The Text and the Voice Author(s): Paul Zumthor and Marilyn C.

Engelhardt Source: New Literary History, Vol. 16, No. 1, Oral and Written Traditions in the Middle Ages (Autumn, 1984), pp. 67-92 Published by: The Johns Hopkins University Press Stable URL: http://www.jstor.org/stable/468776 . Accessed: 12/03/2011 17:29
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=jhup. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to New Literary History.

http://www.jstor.org

The Text and the Voice

Paul Zumthor
SEPTEMBER 1982 at an international convention dealing with the medieval epic, I heard one of my colleagues declare in plain language that "today, the question of orality in the chanson de is a dead issue." geste No doubt it is. Let us take the chanson de geste as an example. The disputes of the 1950s-the best record of which is contained in that issue of the Actes for the 1957 Liege Colloquiuml-resulted only in the weakening of a certain confidence, a narrowing down of the scope of several terms, and the circulation of a small number of collectively shared doubts. As to matters of certainty, there were none. But the question is not one that deals with certainties. It concerns a mode of perceiving and, perhaps more, a willingness to be openminded in approach (which involves an integration) when reading our old texts, and a sort of critical imaginativeness. It is a question of epistemology. From this standpoint, the chanson de geste as such is of little importance-even though it undoubtedly constitutes a field worthy of investigation as a very consequence of all the pitfalls that are strewn along its path! The phenomenon which warrants our attention is a general one, well within the embodiments of any given genre: the phenomenon of the human voice as a dimension of the poetic text-a dimension that is socioculturally determined, and the measure of which calls for consideration of the history, not only of the text, but also our own-and therefore not possessing an inscrutable otherness. I admit here that, apart from some exceptions, every medieval "literary" text, whatever its mode of composition and transmission, was designed to be communicated aloud to the individuals who constituted its audience. Several settings for this communication obviously must be distinguished, from improvisation before an assembled crowd to reading aloud in a private circle. The question of "orality" in the chansons de geste or in any other poetic genre can therefore be raised only in terms of performance,not of origin. At least we may suppose that performance-which gives due recognition to the general rule. This the qualities of the human voice-represented is not an assertion on my part that has no evidence to support it. as well as Examples are numerous and of every sort-iconographic

IN

68

NEW LITERARY HISTORY

literary. We may be assured that even reading by oneself would normally include an articulation of the sounds being decoded. If there has been an almost universal falling off in discussions on the oral nature (or orality) of various medieval traditions, it is due less to an inability to come to any conclusion about the facts than to an error in perspective or approach. In the latest bibliographies published by the Societe Rencesvals, not only do I find that merely three to four percent of the titles relate to this (non?)-issue, but virtually all of the authors cited treat orality as both a technique and a sufficient explanation. Apart from infrequent references to works like those of Jack Goody and Walter Ong, the question concerning the nature or characteristic function of orality never comes up, no more than it does for the "Middle Ages" as a sounding place for the human voice. Through the intermediary of those who, around 1930-40, were the great authorities for my generation, we continue to rely upon scientific positivism on this point. Out of this positivist thinking we have, it is true, acquired a concern for rigor, something that I should hesitate to bring under attack; but in doing this we have lost the intuitive sense which in a perhaps naive but earnest fashion guided Romanticism in its discovery of the Middle Ages: that feeling of "internal subjectivity" (innere Subjektivitdt)which Hegel made central to medieval Christian "chivalry"2-a sense of the inner self that is aroused by face-to-face communication and idealized in what Novalis called "communion of the word," as against the innumerable interferences which for us subjugate all dialogue. Now, in our case quite obviously, it is a matter of dialogue with ancient texts that are masked walk-on character parts in a drama at which we are no longer spectators, and carriers of a discourse that we no longer hear. A whole entity has been shattered and, as such, is beyond recovery and beyond even the imagination. There still exists, however, the possibility of defining and shedding some light upon some of the areas of intersection where distant fields of view converge and, in their axis, a space figuratively reforms itself. I, for my part, intend not so much to claim once again, as on other occasions, the importance of orality in the transmission and indeed in the creation of medieval poetry, but rather to try to appreciate and gauge what this orality implies; not so much to evaluate the size of the "oral part" in the corpus of extant texts as to integrate into my perception and my reading the properties thus ex-plained. The problem of orality in what is incorrectly called medieval "literature" seems to me less a question of fact (supposing reconstitution and evidence) but more one of interpretation (with a view to overcoming a reciprocal alterity by unravelling it).

THE TEXT AND THE VOICE

69

Here it is not so much a question of orality as of vocality. I would borrow what Henri Meschonnic has proposed: sound cannot be equated with the voice.3 Let us agree then that "orality" is the historical authenticity of a voice. It is not a matter of thus starting up once again the debate on writing, initiated fifteen years ago by Jacques Derrida, and digging up terminology from that area to use for the context of the voice in performance. There is no doubt that poetic voice carries the imprint of some "arche-writing," but this imprinted trace is inscribed there in a specific manner, since voiced discourse given aloud has its roots more clearly in the human body and in other narrowly defined areas and lends itself better to the inflections of memory. Certainly a long-established school of thought esteems and sets great importance, to our minds, on the human voice inasmuch as it is the vehicle for language, and in it and through it are articulated the sounds which carry meaning; on this score we could designate it in effect as the "disguised version of a first writing."4 However, what should strike us as more important is the broader function of the voice. The spoken word is, in our culture, the most obvious manifestation of this function, but not the only one and perhaps not the most vital-I mean here the exercise of its physical power, its capacity to produce phonic elements and organize the substance of these sounds. This phone does not take on meaning in any immediate fashion; it only prepares for it the setting where it will be realized.5 No medieval discourse is known to us except through texts. That fact has far-reaching consequences. What was originally the final version of one such text that I came to study still does not seem clear to me in spite of all the information I gleaned from paleography, bibliology, or history. I can distinguish without difficulty between a luxury manuscript, individualized by its dedication, and a quite ordinary minstrel manuscript; between a copy used for teaching and one which a public bookseller could have sold at a profit in an urban setting. But these are circumstantial elements that are generally not enough to define the function fulfilled by the scribe in tracing those symbols onto the paper. Of the two primary functions of setting down in written form-communication one preand preservation-did dominate and perhaps alone determine the whole operation? All this is conceivable, and you come across only individual cases. In that sense, too, each text has its own history. Now, it is clear that depending on how the examination of such a manuscript or such a written tradition as a whole causes us to lean toward one answer or the other, we find ourselves confronted with quite different modalities of poetic discourse.

70

NEW LITERARY HISTORY

What does remain is that, in whatever state they have come down to us, these texts are texts; and nothing authorizes us to take them as simple recordings of spoken words-that is to say, to jump, by and as if it were of no great consequence, from the mode metaphor of our sensory perception of it to another mode. These texts I read and, if I am editing them, I print them so that in this form I may offer them for annotation by scholars, teaching by professors, and for reading by all those who are interested: a work of cultural retrieval has thus been accomplished. I make no claim that this work is in and of itself worthy of condemnation, nor moreover that it has any merit. The only thing that is important in this instance is that it puts into play for us, in our (dare I say it?) "consumption" of the text, those same physical and intellectual faculties we use in the consumption of contemporary writings: our visual sense, but in addition all that our own reading habits imply-in body position and kinesthetic impressions, as well as in the mechanisms of reception and memory processes-or even down to the shape and perhaps the softness of the chair we're sitting in! Even if the twelfth- or the thirteenth-century man was himself a reader, did he read "like us"? Most certainly, no. In this sense "medieval texts" present us with nothing but an empty form that is without a doubt profoundly distorted from what was, in another sensorimotor context, the whole potential of the spoken word. All questions regarding the oral quality of poetry of that period remain subordinate to this general fact. This is true to such an extent that one could speak of the very techniques which today permit us to hear certain medieval texts (such as records of troubadour or Minwhatever their intrinsic quality may benesanger songs)-besides in about the same terms as we would for things we read. The sound transmitted by the record lessens to a great degree the effects of temporal distance and muffled sensory perception. However, it is only an illusion of being there; and if one goes on, as sometimes happens, to a live performance in a hall with a singer and musicians, it is upon nontextual aspects of the performance that the ambiguities of voice become centered: what we listen to from our seats, with (I trust) great pleasure, relates in only an artificial manner to the medieval past. This concert is a present-day performance. From the point of view of our own culture, this is an advantage; and perhaps it is the most efficacious means of breathing a little life into these venerable documents of history. Our understanding of these documents is enhanced through such performances-but no doubt at the same time becomes somewhat distorted. Standing as evidence, among other things, are the quarrels between different schools of thought regarding the interpretation of melodies.

THE TEXT AND THE VOICE

71

During the life of the text, whatever the moment may have been (upon composition, transmittal, or indeed both) when vocality intervened and became operative, imparting to a "work" its concrete reality, the position of the modern reader remains the same. He can, implicitly or explicitly, talk about this text; but he has no means of explicating it. By this I mean grasping and making its sensory implications perceivable to us. He shares this inability with all readers of ancient texts-eminently texts which, during a part of their life span, went back to an oral tradition. Biblical exegetes were the first to become aware of this during the 1920s. The medievalist, who is in a similar position, could benefit by reflecting upon their experience. It is a question here ideally of abandoning (in some fashion that will not cause too much dissension) the qualitative changes formerly brought about by the act of having set something down in written form, of setting up at least a plausible bridge above the discontinuities whose importance we assume without being able to quantify it. We will do nothing more than try to reconstruct the circumstances-not by simple, scholarly accumulation, but because we will have posited straightaway that the text verbalizesa situation that is special, a Sitz-imLeben which is not only definable in general sociohistorical terms but which individualizes it, just as much in the realm of physical perceptions as in that of the intellect.6 I take down from my library an edition of the Song of Roland. I know (or assume) that in the twelfth century this poem was sung to a tune that, for all intents and purposes, I have no means of reproducing. I read it. What I have before my eyes, printed or (in other situations) handwritten, is only a scrap of the past, immobilized in a space that is reduced to the page or the book. I come up against a twofold problem here. On the one hand, there is the discontinuity arising from the historical set of my critical concepts and their presuppositions, which project my own cultural identity onto a different object. On the other hand (for a text over which hangs some presumption of orality), there is my lack of knowledge about the auditory mode of articulation, as opposed to the visual, for a civilization with a predominantly strong oral characteristic.7 Only practical experience allows us, if not to resolve them, at least to shed some light empirically on these contradictions. Through an intertwining of different spheres of knowledge and by shifting our perspective and aims and starting from a point of view that is intuitively selected, we will be to perform the forced to come up with an event-a text-event-and text-in-action, and integrate this representation with the pleasure that we experience in reading-and take this into account, if the need arises, in our study of the text. What we spontaneously seek in all works of art, and in particular

72

NEW LITERARY HISTORY

the poetic works, is a mirror effect. If the work belongs to the past, this reaction-which for us is hard to control-gets stronger, especially since the elements of anecdotal background information are not at our disposal. The further back we go into the past, the more forcefully are we thus swept along, up to an optimal threshold and before too great a spatiotemporal distance wipes out all sense of belonging. There is no doubt that the Middle Ages are for us today at that threshold-hence their "topical interest," just as much for those in research as for the general public. Such is the desire that shapes, to its very core, the discussions that we hold on this subject matter. A desire that is often veiled, and for that all the more vigorous, producing fantasies in criticism and mythologies that tend to popularize the subject matter and which are sometimes ossified in a methodology aimed at integrated unity. It is on account of this fact no doubt that we have seen the poststructuralist "a-topia" which has led to a multiplicity of terms with the prefix de: decoding, deconstruct. That u-topia.... All this, if I might dare to say, should be debunked!-not rejected, but put to use for a redefinition, not so much of history as such (who would risk that?) but of the event. Admitting to the orality of a text is, paradoxically, to become aware in an active fashion of a historical fact. This fact cannot be confused for the setting which we still have the written symbol for, nor will it ever appear in the mirror. For us then it is a question of exploiting this desire for an intrinsic, recognizable likeness in our discourse, as students of history, and of capturing its intensity in order to use it for our own ends: to try to see the other side of the mirror, or at least to scratch away the silvering a little. There behind it, beyond the evidence of our present muland the rationalities of our methods, lies that remnant-that tiple, without a unifying origin or a globalizing end, that "noise" referred to by Michel Serres, any understanding of which resides in our sense of hearing rather than sight.8 It is there, and only there, that we will find the orality of our "medieval literature"-a remnantvocality of our philologies, reluctant to yield to our systems of conceptualization. We shall only evoke it figuratively, that is to say, in what I would call narrative terms-giving to that word its most dynamic sense, referring to the narrative as a receptacle for analogy. My intention here is not to advocate an anecdotal, comparative approach, even though it can sometimes shed light on the field of research: someone or other amongst us will attempt to describe medieval orality by reference to what we can see today, or could have seen in a still recent period of the past, in the setting of an oral culture. This is what in 1955 Jean

THE TEXT AND THE VOICE

73

Rychner did in an outstanding book about the chanson de geste. The problem is one of shifting emphasis: How do we take from ethnological studies information that can be used in the history of texts? We lack the general reference term that might be something like a "poetics of the voice."9 By analogy I am here referring rather to all figurative representations allowing us to perceive and record the signals emitted by history but which have been muffled or suppressed, sometimes by the society of the past, sometimes by our own. It is up to the historian and the critic to recapture these signals, even if starting from a specious similitude, but passing through this stage and then distancing ourselves to the point of forgetting it as the original motivating forces are revealed. The progressive discovery in Germany, England, and France during the eighteenth century of a "popular poetry" thus permitted future medieval studies to take shape.10 There is no question of establishing a "science" based on this, but a probable opinion; less a question of demonstrating (yet another term with the de prefix!) than of revealing, that is to say, of making present to the senses, of causing us on a cognitive plane to hear and feel. What I am seeking to evoke in this way would be a "poetic" discourse, homogeneous with its subject matter, as opposed to an approved master language that is external and reductionist: a language that is both ever-changing and yet producing its own adherence to truth: a gesture.11 Just as important as mastery of the techniques of philology and textual analysis, the ideal task of the medievalist would be to convince him- or herself of the incomparable properties of the human voice; to develop a sensitivity to them; or better still, to live them, for they only exist, so to speak, at the point of utterance and independently of the concepts in which, of necessity, they are bound up in order to describe them. This requirement seems to me fundamental, from the very fact that what we can say about medieval oral activity constitutes the point of departure and arrival for what we will say about medieval writing, and not the other way round. On this aspect, I must be quite emphatic. The point of departure for a medieval poetics ought to be the consideration of that inner beauty in the human voice, "taken closest to its source," as Paul Valery wrote.'2 This beauty can, it is true, be conceived of as being particular or special to the individual who gives it vocalization; for this reason, aside from some nearly unimaginable exception to the rule, it remains impossible for us to grasp it across so many centuries. But it is conceivable also as a social and historical beauty in that it unites human beings and, by the use we

74

NEW LITERARY HISTORY

make of it, shapes a shared culture; and that aspect obviously falls within the scope of this present study. Prior to any phonological differentiation (an inexpressibility that is able to be represented through language), the human voice is a material entity. We can describe its properties in quantitative as well as timbre, loudness, pitch; and to each one of qualitative terms-tone, these properties custom attaches, in the majority of societies and groups, if not a sign function, some symbolic connotation. Out of the multitude of natural sounds, man hears his own voice emerging, like an object rising up from the watery deep. Around this object the social link is formed and a poetry begins to take shape. From this arises the backdrop, filled with entangled potential meanings, which distinguishes the voice from the expression in our eyes. The latter itself is another physical manifestation (no less than the structure of the performance) which many myths associate with the voice. In fact, whatever the expressive power of a look may be, the register of visually perceived things lacks that fullness of the voice, its concreteness, the sensual tactile quality of a puff of air and the vibration of breathing. This capacity of speech is not found in a look-the capacity to continuously stimulate the game of desire for something that is absent, but yet nevertheless present, in the sound of the words. That is why, just as language is unthinkable without the human voice, so too the voice goes far beyond the spoken word. The voice does not carry language: it is the latter which passes through it in transit and leaves no trace. Perhaps, in the deepest recesses of our mind, the voice is assigned the task of serving a protective function: it safeguards a subject matter that its language threatens; it checks the loss of substance which a perfect act of communication would entail. The voice expresses itself by the very means of what it says; in and of itself, it is a pure imperative. Its use procures a sense of pleasure which, at every moment, it aspires to actualize in the linguistic flood which creates interference, at the same time as the voice is producing this flow of words. In the voice, in effect, the spoken word is uttered as a reminder, like memory enacting some primordial contact, the blurred trace of which, however, remains with us in the image of a promise. Raised above that rift "between the transparency of the abyss and the flatness of words," as it was expressed in a beautiful essay by D. Vasse, the voice suggests a "boundless resonance over and above itself."'3 What the voice delivers up to us, prior to and within the spoken word as it is transmitted, is a question about our beginnings-about that point, external to space or time, when the sexes, generations, love and hate were as one. Each syllable acquires a rhythm from the pulsing of

THE TEXT AND THE VOICE

75

blood in our veins, and the energy arising from this converts the question into a declaration, memory into a prophecy, and covers up the tracks the loss of which irreparably affects our language and our time. One paradox of the voice is that it constitutes an event in the world of sounds, just as every movement of the body is an event in the visual and tactile world. Nevertheless, it somehow escapes our full sensory comprehension: in the world of material things, it presents us with a strange incongruity. That is why, in myths no less than in our customs, it turns out to be accompanied by laughter, which is yet another force. The legend of Merlin carried on, well into the thirteenth century in the West, the ancient motif of prophetic laughter: the voice, as revelation, rises out of and is inseparable from a burst of laughter. Uttering the spoken word thereby takes on within itself the value of a symbolic act: by reason of the voice, it is exhibition and gift, aggression, conquest, and hope for victory over its adversary; manifest internalization overcome by the necessity to physically invade the object of its desire: the vocalized sound goes from the inside out and, without any other mediation, links together two lives. The oral poem, therefore, unlike the written poem, cannot become an end in itself. The "enclosure" of the text (in the way we speak of the enclosure of a garden, hortusconclusus,hortusdeliciarum)breaks apart, and through the breach comes something else that is no longer discourse and which crosses over the boundary lines of ordinary language. The referential force of the poem derives largely from its focusing upon the contact between the people bodily and together present at the performance. A trend of thought running through the Middle Ages attests to the fact that people in that period had some appreciation of these qualities. Of course there are very few texts that are explicit on this point. But could it not be that their rarity merely signified that this went without saying and that the ontological implications for the human voice, within the sensibilities of the eleventh, twelfth, and even thirteenth centuries, were perfectly clear from the examples in evidence? The practice of poetic art in those times would have us believe so. In the realm of philosophy I would cite the commentary by Thomas Aquinas on Aristotle's De anima, Lesson 18. Toward the end are the short lines which later, during the fifteenth and sixteenth centuries, would be explicated by the theologian-linguists speculating on the Verb.14 But perhaps too we should bring up as evidence the use within the scholastic tradition of the Latin word vox. It is used to designate not only the materiality of the sound released through the

76

NEW LITERARY HISTORY

throat, and the linguistic sign for which it is the vehicle, but also the meaning that it gives rise to-a word usage that is even attested to in the common speech of the thirteenth century, as in the Rhetorique of Brunetto Latini.15 The inherent properties of the human voice must be a determining force, at least in part, on the formation of poetic messages destined for oral transmission. The texts which they make up thereby acquire a status, however unambiguous it might be, with regard to writing. Thus of all poetry it is true to say, along with Meschonnic, that the human voice must enter into its definition-that the realm of poetry is the realm of the voice.16 For poetry, to which no outlet is offered other than oral performance, this proposition can be applied in a specific way: it heralds and demonstrates the historical realization of a universal virtuality. Another trait is that the written text carries a twofold effect of deferred communication: the one intrinsic, due to self-reflexive activity and polyvalences engendered by poetic formalization, the other extrinsic (but in the long run of great importance), created by temporal and contextual distancing between the moment when the message is emitted and when it is received. The poem that is performed orally displays the first effect but, in principle, not the second; being oral, it is based upon a fiction, at least, of immediacy. In fact, even if the oral reading of the poem takes place long after its composition, as a performance it can only be of an immediate nature. It often ends up (and we can see this empirically) that the effect of intrinsic deferral will weaken upon performance-or perhaps the effect has been lessened since its beginning, with the very prospect of an oral fulfillment having been inscribed within the text by this deferral. This essential need for immediacy is of some consequence: the text transmitted orally must be considered as being fragmentary by nature. Certainly, several of today's critics have so described the modern literary text, by virtue of the noncompletion of a Writing which courses through the text but does not stop within it, and the tension which is created between this endless shifting and the limits of the discourse. These characteristics can be found in the oral text, which, as an organized linguistic sequence, cannot fundamentally differ from the written. But the linguistic plane is only one of its aspects of realization, and it is from the combination of these various levels that a particular fragmentary property arises. The tension out of which the oral poem is formed in effect takes shape between the spoken word and the human voice, and proceeds from a quasi contradiction between their respective finalities: between the finite properties of

THE TEXT AND THE VOICE

77

the discourse structures and the infiniteness of memory; between the abstract nature of language and the spatial world of the body. That is why the oral text is never saturated, never quite fills up its semantic space. It is this that provides us, the present-day readers, with a permanent source of errors of interpretation, for the question that we are inclined to address to the text, with a certain amount of impatience, is often to no avail because of its very urgency and its allencompassing intent. Besides, at the heart of the tradition which it cannot help but be referred to, oral poetic performance stands out like a discontinuity in the continuous-"historical" fragmentation whose effect seems all the more apparent as the tradition gets older and more explicit and embraces elements that are better diversified. So, in the economy of the cycle literatures (the legend, epic, tale, song cycles), we find virtual superunities whose property is to never be realized as a whole-a vast treasure house which the narrator, the singer, or the oral performer seems to draw from according to his own desires at each performance. The form of the oral text thus remains on a desidereal planewhat Max Luthi, in reference to tales, calls Zielform, a finalizing and ideal form which here and there is realized, but not across the entire work, and which, as such, cannot constitute a complete whole. Vast areas of medieval poetry (even those for which an oral form of transmittal is left in little doubt, such as the chanson de geste) illustrate by the way they function the fruitfulness of this notion.17 Perhaps, in this connection, I should restate what I wrote three years ago concerning the facts about medieval intertextuality: "Medieval 'literature' from this viewpoint appears as if it is made up of a tangled intertwining of texts, each one of which barely lays claim to its own autonomy. Fuzzy contours encircle it imperfectly and the lines of communication from one part of this network to another are never cut
off."18

It is within the framework encompassed by these general characteristics that the following question arises: Are there any specific oral forms? Whatever, from another point of view, the circumstances of composition might have been, does the oral rendition of a text predetermine the way it will be formalized? I do not plan to furnish a ready-made solution to this problem. At best, I would hope that what might emerge is an angle of approach that will be a likely, but necessarily problematic, one. The text that is destined for oral transmission in fact is, by its very nature, less accessible than the text that is meant to be silently read. Even more than that, the text resists being

78

NEW LITERARY HISTORY

identified with the words of its author; even more than that, it leans toward setting itself up as common property of the group at whose center it functions. From this fact come two closely interrelated characteristics: (1) The "model" of oral texts is more distinctly concrete than that of written texts; the stock discourse fragments for which it serves as the vehicle are more numerous, better organized, and semantically more stable-for example, the stock lines of chansons de geste or certain "lyric" motifs such as the locus eroticus.19(2) Within the same text during the course of its presentation, and from one text to another (at the synchronic and diachronic levels), we see instances of interference, recurrences, repetitions that are probably allusory in nature: exchanges of discourse that create the impression of a whirl of moving textual elements, at every moment linking up with others in provisional compositions-for example, the whole body of texts formerly published by Karl Bartsch under the title Romances et pastourellesfranfaises des XIIe et XIIIe siecles (Leipzig, 1870), which I dealt with in chapter six of my Essai de poetiquemedievale(Paris, 1972). The above comment would seem to take us back to the famous "oral theory" (an expression currently favored in English- and German-speaking countries) which came out of the research done by Milman Parry and the work of Albert B. Lord.20 This reference, in my scheme of things, is only indirect. The Parry-Lord "doctrine," after a quarter of a century of use, if we reduce it to its basic principle, it seems to me in our studies, can be thought of rather as a starting in heuristic terms but lacking in any universal, hypothesis-useful authoritative weight. Initially being limited to descriptions of epic poems (from the Iliad to Serbian heroic chants), the theory defined a fairly complex means of expression given the name of "formulaic style." From the beginning of the 1970s, a growing number of authors came to consider this as the distinctive and defining characteristic of all oral poetry. Rather than as a type of organization, formulaic style can be described as a narrative strategy: it inserts within the discourse, as it unfolds, lexical and syntactic rhythm sequences borrowed from other preexisting kinds of expression, thus referring the audience to a familiar semantic world. These "formulas" crop up in greater or lesser numbers depending on the period, the poets, and the circumstances. Attempts have been made to measure their density of occurrence in European epics of the Middle Ages; the norm for the oldest French chansons de geste would come to thirty to forty formulaic lines out of a hundred.21 The structure of the line and the constraints it engenders must play a role here.

THE TEXT AND THE VOICE

79

Each poem as such constitutes a unity of original spoken language governed by such laws that are specific to it; the formulas are set within the poem as weakly marked terms, acquiring, by and in this insertion, their function and their meaning. Several definitions of the formula, it is true, are complementary while in part contradicting one another. Everyone is in agreement on the banal fact that here we are dealing with a textual scheme that is endlessly reusable. But the interpretation of this particular fact differs according to the breadth of the theoretical approach. For Hermann Bausinger, all cultural practices can be formularized, the formula constituting a Kulturgestalt,a formalizing dynamic force which is collectively shared by the group. We can distinguish two types, each with a certain number of variables: "functional" formulas that exercise constraints on the discourse, and "play" formulas in which, against these constraints, the tension of interplay is dialectically manifested.22 The poem works up a system of formulas linked to one another by rather complex relationships of equivalence, complementarity, and opposition. The manner in which the poet masters and exploits this system constitutes one of the criteria by which his art is judged at the time of performance. With Lord, his students, and his detractors, the definition has progressively become more flexible and more concise on a few points. People no longer lay so much stress on the lexical sequencing as they do on structural factors such as prosody, syntax, or the distribution of key forms. A good example of such an interpretation is furnished by C. W. Aspland's book on the participial -ant assonance in the chansons de geste.23 W. Nagler has systematized this conception of oral poetic style by proposing a generative cescription of the formula. For him, this term goes back to a set of phonetic, syntactic, lexical, rhythmic, and semantic relationships that, in the mind of the oral poet, form the model immediately underlying all formulaic occurrences. Each formula functions like an allomorph, not of some other formula, but of the model; all the occurring formulas weave within the text a solid but flexible network, and between the strands of this network circulates a meaning. The model functions in fact on its various levels (sounds, words, and so forth) by means of a twofold cog system: internally, on the formulaic phrase; externally, on the entire text. As a result of this, in principle, every formula is at a crossroads of multiple relational perspectives. The formulas exist within a tradition and cannot be dissociated from it. The collective tradition retains a greater or lesser number of these formulas, which are available to every poet well versed in his art. It happens that some of the formulas, like dialectal traits, at the heart of the linguistic field of consideration have only a limited range

80

NEW LITERARY

HISTORY

of diffusion; that does not in any way change the functioning of the favored by a particular poet-arise system. Other formulas-those from a personal competence that is stable and repetitive. The study of medieval texts has suggested moreover a distinction between "internal" formulas, which appear only in a single poem, and "external" ones, common to several poems. It was during the 1950s that a number of medievalists almost simultaneously had the idea of applying to our narrative poems of the later Middle Ages the notion of an oral-formulaic style which seemed from then on to have a guaranteed place in the work of certain Hellenists and Slavists. From 1953, Francis Magoun thus set to work on Beowulf and the Anglo-Saxon tradition. His work was based on a strict and rigorous conception of the theory: the formula is a necessary and absolute proof of orality; its presence excludes the intervention of the written word, except simply as a record of performance. With Beowulf containing seventy-four percent formulaic lines, it was only a matter of drawing the conclusions from this fact. When it was reedited in 1963 by Lewis E. Nicholson, Magoun's article had already given rise to a whole line of progeny: studies by Ronald Waldron, Robert D. Stevick, and others on Old English poetry or on Middle English alliterative verse. One source that was often invoked gave perhaps, in the eyes of English literature specialists, a greater sense of urgency to this problem: in Book IV of the Historia ecclesiastica,the Venerable Bede in the eighth century described with wonderment the activities of the bard Caedmon and gave an illustration of his improvisations-which strongly confirms various of the more obscure allusions in Beowulf and in the Irish Egil saga.24 In 1955 appeared Jean Rychner's book on the chansons de geste, which he proved, with terms borrowed from Parry, to have been intended for an oral delivery. The impact of this was such that, two years later, there was justification for the meeting of a convention possibly designed, in the minds of certain individuals who organized it, to strike down a dangerous heresy in the nick of time. The plan failed. Up until the middle of the 1970s we saw the development of a fairly large body of references to studies of this kind, dealing for the most part with the cultural spheres of the Middle Ages, France, the Germanic and Scandinavian countries, and Spain in particular. The study of formulaic style during those years when it was in vogue had, it is true, ended up by developing into a quasi-autonomous discipline, to the detriment of other poetic elements in the texts under consideration. Often, for young researchers with little experience, it was reduced to a rather pathetic formula hunt. Lord him-

THE TEXT AND THE VOICE

81

self, however, has always taken pains to distinguish between the term formula and what he calls "theme" (in the sense, it seems to me, of what is more commonly referred to as "motif"). This distinction apparently refers to the two component levels of the poem, that of "narrative" (deep-structure narrative) and that of the "story" (manifested as surface-structure semantic components), but the distinction has never really been made explicit nor elaborated upon as much as it should. The theory's reputation has greatly, and in some measure justifiably, suffered from misunderstandings and hasty generalizations arising out of this lack of specificity. At the same time, some enthusiasts were aspiring to universalize the conclusions drawn from their work. With Lord at Harvard, but elsewhere too, people were extending their investigations to nonEuropean sectors of oral poetry: Hindu, Persian, and African poems ran each in turn through the mill; Michael Zwettler yet, in 1978, writing on ancient Arabic poetry, referred to the work of Lord and Nagler.25 Now, these excursions outside the theory's home territory often have the advantage (contrary to the intentions of those responsible for them) of relativizing the theory in a number of ways. Several monographs, by the very precision of their subject matter, have contributed to a redefining of the formulaic system by giving a picture of it that is so complex that any pure and simple application of the model becomes impossible. See, for example, the study by Aspland (mentioned previously) or Genette Ashby's thesis on the descriptive formulas of the single combat in the Roland, or in another vein, Edmond de Chasca's discussion of the nature and functioning of formulas in the Cantar de Mio Cid.26Along the way, especially among the Germanic specialists, thinking about the orality of certain medieval texts led critics to include in their body of works (as in a resurgence of Romanticism) modern forms of popular poetry for which we may assume a medieval origin. From this came a series of studies on Germanic, Anglo-Saxon, and Scandinavian ballads.27 Six years ago Ruth Finnegan took stock of the situation in carefully measured terms and called into question several proposals which, around 1970, were on their way to becoming axiomatic.28 In 1980 Adrian Fochi, in his book on "Esthetics of Orality"29 based on the study of Roumanian doine, suggested that formulaic style, far from being a mark of archaic usage within a poetic tradition, is more likely the relatively late result of an evolutionary process. It is now all over and done with; we get nothing more, and the criterion we have is watered down! We must now in fact realize that the use of formulas (1) is not found in every instance, nor throughout the epic; (2) characterizes even less the whole of oral poetry; and (3) is found on the

82

NEW LITERARY HISTORY

other hand in certain written texts that are designed to be read to oneself. This is why there is a common tendency to reject henceforth any view of formulaic style as being an unquestionable sign of orality. Quite understandably, Michael Curschmann was able in 1977 in Medievalia et Humanistica to publish an article with the provocative title, "The Concept of Oral Formula as Impediment to Our Understanding of Medieval Oral Poetry"!30 His demonstration had to do with the Niebelungenlied-an example, it is true, that is a special case, for few medieval texts present such an obvious combination of voice and written literary effects. The formulaic theory does not sufficiently take into account the internal needs of the poetic text. From the linguistic point of view, in oral or written form a text remains a text, within the scope of the critical methodologies for which as a text it is by definition the object of study. It carries necessarily the mark of this status. But any poetics mindful of doing justice to the vocal properties of the spoken word, and in order to maintain their specificity, should dwell less on these marks than on the unstable relationships. It is from these latter that, by concatenation of elements and their effects of meaning, results the particular economy of the text that is to be sung or spoken. This was what Menendez Pidal, around 1950, using somewhat obsolete terminology, discovered in the "traditional style" of Spain: its intensity and tendency to reduce the form of expression to its essentials (which means neither to its shortest nor to its most simple form); the predominance of the word in action over description; the interplay of echo and repetition; the immediacy of narrations whose complex forms are built up through accumulation; its impersonal nature; its timelessness.31 These more or less evident traits show at the poetic level the functional opposition that distinguishes the voice from writing. Since it continues to exist, the written text can fully assume its capacity in future time. The oral text cannot-being too narrowly subject to the present demands of performance; on the other hand, it enjoys the freedom of constant movement and constant variation in the number, nature, and intensity of its effects. It is through generalizations that the actual substance in all this is revealed. What the ear of the "oralist" seeks to distinguish in the continuous stream of the real are discourses rather than texts, messages in the making and not finalized statements, a pulse rather than stasis, or to borrow a term from Humboldt, energeiarather than ergon. This object or substance has to be ensnared; but first we have to invent the means of snaring it, and we are still only at the first fumbling stages. One point at least is certain: it is only by perceivingand analyzing-the oral work in its life as discourse that we will get a handle on its textual existence. It seems to me, moreover, that it is

THE TEXT AND THE VOICE

83

on this aspect that hinges the very position taken by a growing number of linguists and those interested in syntax when confronted with the problems of Old French.32 In relation to medieval works, what I would retain of the formulaic strictly theory is the manner in which it emphasizes-without speaking making it into a topic of discussion-and indirectly explains the force of all the vocal effects of recurrence, the way in which it allows us to catch a glimpse of the poetic characteristics. Yet we have to broaden our initial view of these characteristics and consider how on all levels poetic expression, from one movement to another, composes a theme of melodic harmony, how it weaves a network of connecting links which becomes more and more complex as the discourse proceeds, while, by reason of that very phenomenon, multiple meanings reduce in number and the personal stamp of the speaker or the variations on a stock theme, singer becomes more evident-replays, within sameness, and the laying down of a technique that is diversity everywhere the same and for which only the mechanisms of usage differ, to a greater or lesser extent, depending on the situation. Every recurrence fixes and maintains this quality; tending toward hyperbole, it stands as proof of the acceptance by the poet of the society for whom he speaks or sings. But he accepts that society less by choice than by virtue of the role conferred on him by the community as a whole, to be the curator and herald of poetic tradition. Such is the constant and universal characteristic of all oral poetry, the eminent functional mark of poetic vocality as such, a characteristic that may be seen equally well in the activity of a twelfth-century troubadour as in that of our rock singers. Undoubtedly, the effects of repetition through which one can perceive this trait are often found also in written poetry. Jakobson even saw in this the essence of all poetic language. Nevertheless, a close tie links these effects to how the human voice is used. At the deep-structure level where the epistemological components of the spoken word (as opposed to the written word) are created, all utterances are spontaneously recursive. They are made up of retrievals of a given expression that they amplify while further interpreting it, and this in such a way that the new element in the discourse goes back in part to this very glossing of the expression. The spoken word thus institutes a dialogue with its own theme. This tendency more or less polarizes every poetic genre that is intended for oral delivery. Rudolf Hirsch has noted that with the spread of written communication in the fifteenth century, this movement extended into all areas, from the style of royal proclamations to the prose versions of the chansons de geste.33 The rhythm resulting from these recurrences comes, in either a

84

NEW LITERARY HISTORY

simultaneous or alternating fashion, not only from sonorous echoes, repetitions of stanzas, entire lines, prosodic or syntagmatic groupings, grammatical forms, and words, but also from semantic effects. These repetitions either submit to the regularizing effect of a parallelism, which sets up their component parts in opposing pairs, or they rise above this numerical rule. They are either confined to some privileged placement positions or they overrun the text. They either keep picking up an identical theme or they implement a partial variation; they build up either in a rigorously sequential manner or according to various schemes of alternation. In spite of considerable differences in other respects, this is the general style typical of our chansons de geste, and also of two types of so-called lyric poetry, labeled "courtly" and "vernacular"-one could even go so far as to say typical of all types of expression in verse up to the thirteenth century. Threads thus weave themselves into the framework of the discourse and, having multiplied and crisscrossed, engender another discourse, working with the elements of the first one, as does a dream with fragments of our waking life, enhancing those fantasies to which it gives an identity. While the words are unfolding, equivalencies and contrasts become established. These (because the context changes, even if imperceptibly) entail subtle nuances, each one of which, received as new information, heightens the understanding toward which this voice beckons us. In this way, three modes of composition are established, each of which is widely attested to in the whole body of medieval literature, with the exception of "modern" narrative genres such as the novel: (1) litany: an indefinite repetition of an identical syntactic and in part lexical structure, with some of the words being altered upon each repetition so as to delineate a progression by a sort of sliding back and edging forward; (2) overlaying: the same sliding repetitions, not at the sentence level but of parts of the text, verses, couplets, and the "laisses of similar type" found in the chansonsde sections-hence pattern echoes: the text is dotted with repetitions at fixed geste; (3) intervals. These are sometimes intertwined and, by framing and sustaining the discourse, they lend a special strength to it. In this way the same element (a sound, word, grammatical form, or sememe) will appear at the beginning or the end of even-numbered lines, and another element on the odd-numbered ones; both of these can reappear in medial configurations. The system lends itself to an unlimited number of variations, as can be seen in virtually all the verse poetry of the thirteenth century. Other techniques are more closely linked to the melodic structure of the poem: (1) double or triple repeat lines; (2) a play on reiteration

THE TEXT AND THE VOICE

85

which makes a couplet from the expansion of one line or even a single word; and more generally (3) the use of the refrain, either constant or varying, distributed at regular intervals or not, serving to gloss the couplet or contrasting with it, and sometimes reduced to an evocative name, suspended in the extrasyntactic void, or to a pure singing exercise. Several poetic genres, attested to from the end of the twelfth century but probably with a much older origin, and which were designed to be performed in song, are based on these techniques: the virelay, balleto, and the round in all their various forms. In whatever way it is achieved, discursive recurrence is the most efficacious means of verbalizing a spatiotemporal experience and of bringing the audience to participate in it. Time unfolds in the fictional atemporality of the song, from the moment the first word is uttered. Then, within the space created by the sound, the image that is perceived by the senses becomes objectivized; a rhythm is born and a fragment of knowledge is legitimized. Concerning texts of the Russian Middle Ages, Dmitry Likhachev spoke of the "literary etiquette" or the highly ceremonial aspect of orally transmitted "literatures."34 He thus emphasized the underlying direction of all oral poetry toward a rigorous formalism. This is precisely what was evident in the minds of medieval intellectuals following in the tradition of Boethius when they related poetry primarily to musica, linking it only secondarily to grammar and rhetoric. In speaking of formalism, we therefore need to take it in the sense of where a form is only by exception stable and fixed, where it possesses a mobility which is its own distinctive characteristic, and where, in the extreme and paradoxically, form can equal force. Form therefore will not be conceived of as obeying a rule, since form is a rule, unceasingly recreated, existing only in and through the emotion aroused by each moment, each encounter, each quality of light-each performance. And every one of these forms that spring up is but one of the comin each poem (of which it is the ponents of a single Form-unique and unique in that it is never replicated and, as such, only form), escapes the passage of time, even though its component parts, on the contrary, have a tendency to reoccur indefinitely. Is it not precisely this aspect that is suggested by the author of the tornada, added by the chansonniers G and N to the song of Bernard de Ventadour Ges de chantar? "Lo vers, aissi com plus om l'au, / Vai melhuran tota via ... " ("The more you hear my song, the more its power grows as it goes along").35 But at what turn along the path in the life of texts (these textswhich are all that remains to us) do we have some chance of perceiving this form? Oral and written poetry (if we can judge them by

86

NEW LITERARY HISTORY

modern examples) make use of a language that is fundamentally the same: the same grammatical structures, the same syntactic rules, the same basic vocabulary. However, neither the patterns of usage nor the strategies of expression are the same. Orality seems to include, in this regard, the very tendencies whose effect we see in almost all of the oldest medieval texts and which survive in certain genres up to the fourteenth or fifteenth century. They culminate in that supremacy of parataxis (which Auerbach has already noted in Mimesis) and various word branchings. The narrative (whether it be in the chanson de geste, the fabliau, a certain number of lays, or many of the historiographic narratives in verse) tends in this way to juxtapose elements in a two-dimensional space without subordinating them. The "lyric" account breaks up the discourse into brief statements and usually separates them with exclamations and imperative expressions in a discontinuous, cumulative series; in the extreme, verbs become eclipsed and nothing is left but a string of free noun-phrase elements. The vocabulary (sometimes grammatical words themselves), compared to what the then current usage could have been (or what is evident from documentary prose of the thirteenth century), seems to be treated in a similar fashion: by being restricted and condensed. Sometimes it happens that the meaning is obscured; it is impossible for us to know whether this was intentional or not. At least it is in accordance with a system of expression from which one can suppose that it was a system of "diction," in the sense that Meschonnic uses this term, or a rhetoric of the human voice. The poetic art of the troubadours here furnishes us with a particularly convincing example. The majority of these processes entail some phonetic rule in their composition. The manipulation of any linguistic element helps to elicit or reinforce rhyme (whose widespread general use from early antiquity on no doubt comes from the extended role played by the voice in poetic communication), alliteration, and all kinds of sonorant echoes; or more generally, it accentuates the scansion of rhythms. When they reach a certain density, these processes influence the production of meaning. At the very most, the sentence and the words themselves become blurred into pure sonorous suggestions of cumulative figures of sound (as found in many refrains), or into a genre such as the fatrasie. Then one is confronted with (1) an absurd text, made up of juxtaposed syntagmemes that bear no grammatical or semantic relationship to one another; or (2) a phrase borrowed from another language which is not understood and therefore much changed; or (3) a litany of accumulated words in isolation with no context; or (4) a series of proper names, in apostrophe, and external

THE TEXT AND THE VOICE

87

to the sentence; or (5) ambiguous monosyllables that one can interpret both as meaningful words and as interjections, onomatopoeia, or exclamations; or (6) a string of sounds having lost all relationship to the linguistic code; or finally, in another way, (7) the effects of bilingualism. I pointed out their significance twenty years ago.36 Today I would interpret these effects from the viewpoint that a whole body of poetic work, in keeping with its oral nature, seems to be seeking to free itself from the semantic constraints of natural language, to escape from language itself toward a fullness which would be nothing more than pure presence. Perhaps the outbursts of this desire, characteristic of texts for oral presentation, are amplified by the position that, up to the fourteenth century, this desire occupied in the collective memory: not isolated as a "literature," not separate from the action, but functionalized as play just as in the play of body movements, an activity in which this desire truly participates. Is it not for this reason that, like all play, it brings pleasure stemming from repetition and true-to-life resemblance? But simultaneously, in spite of what we sometimes believe, poetic language caught up in this activity tends to make discourse structures extremely complicated. The Provencal trobar clus represents only a special case arising out of this tendency, just like the Kenningar of Icelandic Scaldic poetry. The kind of technique peculiar to the epic style of the chansonsde geste is another example. But this trait is evident also in so-called blunders and in the blanks or unexpected jumps in things we say; the much sought after complexity undoubtedly consists in linking with grandiloquence the linguistic, the vocal, and the gestural. One can consider that it is by means of these indirect processes that signs become inscribed within the short-lived tale or song, and that these signs transform it into a "monument," shielding it from the fate of common speech. Poetic structuring in the oral mode acts less through processes of grammaticalization than by the device of dramatizing the discourse. We can define the norm more easily in terms of dramatic art than in terms of linguistics. However, this intrusion of the physical in the grammatical seems to me to indicate another notable characteristic common to a great number of medieval works: the absence of unity, in a traditional, classical sense of the word. How many hypotheses relating to pieces of lost text, interpolations, and supposed reworkings arise simply from the repugnance felt by educated philologists for anything our texts display that is multifaceted, many-colored, and sometimes at variance or even contradictory? Is it not to this same misunderstanding that we owe three-quarters of what has been written about the irregularities of medieval versification, from El Cid to the Chanson

88

NEW LITERARY HISTORY

de Guillaume, and Anglo-Norman texts with so many lyric stanzaswhen anisosyllabism is probably nothing more than a quite ordinary characteristic of orality? All this wells up from the inside, from the initial summoning of a voice. Externally, we see that at best a great corpus of work is building up, in a fairly simplistic manner, in a series of diptychs that are dovetailed together, like the old Chanson d'Alexis, the Roland, Aspremont,and many others. Andre de Mandach contends that this type of composition is the most suitable for performances given from memory.37 I would base a grouping together of these various textual "states of existence" on the way in which narrative poetry makes use of every kind of process destined to integrate within the structure of the discourse the redundant features of its "phatic" function: prospective and retrospective digressions, justifications, static ornamental phrases, apostrophes, rhetorical questions, switching from "he" to "I," from "them" to "you," use of presentative terms such as "see!" or "listen!" to assure the attention of an audience, descriptive schematization, enumerations. From this we get an overall, artificial tension which allows the language to change direction with the demands of factual linearity. Such intermingling of registers reveals, from the point of view of performance, an effort aimed at producing a semantic surplus, at setting up a striking diversification at the very heart of poetic meaning. Conversely, "lyric" genres have developed mixed forms which impose a conventional order upon the discourse. The most widespread are those that I should call "pseudonarratives," typified best by the fine amor songs. Fundamentally, what constitutes the utterance here is the endlessly repeated expression of both a desire enthralled by its own fantasies and an intellectual response that denies their reality. The sometimes chaotic textual surface is summarily put together according to a latent narrative scheme: first visual encounter, the actual meeting, the request, a waiting period, abandonment or rejection; and each one of these terms serves as a mnemonic reference, external to the text, to one of the propositions set forth in the story. This pattern explicitly underlies Dante's Vita nuova. Performance puts forward a text which, during the time it exists, can include neither expunction nor change of heart: a lengthy period spent in writing it would have made sure it contained no rough spots in its oral form. Poetic art, for the poet, consists in holding this instantaneous quality and integrating it into the structure of his discourse. Hence the necessity for a special eloquence, a facility with diction and phrasing, the power of suggestion, and an overall predominance of rhythms. The listener follows the thread; there is no

THE TEXT AND THE VOICE

89

going back: the message must carry (whatever the desired effect may be) on the first go-round. Within the framework sketched out by such constraints, language gravitates toward a transparency, not so much in meaning but in its own nature as the spoken word, which is beyond all prescriptive rules pertaining to written composition. It is the voice and gestures that provide verity and coherence; these are the things that persuade us. The function of the linguistic form is to stylize the thrust of the spoken word without breaking it. From this arise illogical constructions, false starts, backtracking-and also, for the modern reader, the impression sometimes that the verbal aspect of an oral work is less carefully done than its prosodic or musical aspect, a point of view coming from people who are tied to the written word. The difference in sensorial registers which is brought into play by poetry intended for oral delivery on the one hand, and then poetry for reading on the other, implies quite obviously that their respective forms cannot be identical. Neither the levels on which they are formed nor the processes which produce them will even be comparable on an a priori basis. The vocalized text becomes an art form within an emotional setting that is brought to life in performance, from which and toward which stretch all the forces that constitute the living work. This is in part a qualitative setting, the operational ground for the "fantasy function," to use Gilbert Durand's expression. But it is also a topographically definable, concrete place where the spoken word, as it emerges, captures so fleeting a moment that it entrusts this very space with the task of organizing the discourse. There is no doubt in my mind that this is the principal reason for a striking and frequently noted characteristic of our medieval texts: their common ineptness at verbalizing descriptions of people and objects other than by a cumulative string of qualifiers with no wider perspective. With our paucity of information, which is reduced to what our eye sees on the page at that present moment, a paradoxical approach could, even to a very small degree, lighten the darkness. Such an approach would be to consider on principle every text earlier than the thirteenth century (or else, and much more problematically, earlier than the fourteenth century) as a dance. By this I mean (admittedly until we find evidence of distinctions of a different nature) that its actual functioning required the same endowment of qualities and brought into play the same powers of expression as did "dance songs" which are in other respects (and in this particular instance) better known: a text, a melody, and (through a rhythmic analysis or an iconographic study) movements. Examples of this abound, starting with the Chanson du roi Lothaire mentioned by Hildegaire de Meaux

90

NEW LITERARY HISTORY

.in the ninth century.38 What we know from these songs requires, in going to other texts, some more or less long-distance translation. At least the application remains on the order of metonymy, not metaphor. On this point I shall be adamant, since, to my mind, it is crucial. Thus with a certain verisimilitude, we can retain the essential "community-like" nature of the medieval poetic text. The Middle Ages had no dances other than those performed in groups, as figure, circle, and line dances, and the carol (or round dance) for processionsgesture and voice, in step with each other, cementing the unity of play and revealing a common design. The cohesive effect of rhythm can be heightened by clapping the hands or some other means of giving a strong beat. The singing parts, with or without musical accompaniment, are generally done by a soloist or a choir, the dancers responding with a ritornel. The text, determined by its function, is related to the gesture that it verbalizes: in its short form, reduced to an exclamation or a maxim; in a longer form, with strophic returns that lend themselves to emotive inflections. It is to this situation that we, as medievalists, come as interpreters. But here I should like to attach to this word those connotations which are linked to it when applied to the musician or the actor: the idea of producing something, of a special knowledge that is active and why not, too, the capable of bringing about a transformation-and idea of dance?
MONTREAL

(Translated by Marilyn C. Engelhardt)

NOTES 1 La techniquelitterairedes chansons de geste (Paris, 1959). 2 G. W. F. Hegel, Aesthetik,ed. F. Bassenge (Berlin and Weimar, n.d.), I, 530-48. 3 Henri Meschonnic, Critiquedu rythme(Verdier, 1982), p. 280. 4 Jacques Derrida, De la grammatologie(Paris, 1967), pp. 102-8; on this text, for the point of view of an ethnologist, see Dennis Tedlock, Phonographyand the Problem of Time in Oral Narrative Events, Documents du centre international de semiotique d'Urbino, 107 (Urbino, 1981), pp. 2-3. 5 See the excellent article Voce by C. Bologna in EnciclopediaEinaudi (Torino, 1980), Vol. IX. 6 Roger Lapointe, "Tradition and Language: The Import of Oral Expression," in Traditionand Theologyin the Old Testament,ed. Douglas A. Knight (Philadelphia, 1977), pp. 125-42. 7 Hans Ulrich Gumbrecht, "Metahistorische Vorgriffe Gattungageschichtlicher Forschung: Schriftlichkeit in miindlicher Kultur," in Archiiologieder literarischenKommunikation, ed. J. and A. Assman (Munich, 1983).

THE TEXT AND THE VOICE

91

8 Michel Serres, Genese (Paris, 1982), pp. 20-23 and 31-36. 9 My book Introductiona la poesie orale (Paris, 1983) sketches the first outline of such a poetics, on a more theoretical level than Ruth Finnegan, Oral Poetry (Oxford, 1977). 10 Bernard Mouralis, Les contre-litteratures (Paris, 1975), pp. 118-22. 11 Paul Zumthor, Parler du moyendge (Paris, 1980), pp. 100-102. 12 Cited in Meschonnic, Critiquedu rythme,p. 277; discussed pp. 277-96. 13 Denys Vasse, "L'arbre de la voix," Semiologiques,6 (1978), 129-34. De 14 Aristotle's anima, ed. K. Foster and S. Humfries (New Haven and London, 1954), p. 298; Claude Gilbert Dubois, Mythe et langage au seizieme siecle (Bordeaux, 1970), ch. 1. 15 Brunetto Latini, Li livres dou tresor, ed. Francis Carmody (Berkeley and Los Angeles, 1948), pp. 359-60; see also Walther von Wartburg, FranzosischesEtymologisches Worterbuch (Basle, 1967), VI, 638a. 16 Latini, pp. 279, 289. 17 Max Luthi, Das Volksmarchen (Dusseldorf, 1975). 18 Paul Zumthor, "Intertextualite et mouvance," Litterature,41 (1981), 15. 19 Pierre Bec, "L'acces au lieu erotique: motifs et exorde dans la lyrique popularisante du moyen age a nos jours," in Love and Marriage in the Twelfth Century,ed. Willy Van Hoeke and Andries Welkenhuysen (Leuven, 1981), pp. 250-99. 20 Summary presentation by Albert B. Lord, "Perspectives on Recent Work on Oral Literature," in Oral Literature, ed. Joseph J. Duggan (Edinburgh and London, 1975), pp. 16-21. 21 Joseph J. Duggan, The Song of Roland: Formulaic Style and Poetic Craft (Berkeley and Los Angeles, 1973), pp. 21-25. 22 Hermann Bausinger, Formen der Volkspoesie (Berlin, 1968), pp. 66-90. 23 C. W. Aspland, Epic Formulas and Formulaic ExpressionsContaining the -ant Form in 12th CenturyFrench Verse (St. Lucia, Australia, 1970). 24 Francis P. Magoun, Jr., "Oral-Formulaic Character of Anglo-Saxon Narrative Poetry," Speculum, 28 (1953), 446-75; taken up by Lewis E. Nicholson, ed., An Anthology of Beowulf Criticism (Notre Dame, 1963), pp. 189-222; Ronald A. Waldron, "Oral Formulaic Technique and Middle English Alliterative Poetry," Speculum, 32 (1957), 792-801; Robert D. Stevick, "The Oral Formulaic Analyses of Old English Verse," Speculum, 37 (1962), 382-89; Donald K. Fry, "Caedmon as a Formulaic Poet," in Oral Literature, pp. 41-61. 25 See Lord, pp. 4-6; Michael Zwettler, The Oral Tradition of Classical Arabic Poetry (Columbus, 1978), p. 11. 26 Genette DeWitt Ashby, "A Generative Grammar of the Formulaic Language in the Single Combat of the Chanson de Roland," Diss. Columbia Univ. 1976 (see Olifant, 8, No. 1 [1980]); Edmund de Chasca, "Toward a Redefinition of the Epic Formula in the Light of the Cantar de Mio Cid," Hispanic Review, 38 (1970), 251-63; Marjorie Windelberg and D. Gary Miller, "How (Not) to Define the Epic Formula," Olifant, 8, No. 1 (1980), 29-50. 27 Otto Holzapfel, "Die epische Formel in der deutschen Volksballade,"Jahrbuchfiir Volksdichtung,18 (1973), 30-41; Wolfhart Anders, Balladensanger und mundlicheKomnella poesia germanicaantica position (Munich, 1974); Teresa Paroli, Sull'elementoformulare (Rome, 1975) (from the Eddas to the Niebelungen). 28 Finnegan, pp. 58-72. 29 Adrian Fochi, Estetica oralitatii (Bucharest, 1980), ch. 5. 30 Michael Curschmann, "The Concept of Oral Formula as Impediment to Our Understanding of Medieval Oral Poetry," Medievalia et Humanistica, NS, No. 8 (1977), 63-76.

92

NEW LITERARY HISTORY

31 Menendez Pidal, Romancerohispanico, 2nd ed. (Madrid, 1968), I, 58-62. 32 Cf. Bernard Cerquiglini, Jacqueline Cerquiglini, Christiane Marchello-Nizia, and Michele Perret-Minard, "L'objet 'ancien francais' et les conditions propres a sa description linguistique," in Methodes en grammairefrancaise, ed. Jean Claude Chevalier and Maurice Gross (Paris, 1976), pp. 185-200; and the special issue of Langue francaise, 40 (1978), Grammairesdu texte m6dieval, as well as Bernard Cerquiglini, La parole medievale (Paris, 1981). 33 Rudolf Hirsch, Printing, Selling and Reading, 1450-1550 (Wiesbaden, 1967), p. 148. 34 Dmitry Likhachev, "L'etiquette litteraire," Poetique, No. 9 (1972), 116-23. 35 No. 21 in C. Appel, Bernart von Ventadorn,Seine Lieder (Halle, 1915), p. 121. 36 Paul Zumthor, Langue et techniques poetiquesd l'epoqueromane(Paris, 1963), pp. 82111. 37 Andre Bernard de Mandach, Naissance et developpement la chanson de geste en de Europe, III (Geneva, 1975), 5-20. 38 Zumthor, Lanque et techniques,pp. 51-53; Friedrich Gennrich, Rondeaux, Virelais und Balladen, 2 vols. (Dresden, 1921; Gottingen, 1927); Margit Sahlin, Etude sur la carole medievale (Uppsala, 1940); Nico van den Boogaard, Rondeaux et refrains, du XIIe siecle au debut du XIVe (Paris, 1969).