events usually of such character as to serve to explain some practice, belief, institution, or natural phenomenon, and that is especially associated with religious rites and beliefs." This indeed is as good a definition as any, clear and all-inclusive, highlighting the essential meanings of the word in its most general sense. Many specialists in the field of mythology, however, are not satisfied with such a broad interpretation of the term "myth." They attempt to distinguish 44 true myth" or "myth proper" from other varieties, and seek to draw distinctions in terminology between it and other words often used synonymously, such as legend, saga, and folk tale. Thus, for some, myth must be ancient and of anonymous origin (according to Gayley, "myths are born, not made); others demand that they invariably be concerned with nature and provide explanations for its phenomena (i.e., real myths are nature stories and aetiological); still others would expand their definition to include tales dealing with all circumstances of human life, although they might also insist upon the presence of the divine or supernatural for authenticity. There are those who argue that myths must have once been believed, although subsequently they come to be regarded as incredible. Thus Christian dogma for them may be excluded from mythological study (although to non-Christians it may be viewed as incredible): and so on.

Robert Graves, for example, distinguishes true myth, which for him is "the reduction to narrative shorthand of ritual mime performed in public festivals, and in many cases recorded pictorially on temple walls, vases, seals, bowls, mirrors, chests, shields, tapestries, and the like," from twelve other categories, such as: philosophical allegory, satire or parody, minstrel romance, political propaganda, theatrical melodrama, realistic fiction. The definitions set forth by Rose in his invaluable handbook have deservedly won wide acceptance. His distinctions are clear, serviceable, and enlightening; although inevitably artificial per se, they are by no means intended to be adopted as rigid. In reality the criteria merge and the lines of demarcation blur. For him, true myth or myth proper is ultimately "the result of the working of naive imagination upon the facts of experience". It is to be distinguished from folk tale, whose major function is to present a story primarily for the sake of entertainment, and from saga, which, however fanciful and imaginative, is rooted in historical fact? Rarely, if ever, do we find a pristine, uncontaminated example of any one of these forms. Yet the traditional categories of myth, folk tale, and saga are useful guides as we try to impose some order upon the multitudinous variety of classical legends. Myth is a comprehensive (but not exclusive) term for stories primarily concerned with the gods and man's relations with them; saga, as we have seen, has a perceptible relationship to history. These two categories underlie the basic division of the first two parts of this book into The Myths of Creation; the Gods and The Greek Sagas: Greek Local Legends. Interwoven with these broad categories of legend are folk tales, which are often tales of adventure, sometimes peopled with fantastic beings and enlivened by ingenious strategies on the part of the hero. How loose these categories are can be seen, for example, from the stories of Odysseus or the Argonauts. In view of the shifting limits of definitions of myth we could be forgiven for taking refuge in a vague and comprehensive phrase such as Geoffrey Kirk's "traditional tale." As Kirk points out, this phrase emphasizes that "a myth is a story, a narrative with a dramatic structure and a climax," and that myths "have succeeded in becoming traditional . . . important. enough to be passed from generation to generation."

THEORIES AND INTERPRETATIONS OF MYTH

The impossibility of establishing a satisfactory definition of "myth" has not deterred scholars from developing theories on the meaning and interpretation of myth, which usually provide bases for a hypothesis about origins. Useful surveys of the principal theories are easily available, so that we shall attempt to touch upon only a few theories that are likely to prove especially fruitful to the student or are persistent enough to demand his attention. One thing is certain: no single theory of myth can cover all myths. The variety of traditional tales is matched by the variety of their origins and significance, so that any monolithic theory cannot succeed in achieving universal applicability.   The desire to rationalize classical mythology arose far back in classical antiquity, and is especially associated with the name of Euhemerus (ca. 300 B.c.), who claimed that the gods were men deified for their great deeds. At the opposite extreme are the antirationalists who favor metaphorical interpretations, which stem from the belief that traditional tales hide profound meanings. At its best the metaphorical approach sees myth as allegory, where the details of the story are but symbols of universal truths. At its worst the allegorical approach is a barren  exercise in cryptology: to explain the myth of Ixion and the Centaurs (cf. pp. 444-45) in terms of clouds and weather phenomena is hardly enlightening and not at all ennobling. The metaphorical approach has taken many forms in the present century through the theories of the psychologists, most especially those of Freud and Jung. Freud's discovery of the significance of dream-symbols led him and his followers to develop the similarity between dreams and myths. Myths, therefore, in the Freudian interpretation, reflect waking man's efforts to systematize the incoherent visions and impulses of his sleep world. The patterns in the imaginative world of children, savages, and neurotics are similar, and these patterns are revealed in the motifs of myth. One of the earliest and basic patterns is that of the Oedipus story, in which the son kills the father in order to possess the mother. From the ensuing sense of guilt and sin emerges the conception of God as Father who must be appeased and to whom atonement must be made. In fact, according to Freud, the Oedipus complex has inspired the beginning not only of religion but also of all ethics, art, and society.

It is clear that Freud's connection between dreams and myths is important, but it is equally clear that such an interpretation cannot cover all myths. Nevertheless, Freud's approach is especially illuminating for some myths, for example, the legend of the Minotaur or the saga of the House of Atreus, which deal with some of the most persistent, if repressed, human fears and emotions and, by their telling, achieve a kind of catharsis.   Freud's former disciple, Jung, went beyond the mere connection of myths and dreams to interpret myths as the projection of what he called the "collective unconscious" of the race, that is, as a revelation of the continuing psychic tendencies of society. Thus myths contain images or "archetypes" (to use Jung's term) that are the traditional expressions, developed over thousands of years, of symbols upon which the society as a whole has come to depend. Such archetypes are the persona (character of the individual), the animus and anima (respectively man's idea of woman, woman's of man), the old wise man, the great mother, and so on. These appear in the dreams of individuals or are expressed in the myths of societies. The great value of Jung's concept is that it emphasizes the psychological dependence of all societies (sophisticated as well as primitive) upon their traditional myths, often expressed also in religion and ritual; but Jung's theories are open to the same limitation as others, that they do not cover all myths.   We have referred to H. J. Rose's definition of myth as "the result of the working of naive imagination upon the facts of experience." The word "naive" implies a type of imagination less sophisticated than that, for example, of the articulate and literary Greeks.

This is both instructive and misleading; it is important to realize that classical mythology developed from something less sophisticated than the form in which we find it in, say, Homer or Euripides, and if we do this we shall avoid a lot of the romantic nonsense that has been perpetuated about the nobility and purity of the Greek imagination, as if the Greeks had no irrational side to their psychology.   On the other hand, it is misleading to press analogies and parallels in the legends of widely divergent societies, especially where, as in Greek mythology, even our earliest literary sources (Homer and Hesiod) appear after a lengthy period of evolution, far removed from the myths' primitive origins. It is equally misleading to posit a "primitive" mentality as if it were something childlike and simple, in contrast to the "sophisticated" mentality of more advanced societies such as the Greeks". In fact, anthropologists have proved how far the myths of  primitive societies reflect the complexities of social family structures; and their tales are far from being merely alogical and mystical. Despite these warnings, however, we may assert that the comparative study of myths, especially by anthropologists (as opposed to philologically -trained classicists), has been one of the most fruitful approaches to the interpretation of myths. Sir J. G. Frazer's The Golden Bough remains a pioneering monument in the field, for all its faults. It is full of comparative data on kingship and ritual, but its value is lessened by Frazer's ritualist interpretation of myth (to be explained below) and by his eagerness to establish analogies between myths of savage tribes and classical myths. Equally important in the development of modern theories is the work of Bronislav Malinowksi, who was stranded among the Trobriand islanders (off New Guinea) during the First World War; he used his enforced leisure to study the Trobrianders and published his "Myth in Primitive Psychology" in 1926. His great discovery was the close connection between myths and social institutions, which led him to explain myths not in cosmic or mysterious terms, but as "charters" of social customs and beliefs. To him myths were related to practical life, and they explained existing facts and institutions by reference to tradition: the myth confirms (i.e., is the "charter" for) the institution, custom or belief.

Clearly such a theory will be valid for only certain myths (for example, those involving the establishment of a ritual), but any theory that excludes the speculative element in myth is bound to be too limited. We have referred earlier to three other approaches to myth that have been important since the nineteenth century, and we end our survey of general interpretations with a brief summary of each. The oldest of the three was the theory of Max Miiller (whose essay on "Comparative Mythology" was published in 1856) that myths are nature myths, all referring to meteorological and cosmological phenomena. This is, of course, an extreme development of the allegorical approach, and it is hard to see how or why all myths can be explained as allegories of, for example, day replacing night, winter succeeding summer, and so on. True, some myths are nature myths, and certain gods, for example, Zeus, represent or control the sky and other parts of the natural order; yet it is just as true that a great many more myths have no such relationship to nature. Further, the Greeks (as opposed to other myth-making societies) do not seem to have given as much attention to nature in their myths as this theory implies. Anyone who attempts to fix the details about Mt. Olympus, or about the sun-god, will soon see how vaguely defined the concepts of Greek mythology were in these areas.   A second universalist theory is that which says that traditional tales should be interpreted as explanations of some fact or custom. Hence the theory is called "etiological," from the Greek word for a cause (aition). On this theory the mythmaker is a kind of primitive scientist, using myths to explain facts that cannot otherwise be explained within the limits of society's knowledge at the time. This theory, again, is adequate for some myths (for example, those that account for origin of certain rituals), but it does not allow for the imaginative or metaphysical aspects of mythological thought. The third and final general theory is the ritualist interpretation of myth, which underlies Robert Graves' definition of "true myth" quoted earlier. Stated most bluntly, this theory says that "myth implies ritual, ritual implies myth, they are one and the same". True, many myths are closely connected with rituals, and the theory is valuable for the connection that it emphasizes between myth and religion. But it is patently untenable to connect all myth with ritual, and an examination of known Greek rituals will soon show how far Greek mythology fails to be covered by the ritualist theory.

 The conclusion of G. S. Kirk seems inescapable: "On all the evidence, the great majority of Greek myths were developed without any special attention to ritual". More recently the structural theories of Claude Lévi-Strauss have enriched the anthropological approach to myth, and they invite the student to observe Malinowski's most important concept, that is, the connection between myth and society. Lévi-Strauss sees myth as a mode of communication, like language and music. In music it is not the sounds themselves that are important but their - structure, i.e., the relationship of sounds to other sounds. In myth it is the narrative that takes the part of the sounds of music, and the structure of the narrative can be perceived at various levels and in different codes (for example, culinary, astronomical, sociological). From this it follows that no one version of a myth is the "right" one; all versions are valid, for myth, like society, is a living organism in which all the parts contribute to the existence of the whole. As in an orchestral score certain voices or instruments play some sounds, while the whole score is the sum of the individual parts, so in a myth the different, partial, versions combine to reveal its total structure, including the relationship of the different parts to each other and to the whole. Lévi-Strauss' method is therefore rigorously analytical, breaking down each myth into its component parts. Underlying his analytical approach are basic assumptions, of which the most important is that all human behavior is based on certain unchanging patterns whose structure is the same in all ages and in all societies. Second, he assumes that society has a consistent structure and therefore a functional unity in which every component plays a meaningful part. Myths are part of the working of this social machine and are derived ultimately from the structure of the mind. And the basic structure of the mind, as of the myths that it creates, is binary. That is, the mind is constantly dealing with pairs of contradictions or opposites, and it is the function of myth to mediate between these opposing extremes-rawlcooked, lifeldeath, hunterlhunted, nature/culture, and so on. "Mythical thought always progresses from the awareness of oppositions towards their resolution".  Myth, then, is a mode by which a society communicates and through which it finds a resolution between conflicting opposites. The theories of Lévi-Strauss have aroused passionate controversy among anthropologists and students of mythology.

The only point that is generally agreed upon is that they are so significant that no student can afford to ignore them. The best starting-point for a study of Lévi-Strauss is the Overture to The Raw and the Cooked, and students of classical mythology should at least read his article "The Structural Study of Myth," which includes his interpretation of the Oedipus myth. As we have already seen, L6vi-Strauss likens each element in the structure of the myth to the part to be played by an instrument in an orchestral score. Thus, by assigning to each element a number, one achieves a structure analogous to a musical score, so that at each stage of the myth one or more structural elements appear together, while the part played by each element can be expressed in linear fashion by reading the "score" from left to right. In the Oedipus myth Lévi-Strauss distinguishes four columns of elements that have common aspects: (1) overrating of blood relations, (2) underrating of blood relations, (3) killing of monsters, and (4) names expressive of physical and moral handicaps. Thus the structure shown on p. 9 evolves. Reading the score vertically, the student sees groups of common elements in the myth. Reading horizontally, one sees the development of an individual's part in the structure of the myth. From all this, L6vi-Strauss deduces the elements common to each vertical column (a deduction that we have anticipated in our account); from column 3 he deduces that the myth concerns "the autochthonous origin of mankind" (serpents being chthonic symbols) and its denial (symbolized by killing). He deduces that column 4, because its names suggest an inability to walk properly or clumsiness, implies the persistence of the autochthonous origin of mankind, because in mythology "it is a universal character of men born from the earth that at the moment they emerge from the depth, they either cannot walk or do it clumsily." Thus column 4 denies column 3, just as column 2 contrasts with column 1. So (although Lévi-Strauss does not draw these conclusions), the Oedipus myth attempts to mediate between opposites-overratinglunderrating of blood relations; mankind sprung from the earthlmankind sprung from the union of human parents, and so on. Lévi-Strauss would maintain that all versions of the myth (including Freud's use of it) are equally authentic for a student who is exploring the myth's structure. The structure of the myth provides a "logical model by means of which the human mind can evade unwelcome contradictions," according to Edmund Leach, and so provides a means of "mediating" between opposites that would, if unreconciled, be intolerable. Whatever one may think of Levi-Strauss' interpretation of the Oedipus myth (and it has been widely criticized), there is no doubt that this structural approach can be illuminating for a qumber of Greek myths, especially with regard to the function of "mediating." But it is open to the same objections as other comprehensive theories, that it establishes too rigid, too universal a concept of the functioning of the human mind. Indeed, the binary functioning of the human mind and of human society may be common, but it has not been proved to be either universal or necessary. Finally, Lévi- Strauss draws most of his evidence from primitive and preliterate cultures, and his theories seem to work more convincingly for them than for the literate mythology of the Greeks. His approach is better applied, for example, to the early Greek succession myths than to the Sophoclean, literate version of the legends of Oedipus and his family. We should all the same be aware of the potential of Lévi-Strauss' theories and be ready to use them as we seek to make meaningful connections between the different constituent elements of a myth, or between different myths that share constituent elements.

Structuralism has often proved to be a liberating force and, despite its rigid logic, to be able to give flexibility to the student, who might otherwise be perplexed by conflicting elements within, or conflicting versions of, a myth. The structural interpretation of myth was developed long before the work of Lévi-Strauss by Vladimir Propp in his study of the Russian folk tale.  Like Lévi-Strauss, Propp analyzed traditional tales into their constituent parts, from which he deduced a single, recurrent structure, applicable to all Russian folk tales. Unlike Lévi-Strauss, however, he described this structure as linear, that is, having an unchanging temporal sequence, so that one element in the myth always follows another and never occurs out of order. This is significantly different from the pattern in Lévi-Strauss' theory, where the elements may be grouped without regard to time or sequence. A diagram of a Lévi-Straussian structure may be "read" vertically and horizontally, but Propp's structure must be read horizontally, left to right. In technical terms, Propp's structure is diachronic, that of Lévi-Strauss, synchronic. Propp divided his basic structure into thirty-one functions or units of action (which have been defined by others as "motifemes," on the analogy of "morphemes" and "phonemes" in linguistic analysis). These functions are constants in traditional tales: the characters may change, but the functions do not. Further, these functions always occur in an identical sequence, although notall the functions need appear in a particular tale. Those that do, however, will always be in the same sequence. Finally, Propp states that "all fairy tales are of one type in regard to their structure". Propp was using a limited number (100) of Russian folk tales of one sort only, that is, the Quest. Yet his apparently strict analysis has proven remarkably adaptable and valid for other sorts of tale in other cultures. The rigid sequence of functions is too inflexible to be fully applicable to Greek myths that have a historical dimension (for example, some of the tales in the Trojan cycle of saga), where the "facts" of history, so far as they can be established, may have a sequence independent of structures whose origins lie in psychological or cultural needs. On the other hand, Propp's theories are very helpful in comparing myths that are apparently unrelated, showing, for example, how the same functions appear in the myths, whatever the names given to the characters who perfom them. As any student of mythology soon discovers, mythological names are a strain on the memory.

Merely to master these catalogues is to have achieved very little, unless they can be related in some meaningful way to other tales, including tales from other mythologies. The dreary labor of memorization, however, becomes both easier and purposeful if underlying structures and their constituent unit can be perceived and can be arranged logically and consistently. A very simple example would be the structural elements common to the myths of Heracles, Theseus, Perseus, and Jason, whose innumerable details can be reduced to a limited sequence of functions. It is more difficult to establish the pattern for, say, a group of stories about the mothers of heroes (for example, Callisto, Danae, 10, and Antiope). Yet they resolve themselves (as Burkert has shown) into a clear sequence of five functions: (1) the girl leaves home; (2) the girl is secluded (beside a river, in a tower, in a forest, etc.); (3) she is raped by a god; (4) she suffers punishment or rejection or a similar unpleasant consequence; and (5) she is rescued, and her son is borm.   In concluding this discussion of structural explanations of myth, we can say definitely that in most cases it is helpful to the student to analyze a myth into its constituent parts. There should be four consequences:  

1. A perceptible pattern or structure will emerge.  

2. It will be possible to find the same structure in other myths, thus making it easier to organize the study of myths.  

3. It will be possible to compare the myths of one culture with those of another.  

4. As a result of this comparison, it will be easier to study the development of a myth prior to its literary presentation.

Structuralism, then, is basically a means toward establishing a rational system for studying mythology. Its theories need not beindeed, cannot be-applied to all classical mythology, nor need one be enslaved to either L6vi-Strauss or the more rigid but simpler structure of Propp's thirty-one functions. On the other hand, if the study of mythology is to become something more than the sterile memorization of names, a student who uses the help of structural theories will be more likely to discover a rational organization in mythology. In 1977 Walter Burkert, in a series of lectures at the University of California? persuasively attempted a synthesis of structural theories with the more traditional approaches to classical mythology. In defining a theory of myth he developed four "theses," which are in part based upon structural theories and in part meet the objection that these theories are not adequate for many Greek myths as they have come down to us after a long period of development. According to Burkert, classical myths have a "historical dimension" with "successive layers" of development, during which the original tale has been modified to fit the cultural or other circumstances of the time of its retelling. This will be less true of a tale that has sacred status, for it will have been "crystallized" in a sacred document- for example, the myth of Demeter in the Homeric Hymn to Demeter. In contrast, many Greek myths vary with the time of telling and the teller-for example, the myths of Orestes or Meleager appear differently in Homer from their treatment in fifthcentury Athens or in Augustan Rome. Burkert therefore believes that the structure of traditional tales cannot be discovered without taking into account cultural and historical dimensions. With regard to the former, the structure of a tale is shaped by its human creators and by the needs of the culture within which it is developed. Therefore the structure of tale is "ineradicably anthropomorphic" and fits the needs and expectations of the teller and audience. (Indeed, as Burkert points out, this is why good tales are so easy to remember: "There are not terribly many items to memorize, since the structure has largely been known in advance.") Further-and here we approach the historical dimension-a tale has a use to which it is put, or, expressed in another way, "Myth is traditional tale applied." This refinement of the structural theory allows for the development of a tale to meet the needs or expectations of the group for whom it is told-family, city, state, or culture group, for example.

A myth, in these terms, has reference to "something of collective importance." This further definition meets a fundamental objection to many earlier "unitary" theories of myth. If myth is a sacred tale or a tale about the gods, how do we include, for example, the myths of Oedipus or Achilles? Similar objections can easily be made to other theories that we have been describing. The notions of "myth applied" and "collective importance" avoid the objection of rigid exclusivity, while they allow for the successive stages in the historical development of a myth without the Procrustean mental gymnastics demanded by the theories of Lévi-Strauss. Here, then, are the four theses of Burkert7s modified synthesis of the structural and historical approaches: 1. Myth belongs to the more general class of traditional tale. 2. The identity of a traditional tale is to be found in a structure of sense within the tale itself. 3. Tale structures, as sequences of motifemes, are founded on basic biological or cultural programs of actions. 4. Myth is a traditional tale with secondary, partial reference to something of collective importance. These theses form a good working basis upon which to approach the interpretation of myth. They make use of the significant discoveries of anthropologists and psychologists, while they allow the student flexibility in exploring the structure of classical myths. Finally, they take account of the historical development of myths and of the culture within which they were told. It will be useful to refer back to these theses when studying individual traditional tales. Over the last few decades comparative mythology has proved essential to the understanding of the myths of any one culture. Greek mythology, largely because of the genius of the authors who told the stories in their literary form, has too often in the past been considered as something so unique that it can be set apart from other mythologies. Indeed, even the earliest literary sources for Greek mythology (Homer and Hesiod) provide artistic presentations of intellectual and spiritual concepts of the highest order, while the sophisticated versions of later Greek and Roman authors are clearly to be differentiated from many preliterate tales gathered from other cultures by anthropologists. The work of Lévi-Strauss and many others has shown that classical myths share fundamental structures with this latter class of traditional tales. It is important to be aware of this fact and to realize that there are many successive layers (to use Burkert's phrase) in the development of Greek and Roman myths before their crystallization in literary form.

Often, and especially in structural interpretations, the earlier stages of a myth are discovered to have been rooted in another culture, or at least show the influence of other mythologies. For example, there are obvious parallels between the Greek creation and succession myths and myths of Near Eastern cultures. The myth of the castration of Uranus by Cronus is better understood if we compare it with the Hittite myth (first published in 1945) of Kumarbi, in which Anu, the sky-god, is castrated by Kumarbi, who rises against him. Kumarbi swallows Anu's genitals, spits them out when he cannot contain them, and is finally replaced by the storm-god. The structure of this tale is paralleled by the myth of Uranus, castrated by Cronus, who, in his turn, cannot hold what he has swallowed (in this case, his children) and is eventually replaced by the sky-god, Zeus. Some details in the two tales, of course, are different, but the basic functions (kingship, revolt, castration, swallowing, regurgitation, replacement by a new king) are the same and occur in the same sequence. Thus the basic structure is the same and a better understanding of the origin and purpose of the Greek myth, as narrated by Hesiod, is achieved by comparison with the older myth from the Near Eastern culture. Whether or not direct influence can be proved (and scholars are not agreed upon this point), the structural similarities do at least show that Greek myths cannot be studied in isolation from those of other cultures.