"La Tierce Geste" collects twelve of Madeleine Tyssens' essays on the Old French epic geste of Guillaume d'Orange. The selected works, most of which were previously published in edited volumes such as festschrifts or conference proceedings, represent only a fraction of Tyssens' remarkably prolific scholarly output (indeed, the volume opens with a bibliography of some fifty works authored by Tyssens between 1956 and 2008). In a short preface, the author signals the principle of selection: these essays constitute, along with her 1967 thèse , "l'essentiel de ce que j'ai cru pouvoir avancer au sujet de cette geste" (7). No principle of organization is signaled (the essays are not arranged chronologically)--I will thus discuss them in groupings that occur to me as a reader.
The theory that Tyssens outlines in the remainder of the preface is the result of her decades of philological labor: while the variance and instability of surviving texts have led some to imagine an oral of culture of endless improvisation and variation (cf. Paul Zumthor's concept of "mouvance"), Tyssens maintains that the careful study of the existing manuscripts reveals an intentional, organized textual culture with an aim of preserving, combining, and improving pre- existing texts. Furthermore, the study of surviving fragments at times allows for glimpses of earlier phases of textual development. These assertions are supported by the minute textual analyses of a scholar committed to "voir la réalité concrète des textes conservés plutôt que des projections théoriques ou des stades légendaires inaccessibles" (57). This insistence on the material reality of textual transmission- -"les textes sont lÀ, qui disent ce qu'ils disent," she asserts matter-of-factly in one essay (150)--is both a rejection of the speculative romanticism that characterized early medievalism as well as a implicit assertion of philology's self-sufficiency as a critical practice.
The two essays that open the collection also count among the chronologically earliest. "Le style oral et les ateliers de copistes" (1964) and "Le jongleur et l'écrit" (1966) find Tyssens looking over her shoulder at earlier scholars' preoccupations and theories. Charting a middle path between Joseph Bédier's "individualism" and the theory of oral formulaic composition promoted by Jean Rychner, J. Joseph Duggan, and others, Tyssens focuses on the issue of the "publication"--literally the "making public"--of chansons de geste as a conscious, collective effort of copyists laboring in organized ateliers. Variance, she argues, is not the result of de-centered, unregulated oral performance but rather of the copyists' concerted effort to stitch up the various pre-existing poems into the long cyclical compilations preserved in the majority of manuscripts. But what of the various allusions to orality, performance, and memory in the poems themselves? These, Tyssens argues, are built into the texts by authors who knew that their works would be performed aloud by jongleurs reading from a written source. In Tyssens' view, writing must always take precedence: it is the means of literary creation and transmission, as well as the physical traces that remain of those phenomena.
As promised in the title, the bulk of the essays focus on the Guillaume cycle, the broader cycle of Garin de Monglane/the Narbonnais, or individual poems from the cycle. In "Relectures de la geste des Narbonnais" (1987), first given as a plenary lecture for the International Congress of the Société Rencesvals, Tyssens provides an ambitious overview of the textual tradition of Guillaume d'Orange, from early fragments like the Hague fragment and the Nota Emilianense to the grand cyclical manuscripts constituting the "vulgate" of the Guillaume cycle; idiosyncratic manuscripts versions of certain of the works of the cycle; and even the fifteenth-century Roman de Guillaume d'Orange in prose. Though this survey must necessarily begin with a lost text (a first Prise d'Orange whose existence can be inferred from other sources), and despite her assurance that the poems of the "petit cycle" (the Couronnement de Louis, the Charroi de Nîmes, and the Prise d'Orange) all provide details allowing us to reconstitute "un archétype unique 'parfaitement fixé dans sa lettre'" (61) , the essay's primary pleasures and most valuable lessons derive from Tyssens' scrupulous attention to the details of textual variance: her magisterial lecture bears witness to a literary tradition that is varied, complex, ever- changing, and extremely popular.
In essays based on individual poems, Tyssens alternates between these dual preoccupations: the search for archetypes and "original" works on the one hand, and an engagement with the fascinating mess of the texts we have on the other. In "Le siège de Narbonne assonancé" (1969), Tyssens argues that a then-recently-published fragment proves that the text Hermann Suchier published in 1898 as Les Narbonnais is not in fact one poem, but two: a "Département des fils d'Aimeri" followed by a rhymed "Siège de Narbonne" of which the assonanced fragment is an earlier version. Similarly, essays on Aliscans, a poem whose textual tradition is especially complicated, point to earlier versions of a Chanson de Rainouart than the one that seems to constitute the final section of the Chanson de Guillaume (sometimes referred to as G2). In these essays--Aliscans: Fragment B.N. fr. n.a. 934" (1973) and "Encore Aliscans: Les enseignements du manuscrit Savile" (1982)--Tyssens advances her theory that medieval copyists worked as philologists avant la lettre: older and newer versions of poems circulated simultaneously, she argues, and copyists combined them to establish "éditions critiques" in the great cyclical manuscripts (see 190 and 215, in each case she cites from La geste de Guillaume d'Orange 249-50). Considering "Vestiges lexicaux dans le manuscrit D de la Prise d'Orange" (1980), by contrast, Tyssens invites us to appreciate an especially unusual variant. A line in which Guillaume and his men are likened to pigs in a sty may represent, she concludes, an original lesson, or it may be simply "une innovation heureuse, où se conjuguent la verve hardie de notre remanieur er la saveur des mots régionaux" (161).
Two further essays consider an often-overlooked chapter in the long history of the chanson de geste in Europe--the circulation of epic poems on the Italian peninsula, often in a mixed Franco-Italian koinè. In "Poèmes franco-italiens et Storie Nerbonesi. Recherches sur les sources d'Andrea da Barberino" (1989), Tyssens poses a very concrete literary historical question: if Franco-Italian versions of French epics exist, are they indeed the versions that would have served Andrea, the fifteenth-century Florentine composer of prose texts? Via an analysis of three Franco-Italian texts-- Aliscans, Fouque de Candie, and a Testament de Charlemagne--Tyssens answers in the affirmative, only to arrive at a further question: are all existing Franco-Italian poems based on French originals? This question is of no small consequence, since romance philology has maintained throughout its history a strong preference for the early over the late, for the original over the copy, and for the national over the "international." This open- mindedness toward the Franco-Italian tradition is surely a factor in Tyssens' unorthodox suggestion, in "La version franco-italienne d'Aliscans" (2001), that the version of the poem recorded in the Venice Marciana manuscript serve as a base text for an edition of Aliscans. It is in this linguistically mixed version, Tyssens argues paradoxically, that the earliest French version of the poem is to be found.
In the remaining three essays, Tyssens gives the most direct defense and illustration of the philological science practiced by scholars of her generation, and inadvertedly reveals some of that philology's limitations. In "Typologie de la tradition des textes épiques: les poèmes français" (1990) and "Philologie 'chevronnée,' nouvelle philologie" (2002), Tyssens quarrels explicitly with new modes of critical practice. In the first essay, a lengthy meditation on the tricky business of localizing manuscripts and the texts they preserve, she takes aim at computer-aided research on dialectology (specifically Anthonij Dees' Atlas des formes linguistiques des textes littéraires de l'ancien français) . Here Tyssens introduces a distinction between copies and texts that would seem at odds with her stated intention of sticking to the reality of what remains rather than speculating on what is lost (49). Copies may be digitized and fed through computers as data; but the text, the archetype, the "poème premier" (56), she insists, can only be discovered through the philologist's craft. In the second essay, Tyssens mounts an even more spirited and explicit defense of traditional philology, or "philologie chevronnée." In response to two Dutch scholars' critique of traditional editing methods , Tyssens reasserts the practiced philologist's unique ability to discover, correct, and publish a literary work. Rejecting a younger generation of philologists' fantasy of a computer-generated and digitally-disseminated "édition 'totale'" (175) reproducing all variants, she insists on the goals of traditional text editing: "donner À lire un texte" while eliminating errors "qui viennent perturber le sens" (176, emphasis in original). The practice required to achieve these objectives is not data analysis, but interpretation (my emphasis), an imaginative ability to probe the mysteries of the text that no computer program could ever emulate.
Where, then, is the medieval literary text? In the material remains of textual culture, the variant manuscripts themselves, or behind them, in a realm only accessible through philological science and interpretation? While Tyssens remains rigorously focused on the former, the ideology of her discipline forces her to interpret this evidence according to a set of implicit and explicit presuppositions: every copy must have an original; every manuscript is a "witness" to a model; every text has an author; one lesson must be better than the others. The limitations of this approach are perhaps most in evidence in Tyssens' only attempt to engage with modern literary theory, "Aspects de l'intertextualité dans la geste des Narbonnais" (1997). This essay begins with the promising assertion that the Narbonnais cycle can be read through the lens of Julia Kristeva's concept of intertextuality, since "La Geste semble se réverbérer en chacune de ses chansons" (83). Tyssens' reading of the corpus is characteristically exhaustive as she catalogues the various ways in which each song of the cycle references the others: stock phrases and formulaic language; consistency of characterization across the corpus (Guillaume is always recognizably Guillaume, etc.); family resemblance between characters; and the recapitulation of geste narratives within individual works (as in when characters remind each other of events recounted in other poems). But the radical thrust of Kristeva's concept--that texts are not intertextual because they refer to other texts, but rather because they are themselves inescapably made up wholly of other texts, both at the moment of composition and at the moment of reception or reading--is dodged, and quite willfully. The essay concludes with a rejection of Zumthor's comparable concept of "mouvance," which haunts every essay in the volume. Taken together, Kristeva's and Zumthor's visions of the literary undermine any notion of originality, authenticity, or singularity. For Tyssens, this is a bridge too far: existing versions should be viewed neither on their own terms nor as a decentered network. "Ces remaniements," she insists, "ont nécessairement été élaborés dans et pour le livre et par un recours direct À des textes, in praesentia (copies disponsibles) et parfois, mais rarement, in absentia (recours À la mémoire)" (99). Traditional philology thus permits its practitioners to imagine that which is not there (the lost text, archetype, model, intermediary step, etc.), but it does not seem to permit one to imagine the literary differently.
1. La geste de Guillaume d'Orange dans les manuscrits cycliques. Paris: Les Belles Lettres (Bibliothèque de la Faculté de Philosophie et Lettres de l'Université de Liège, CLXXVIII), 1967.
2. Tyssens cites her own phrase from "Le Charroi de Nîmes et la Prise d'Orange dans le manuscrit B.N. fr. 1448" Cahiers de Civilisation Médiévale t. III (1960): 98-106 at 106.
3. Tübingen: Niemeyer, 1987.
4. Lene Schøsler and Pieter can Reenen, "Le désespoir de Tantale ou les multiples choix d'un éditeur de textes anciens. À propos de la Chevalerie Vivien, éditée par Duncan McMillan" Zeitschrift für romanische Philologie CXVI (2000): 1-19.