Chapter 1. Introduction

In the early thirties of this century, when Milman Parry began to write the book from which this one takes its name, [1] what was needed most in Homeric scholarship was a more exact knowledge of the way in which oral epic poets learn and compose their songs. Now in the late fifties of the same century the need is still great; in spite of the number of books about Homer and his poems, about epic poetry in general, and about specific epic traditions in various parts of the world, the student of epic still lacks a precise idea of the actual technique of poiesis in its literal meaning. Thanks to Parry, however, we have the material for the research necessary to determine what this technique is. He has left us his collection of South Slavic texts, which is the record on phonograph discs and in manuscripts of experiments in the laboratory of the living epic tradition of the Yugoslavs. [2]

In 1935 Milman Parry was Assistant Professor of Classics at Harvard University. He had already made a name for himself in classical scholarship by his masterly analysis of the technique of the formulaic epithets in the Iliad and the Odyssey. [3] This work had convinced him that the poems of Homer were traditional epics, and he soon came to realize that they must also be oral compositions. [4] He therefore set himself the task of proving, incontrovertibly if it were possible, the oral character of the poems, and to that end he turned to the study of the Yugoslav epics. In the autumn of 1935, he wrote:

In Part I of this book I shall attempt to fulfill Parry's purpose of setting forth with exactness the form of oral narrative poetry, drawing my illustrative {3|4} material from this collection; in Part II I shall use the principles presented in Part I in studying the form of some of the great epic poems from the past. Because I intend to limit the scope of this book to a consideration of oral form and manner of composition, a discussion of a broader sort which would aim at seeing "how the way of life of a people gives rise to a poetry of a given kind and a given degree of excellence" will not be fully entered upon. Yet considerations of this kind will inevitably occupy us to some extent in this book. It is hoped that what is said here will be of use for future comparative study of oral poetry.

***

The burden of the first few chapters of Part I will be to work out in fullness of detail a definition of oral epic song. Stated briefly, oral epic song is narrative poetry composed in a manner evolved over many generations by singers of tales who did not know how to write; it consists of the building of metrical lines and half lines by means of formulas and formulaic expressions and of the building of songs by the use of themes. This is the technical sense in which I shall use the word "oral" and "oral epic" in this book. By formula I mean "a group of words which is regularly employed under the same metrical conditions to express a given essential idea." This definition is Parry's. [6] By formulaic expression I denote a line or half line constructed on the pattern of the formulas. By theme I refer to the repeated incidents and descriptive passages in the songs.

These definitions are but the bare bones of the living organism which is oral epic. We shall peer into the structural heart of the formulas to discern the various patterns which merge to give them form. We shall see that the formulas are not the ossified clichés which they have the reputation of being, but that they are capable of change and are indeed frequently highly productive of other and new formulas. We shall come to realize the way in which themes can be expanded and contracted, and the manner in which they are joined together to form the final product which is the song. We shall note the difference both in the internal structure and in the external connection of themes as they are used by different singers.

Finally we shall turn our attention to the song itself. We shall see that in a very real sense every performance is a separate song; for every performance is unique, and every performance bears the signature of its poet singer. He may have learned his song and the technique of its construction from others, but good or bad, the song produced in performance is his own. The audience knows it as his because he is before them. The singer of tales is at once the tradition and an individual creator. [7] His manner of composition differs from that used by a writer in that the oral poet makes no conscious effort to break the traditional phrases and incidents; he is forced by the rapidity of composition in performance to use these traditional elements. [8] To him they are not merely necessary, however; they are also right. He {4|5} seeks no others, and yet he practices great freedom in his use of them because they are themselves flexible. His art consists not so much in learning through repetition the time-worn formulas as in the ability to compose and recompose the phrases for the idea of the moment on the pattern established by the basic formulas. He is not a conscious iconoclast, but a traditional creative artist. His traditional style also has individuality, and it is possible to distinguish the songs of one singer from those of another, even when we have only the bare text without music and vocal nuance.

***

The need for a clarification of the oral process of composition is reflected in the many terms which are used for oral narrative poetry. To no small degree difficulties have arisen because of the ambiguity of terminology and because each school has chosen a different facet of this poetry as distinctive. The term "oral" emphasizes, I believe, the basic distinction between oral narrative poetry and that which we term literary epic. But it too carries some ambiguity. Certain of the misunderstandings of Parry's oral theory arise from the failure to recognize his special use of the word "oral." For example, one often hears that oral poetry is poetry that was written to be recited. Oral, however, does not mean merely oral presentation. Oral epics are performed orally, it is true, but so can any other poem be performed orally. What is important is not the oral performance but rather the composition during oral performance. [9]

There may be ambiguity also when we say that the oral poet learns his songs orally, composes them orally, and transmits them orally to others. Like so many statements made in the debate on the oral theory, this one too is perfectly true if the word "oral" is understood in the technical sense in which it will be presented in this book. But if the reader interprets oral learning as listening to something repeated in exactly the same form many times, if he equates it with oral memorization by rote, then he will fail to grasp the peculiar process involved in learning oral epic. The same may be said for oral composition. If we equate it with improvisation in a broad sense, we are again in error. Improvisation is not a bad term for the process, but it too must be modified by the restrictions of the particular style. The exact way in which oral composition differs from free improvisation will, I hope, emerge from the following chapters. It is true also that oral epic is transmitted by word of mouth from one singer to another, but if we understand thereby the transmission of a fixed text or the kind of transmission involved when A tells В what happened and В tells С and so on with all natural errors of lapse of memory and exaggeration and distortion, then we do not fully comprehend what oral transmission of oral epic is. With oral poetry we are dealing with a particular and distinctive process in which oral learning, oral composition, and oral transmission almost merge; they seem to be different facets of the same process. {5|6}

The word "epic," itself, indeed, has come in time to have many meanings. Epic sometimes is taken to mean simply a long poem in "high style." Yet a very great number of the poems which interest us in this book are comparatively short; length, in fact, is not a criterion of epic poetry. Other definitions of epic equate it with heroic poetry. Indeed the term "heroic poetry" is sometimes used (by Sir Cecil M. Bowra, for example) to avoid the very ambiguity in the word epic which troubles us. Yet purists might very well point out that many of the songs which we include in oral narrative poetry are romantic or historical and not heroic, no matter what definition of the hero one may choose. In oral narrative poetry, as a matter of fact, I wish to include all story poetry, the romantic or historical as well as the heroic; otherwise I would have to exclude a considerable body of medieval metrical narrative.

That whole body of verse that we have now agreed to designate as oral has been called by many names; the terminological battle is a serious one. Those who call it "folk epic" are carrying on a nineteenth century concept of composition by the "folk" which has long since been proved invalid. At one time when "folk epic" referred to a theory of composition, it was a justifiable term. It pointed to a method of composition as the distinction between oral narrative poetry and "written" poetry. It was looking in the right direction. But when its theory of composition was invalidated, because no one could show how the people as a whole could compose a poem, then the technical meaning of the term was lost and it came to be equated in a derogatory sense with "peasant." The attention was then shifted from the way in which the poetry was made, first to the social status of those who practiced it, and then to the content and quality of the poetry itself. Although it may be true that this kind of poetry has survived longest among peasant populations, it has done so not because it is essentially "peasant" poetry, but rather because the peasant society has remained illiterate longer than urban society. [10] Indeed this poetry has more often been aristocratic and courtly than of the folk. It would seem even from its origins to have belonged to serious ceremonial occasions, to ritual, to celebration. The term "folk poetry" becomes more and more inadequate, more and more restricted in time and place. To apply the term to the medieval epics or to the Homeric poems is ever more inadmissible.

Another reason why this poetry should cease to be denominated as "folk epic" is that outside the circle of folklore enthusiasts the connotations of "folk" in many countries tend to be derogatory. One thinks of the simple peasant with his "quaint" ideas, his fairy stories, and children's tales. The use of folk stories as entertainment for young children has its ironic aspects; we are beginning to realize the serious symbolism and meaning of folk tales, which, if rightly understood, would be far from proper fare for children. Moreover, if we mean by "folk epic" to indicate that oral epic shares some of its subject matter with folk tale and all that is seriously {6|7} implied in that term, we are ignoring or underestimating all the other subjects of oral epic, historical, legendary, and heroic: we have outgrown the appellation "folk epic." It is no longer exact, and in time it has come to misrepresent oral epic poetry rather than to describe it.

Similar objections can be brought against the term "popular," the Latin derivative equivalent to "folk." While this term avoids the "simple peasant" connotations of "folk," its literal meaning has been overlaid with another set of unfortunate implications from its use in English to denote "popular music" and "popular songs."

The fever of nationalism in the nineteenth century led to the use of oral epics for nationalist propaganda. The poems glorified the heroes of the nation's past; they depicted the struggles of the nation against outside foes. Hence the hero emerged as a "national" hero, and the poems themselves were labeled "national" epics. In some of the Slavic countries the word narodni has a useful ambiguity, since it means both "folk" and "national." As a term to designate oral epic "national" is woefully inadequate and an insidious imposter.

Some scholars have sought to avoid the pitfalls of the three terms already discussed, folk, popular, and national, by recourse to the word "primitive." It sounds somehow more "scientific" because it has been borrowed from the social science of anthropology. But here too the ambiguity is great and the connotations hardly less flattering than those of "folk" in some countries. If the idea behind the use of "primitive" for this poetry is that oral epic poetry precedes written poetry in time in the cultural growth of a society, then its use would be legitimate, because as a rule oral poetry does precede written poetry, but it would, like the other terms, still miss the fundamental difference in form between the two.

In summary, any term that is used to designate oral narrative poetry in an attempt to distinguish it from written narrative poetry must contain some indication of the difference in form. It is because the terms which we have discussed above failed to comprehend this distinction that they have proved themselves to be inadequate. Any terms, also, carrying implications derogatory to either oral narrative poetry or written poetry (as, for example, such terms as "authentic" and "artificial"; "primary" and "secondary") must be abandoned, for they represent an attitude that is neither scholarly nor critical. Both these forms are artistic expressions, each with its own legitimacy. We should not seek to judge but to understand.

If the need for a clarification of the process which produces oral narrative poetry is reflected in the confusion of terms which have been used to designate that poetry, this need is even more apparent, of course, in the variety of theories put forth in the last two centuries (and which still survive in one form or another today) to explain the peculiar phenomenon of oral epic. On the one hand there has been a solid block of loyalists to the literary tradition who have maintained through thick and thin that the {7|8} Homeric poems, as well as the great epics from medieval times, are written literary productions by a single author.

These loyalists have found themselves defending their position from attacks by those who from time to time raised annoying questions. One of the earliest questions posed was whether writing existed in the ninth century B.C., the traditional date of Homer. This was first raised by Josephus; [11] it came to the fore again in D'Aubignac [12] in the early eighteenth century and reached its classic expression in Friedrich August Wolf's famed Prolegomena (1795). A second problem was formulated during the seventeenth century and played a great role in the Querelle des Anciens et des Modernes; this was the problem of the "errors" or inconsistencies in the Homeric poems. D'Aubignac, Perrault, Giambattista Vico, Robert Wood, and others led once again to Wolf and the later Separatists. A third question concerned the unusual length of the Homeric poems. If there was no writing in Homer's time, how could such long poems be preserved until the time of writing? In fact, how could poems of such length come into being at all without the aid of writing? Clearly this was a corollary of the first question raised. Among the earlier scholars who attempted to answer this question the name of Robert Wood stands out. A fourth problem arose from the increased knowledge of and interest in medieval minstrelsy and contemporary oral poetry during the eighteenth century and later. Here again we may begin with D’Aubignac and continue with Thomas Blackwell, Percy, Macpherson, Herder, Goethe, and a host of others. There was, fifthly, also the problem, inherited from ancient times, of the meaning of the Peisistratean legend about the recension of the Homeric poems. And finally with the development of linguistic studies in the nineteenth century the question was raised about the possibility of one man using dialect forms from several regions and archaisms from different periods.

These were the chief questions that were current in Homeric scholarship and still are. In answering them some scholars have gone so far as to deny even the existence of Homer, but the usual answer has been some form of multiple authorship for the poems with Homer at one end or the other of a series of poets. Sometimes he was the originator whose poems were carried through oral tradition or whose works were modified by later poets; more often he was the last of the redactors or compilers or, in an attempt to bridge the gap between Unitarians and Separatists, he was the great poet who reworked oral tradition into a "literary" poem. The concept of multiple authorship led scholars naturally to the dissection of the Homeric poems in an attempt to see what parts were done by different authors. They were thus led also to seek the "original" or archetype of the poems.

The doubt as to the existence of writing in Homer's time has given Homerists three choices: to seek proof that the doubt was ill founded and that there was writing as early as the traditional date of Homer; to change Homer's date, bringing it down to a period when writing was possible; to {8|9} leave Homer's date where it was and to cover the intervening years to the age of writing by oral transmission of Homer's poems.

The most significant step in proving that writing was possible and indeed existed not only in the ninth century but earlier has been made only recently in the discovery that Linear В is Greek. Although none of the texts yet deciphered is literary, [13] the old argument of Josephus and Wolf has had the ground cut from under it. Ironically enough, however, the proof has come at a time when many scholars realize that the existence of writing or even of a literary tradition does not necessarily mean that Homer's poems belong in the category of "written" literature; and many realize, and have realized for some time, that obviously our Homeric texts could not have been preserved had there not been writing in Homer's day. Valuable as the decipherment of Linear В is, it is no longer relevant to the Homeric problem.

Unless not only literary texts are discovered in Linear В but also some evidence can be unearthed to prove that epic poetry was being written down during the period of Linear B, its decipherment cannot help us much in determining Homer's date. Some Homeric names seem to have been deciphered, Hector and Achilles, for example, but this might indicate no more than that these were common names or that songs about these heroes existed; it tells us nothing about our poems. There is no evidence at all at this point that Homer was written down in Linear В and later copied in the Greek alphabet that we know. Were there such evidence, we would be justified in moving Homer's date back. So the problem of the date of Homer still remains with us.

Wolf and some of his predecessors turned to the Peisistratean legend to answer the question of date, as has Carpenter in our own century, although the latter's reasons are different from those of Wolf. Carpenter reflects our growing knowledge of oral literature and seeks a time when the writing down of the poems would make sense in the context of this knowledge. To him Peisistratus seems the most likely person to sponsor this. Certainly he is right that the Peisistratean legend is an invaluable clue, one that cannot be ignored but which demands explanation and interpretation. But Carpenter has been little heeded, and the date accepted now by most scholars is the second half of the eighth century.

Those scholars who made the third choice were moving in the right direction, namely towards oral tradition, but in putting the poet of our Homeric texts before the period of writing, they were unwittingly creating more problems than they were as yet equipped to handle. Their choice was a compromise. Oral tradition was a fickle mistress with whom to flirt. But scholars could call in to their help the "fantastic memories" so "well attested" of illiterate people. They felt that a text could remain from one generation to another unaltered, or altered only by inconsequential lapses of memory. This myth has remained strong even to the present day. The main points of confusion in the theory of those scholars who made the third {9|10} choice arose from the belief that in oral tradition there is a fixed text which is transmitted unchanged from one generation to another.

The quarrel about the errors and inconsistencies in the Homeric poems, inherited from the seventeenth century, has continued steadily until our own times. In its narrower aspects the strife has resulted in a stalemate; the real inconsistencies still remain unexplained in spite of the ingenuity of the Unitarians, who have to their credit the checking of the excesses of the Separatists. [14] The picture of the great Homer nodding has looked more like an excuse than an explanation. Bowra's remark in Tradition and Design in the Iliad that the oral poet concentrates on one passage at a time is closer to the truth. We have been gaining much perspective on the inconsistencies, despite the hard core of apologist opinion. Of far greater importance than the labeling of the inconsistencies themselves has been the theory of multiple authorship which emerged when the quarrel over the "errors" began to find added fuel from the attention paid to still living bards and bardic traditions and to medieval minstrelsy. The many theories of multiple authorship of the Homeric poems have contributed more to Homeric scholarship than any other single concept. Not that they were right. But they led in productive directions. They were honest attempts to meet the challenge offered by the growing body of oral ballads and epics as well as of medieval epic.

The theories of multiple authorship can be divided into two general classes. The first, and the earliest, saw the Homeric poems as compilations of shorter songs, stitched together by their compiler. D'Aubignac presented this in his Conjectures in 1715, but it was Lachmann and his followers in the nineteenth century who made serious attempts to dissect the poems according to this lieder theorie. The attempts were unsuccessful and unconvincing; for the dissectors could not agree on where to use the scalpel. The theory was discredited.

A second general approach moved in vertical rather than horizontal lines. The scholars who used this approach abandoned the idea of a compilation, even of poems of different times and places, and conceived of an original kernel which was modified by a succession of later authors. Usually to them Homer was not a compiler but the last and greatest of the redactors. They too whisked out the scalpel and began to peel off the layers in the Homeric poems. Linguistic and dialect evidence came to their assistance. But they were equally unsuccessful and their theories too have been discarded, although with considerably less finality than those of the first group.

The work of the first group of dissectors led to several valuable compilations. Lönnrot put together the Finnish Kalevala, the Estonians entered the competition with the Kalevipoeg, and the Serbs attempted a number of "national" epics on the Kosovo theme. But nothing comparable to the Homeric poems was produced. The problem of the way in which the {10|11} Homeric poems had attained their length, if they were not literary productions of a single author, remained unsolved.

So also did the problem of the variety of dialect and archaic forms in the poems. One of the attempts to solve this problem was the theory of a special poetic language, a kind of artificial dialect which was the property of epic poets. This of course did not solve anything. It merely put a label on the diction as found in the poems and pushed into the background the question of how such a diction could have been formed.

Hand in hand with the theory of multiple authorship went the emphasis on a search for the archetype. Leaf's work is typical of this trend. His five strata began with an original and then discerned expansions and interpolations of later periods. One still hears echoes of this kind of dissection, for example, in MacKay's The Wrath of Homer.

The work of all these theorists should not be dismissed as without avail, certainly not with the tired yet vituperative cynicism of Allen in his Origins and Transmissions. The service of these scholars has been in essence to point out the peculiarities of language and structure of the Homeric poems, peculiarities that we now recognize to be those of oral poetry. The inconsistencies, the mixture of dialects, the archaisms, the repetitions and epic "tags," and even the manner of composition by addition and expansion of themes have been noted and catalogued by these scholars. The questioned existence of writing led them to use the word "oral" and their experience of folk epic seemed further to justify this term. The elements that were needed to crystallize the answers to their questions were there. It is a strange phenomenon in intellectual history as well as in scholarship that the great minds herein represented, minds which could formulate the most ingenious speculation, failed to realize that there might be some other way of composing a poem than that known to their own experience. They knew and spoke often of folk ballad and epic, they were aware of variants in these genres, yet they could see only two ways in which those variants could come into being: by lapse of memory or by willful change. This seemed so obvious, so much an unquestioned basic assumption, that they never thought to investigate exactly how a traditional poetry operated. They always thought in terms of a fixed text or a fixed group of texts to which a poet did something for a reason within his own artistic or intellectual self. They could not conceive of a poet composing a line in a certain way because of necessity or because of the demands of his traditional art.

I believe that the greatest moment in recent Homeric scholarship was expressed by Milman Parry when he wrote his field notes for his collection of South Slavic texts and spoke of his growing realization that what he had been calling traditional was in fact oral:

The real impact of this revelation of Milman Parry has not yet been fully felt in Homeric scholarship, which has chosen to disdain oral epic and to move into the more abstruse kinds of literary criticism. [16] Although often referred to, the oral theory of Milman Parry is at best but vaguely understood. It is the purpose of this book to present that theory as fully and yet as simply as possible. Parry himself did not live long enough after making his monumental collection to think out his theory in detail, let alone to develop it and present it to the learned world in completeness. Working from the clues that he left, I have tried to build an edifice of which he might approve. {12|}