Island Philology

Les fabliaux

Bédier’s doctoral thesis, Les fabliaux, critiques theories of origins in order to rid French comic tales of foreign influences. When Bédier began his study, critics sought to explain the tales’ origins through comparative analysis, and most argued that they spread to Europe from India through a series of direct influences. Bédier, for his part, rejected both the comparative method and the thesis of Indian origins. Throughout the Fabliaux, he expresses annoyance at being “brought back to India.”46 He methodically blocks the “caravan of tales” that Gaston Paris saw traveling from the banks of the Ganges to the banks of the Seine.47 He also ridicules the methods of comparative anthropology, which used evidence from contemporary tribal culture to explain ancient Europe: “Here it’s a ancient totem, there a tabou, and to explain these marvels, one must sometimes look to the Bassoutos [of South Africa], the Hurons [of Canada], the Kamchadales [of Siberia].”48 Bédier here rejects the colonialist medievalism prevalent in the late nineteenth century. Rather than looking to the time when “the ancestors of the Germans resembled the Zoulous,” Bédier posits the futility of the very idea of a single origin.49 Instead, he argues that comic tales exist in many different places—Antiquity, France before the Crusades, India, Africa, Brittany, Germany, etc. Bédier refuses, in other words, to see India or anywhere else as the source of French literature.

Bédier’s outrage at the idea that French literature originated in India is scientific, patriotic, and also colonial. In the tradition of nationalist medievalism, he blames German scholars for extending a relatively modest French theory to monstrous proportions.50 Equally important for a native of Réunion from a prominent family involved in conquering and administering French Indochina, the Indianist theory granted cultural privilege to a newer colony seen as subordinate and distant from French values. French India also competed with Réunion for scarce metropolitan resources. Bédier’s adolescent years on Réunion, moreover, coincided with a period of substantial migration from British India under indentured labor agreements, almost all passing through Saint-Denis.51 These arrangements signaled an “embarrassing” dependence on a rival empire. Indian immigration led metropolitans to associate Réunion with “orientalism,” which, in the eyes of elite creoles, distanced the island from the privileges that should accrue to their faithfully French society. To combat this reputation, the organizers of the Réunion pavilion for the colonial exposition of 1931 minimized the visibility of Indian culture; Marius Leblond (in his own assessment of fabliaux) claimed that Réunion’s own popular tales imitated Horace and La Fontaine, owing nothing to Indian sources.52 The derivation of French national literature from India posed, in short, a sharply colonial question: Bédier’s theory erased any signs that France had ever depended on “inferior” colonies for inspiration.

In his arguments against the Indianists, Bédier posited that contacts between cultures play a minimal role in literary creation.53 Rather, different cultures spontaneously and independently produce similar stories. Bédier’s proof of this theory of “polygenesis” rests partly on a tale from his own imperial experience. Toward the end of his refutation of Indian origins, he recounts a story from his trip back from Bourbon:54

In the month of October, 1887, at the level of Cap Gardafui, on the steamship the Yarra of the Australian line, I heard stories told. The narrator was a old inhabitant of Mauritius, who was leaving his island for the first time. He told, among other tall tales, the story of a certain examination that a father makes his three daughters pass in order to see which one of them needed to be married first. The tale, which it is impossible for me to analyze more precisely, is a fabliau. It is likewise impossible for me to say the title of the fabliau, but it can be found in volume V of the collection by Montaiglon and Raynaud, under the number 122. Even though I must have read a hundred collections of Κρυπτάδια [bawdy tales], I have never encountered this tale anywhere else, and I doubt if it’s ever been written since the thirteenth century. The old Mauritian planter, however, told it like a [medieval] jongleur, without adding or cutting a single episode. I asked him where he had gotten his story, and I received the response well known to collectors of tales: “Who knows? I heard it told just this way, undoubtedly at Port-Louis, I don’t know anymore when or by whom.” He was therefore a witness to oral tradition. I noticed then that among the listeners were an English businessman who was from Sydney and a seaman whom we called the Martigaw, because he was from the Martigues. The next day, I heard the Martigaw recount the fabliau to a circle of seamen. The crew was almost exclusively composed of Basques and Corsicans, but the listener who seemed the most amused, and who showed the biggest teeth when he laughed, was an Arab stoker who had just refueled the machine and who, his naked body glistening with sweat, was drinking his minuscule cup of coffee. One can say that, that day, this tale had passed from the Mascarene Islands to the Basque country, to Corsica, to Australia, to Arabia. Besides which, on the ship itself, it could have passed further to the Chinese boys and the Piedmont [Italy] laborers who were coming back from Bourbon, the Arab could have told it in Aden, the Martigaw in Provence, a Corsican in Bastia. Collectors of tales who perhaps, in these last five years, have gathered this story in Aden or in Moka [Yemen], in Marseille, in Dax [southwestern France], will compare gravely these versions that they’ll proclaim Arab, Provençal, Basque, and they’ll look for the laws of propagation of this tale. What likelihood that they’ll ever be discovered?55

Cap Gardafui forms the tip of present-day Somalia, and marks the boundary between the Gulf of Aden and the Indian Ocean. Moving through this geographic nexus of east and west, south and north, Bédier hears a medieval tale. His own telling is fraught with contradiction. While he identifies the tale’s genre and date with precision, its linguistic form remains entirely unclear: the Mauritian could have spoken French, Creole, or English (Mauritius, formerly Ile de France, was captured by the English in 1810). The description of the audience does little to clarify, since Bédier mentions both an English speaker (the businessman from Australia) and a French speaker (the seaman from Martigues, near Marseille). The languages of transmission are even less clear when the Provençal seaman repeats the tale to Basques, Corsicans, and an Arab—and Bédier goes on to imagine Chinese and Italian listeners (who could have spoken Piedmontese or Occitan). As a parable of literary transmission, then, the account hardly inspires confidence. The only two people who understood the story for sure are the Martigaw and Bédier: the other transmissions are entirely imagined.

What’s more, Bédier himself does not repeat the tale, referring readers by number to the published edition (after having shifted into the present tense of oral story telling). His silence arises not only from gentlemanly propriety, which prevents him from repeating the tale’s title, “Trial of the Cunts” [Le jugement des cons].56 It also enables Bédier to stand outside the stream of narrative that flows simultaneously from the Indian Ocean and the Middle Ages, a stream shaped by multiple colonial histories. By not repeating the tale, Bédier keeps the Euro-French version firmly on French soil. His double-processing of memory (one conditioned by colonial experience, one by scholarly study), however, suggests the impossibility of actually disentangling the lines of transmission: it becomes difficult to say which came first, the medieval tale or the Mauritian. In this sense, Bédier’s anecdote “proves” polygenesis.

While Bédier’s anecdote ostensibly supports his thesis that the fabliaux have no single origin, he actually does describe a linear transmission process. The Arab, Australian, Corsican, Chinese, and Italian versions of the tale that Bédier imagines do in fact have a single origin—the Mauritian planter. The problem of the tale collectors whom Bédier ridicules is not so much they are wrong to seek the tale’s origin but that they lack sufficient information to find it. They specifically lack the imperial imagination that could trace people from such disparate places to a single location, a single conversation, in the Indian Ocean. Moreover, Bédier’s insistence on the tale’s exact conformity with the medieval French tale implies that the Mauritian version extends, unchanged, a direct line of transmission from medieval France. Could the tale indeed have traveled directly from France to Ile de France, just as it was now clearly returning to France? Upon closer inspection, then, Bédier’s anecdote profoundly contradicts the thesis he wishes to support.57 The anecdote actually shows potentially broad propagation (in both time and space) from a potentially single source. The contradictory nature of the anecdote, which both proves and disproves Bédier’s thesis, reveals the strategic nature of his insertion of colonial explanation into medieval literary history.

Bédier’s anecdote demonstrates the remarkable continuity of French culture (stretching from the thirteenth century to the late nineteenth, from the metropole to a distant former colony). It also illustrates how casually cultural material can pass from one “national” context to another. Bédier concludes that it suffices for two people knowing the same language to meet, anywhere, for stories to take root in new contexts. He characterizes his surprising encounter aboard the Yarra as proof that popular tales belong wherever they are repeated: “The homeland [patrie] of tales is not where they are born, but where they are comfortable.”58 This conclusion echoes the famous phrase of Roman imperialism, “ubi bene ibi patria” [where you prosper, you’re home]. Bédier’s commitment to fixed identities, however, belies this flexibility. His theory of “polygenesis” actually relies on clearly distinguishable (and unchanging) cultural traits: “India invents Indian tales, France French tales, Armorica Celtic tales, Zoulouland Zoulou tales.”59 Covering both the French and British empires, Bédier’s examples isolate cultural groups from one another—and project an identifiable “French” culture into the distant past. Medieval crusaders, he continues, encountered both Arabs and other Europeans overseas, but their respective cultures—and popular tales—remained unchanged by the experience. The contradictory anecdote from the Yarra thus supports a fundamentally contradictory thesis: on the one hand, tales do have specific origins (however invisible to later observers); on the other, they migrate freely such that their origins cease to carry meaning. Aboard the Yarra, the passengers remain identified with their homelands; the ship itself is no patria even though tales are being invented there.

Despite Bédier’s claim that tales reroot themselves wherever they find a comfortable home, his overall approach remains deeply nationalist. Indeed, his thesis extends the patriotic medievalism undertaken by Charles Lenient. Like Bédier, Lenient complained of being “taken back to India” by German scholars (“charlatans érudits” studying the origins of the epic). Lenient actually mocks the motto of Roman imperialism (“ubi bene ibi patria”), countering with the example of French colonial settlers who remain faithful to their homeland even after generations of life overseas.60 This approach to faithful national identity coincides fully with Bédier’s conception of creole identity. In the Fabliaux, Bédier cites Lenient’s popular La satire en France au moyen âge (1859, 1877, 1883) to define the fabliaux’s “esprit gaulois” as a quintessentially French national trait. Lenient, in line with the “decolonization” of the French Middle Ages after 1870, treats “Gaul” as a synonym for “France.”61 The fabliaux’s Gaulish dimension represents yet another way in which the genre, in Bédier’s analysis, lends credence to the deep historical formation of a continuously national literature.

Bédier’s thesis on polygenesis largely succeeded in displacing genealogical approaches to textual filiation. The question of the fabliaux’s “Indo-Aryan” origins has largely fallen from prominence (although the comparative mythology of Georges Dumézil still attracts adherents). The study of the fabliaux today is more likely to concern eroticism, economics, humor, or manuscript collections. On one level, this shift in scholarly interest merely reflects general twentieth-century critical trends. On another level, though, Bédier contributed to severing the genre from ethnographic historiography. Having set aside the question of origins, he turns in the second half of his book to literary questions. Scholars have since pursued any number of ingenious topics, yet arguably many studies still fall within the general parameters Bédier defined: historical setting (such as class relations and audience) and literary representation (especially women, marriage, and religion).62 Bédier accomplished this shift largely by insulating French literature from outside influences: the medieval fabliau and the Mauritian jongleur may tell the same tale, but they have nothing to do with each other. Bédier’s “decolonization” of the fabliaux (through the rejection of Indian origins and an emphasis in Gaul) also “colonizes” them as pure French forms.

Bédier’s “nationalization” of the fabliaux was immediately criticized by Brunetière, in ways that are still useful:

Although not resembling in every way those of our European women, are the dreams of a negress, in the Sudan, less feminine? And subjected to the same necessities, exposed to the same challenges, do the needs and ambitions of a Chinaman differ very much from ours?63

Brunetière counters Bédier’s refusal to compare tales with an openness to cultural communication, and an appeal to common humanity. Although this projection of universality can lead to oppressions and misprisions of its own, it usefully resists Bédier’s Francocentrism and racial ontology. To the extent that critics today treat the fabliaux as “French” literature, they work within paradigms that isolate the genre from global migrations. Setting aside analogues that may persist around the world, or traces of multi-culturalism within the French-language versions, means acquiescing in part to the formations of colonial philology. Tale-tellers like the Mauritian did travel; his tale may indeed have taken root in Aden. While specific origins usually lie beyond our view, and the broad sweep of comparative Indo-European mythology raises more questions than it answers, judicious comparisons might yet reveal new literary and cultural histories. Regardless of the approach, fabliaux study can engage the tales’ contradictory imperial formations.

The Mauritian planter made at least one return trip to the Indian Ocean: Bédier’s Fabliaux arrived on Bourbon in May 1893, to the delight of his mother and stepfather.64 Although the book could not inspire the further propagation of tales (since Bédier tells hardly any), it could reassure a colony far from the metropole of its closeness to France. Bédier’s proclaimed conviction that the Mauritian jongleur provided an absolutely faithful performance of medieval oral tradition could strengthen creoles’ belief in their own fidelity to France: the preservation of ancient traditions underwrites creole claims to special treatment within the empire. Bédier’s parallel argument, moreover, that cultures can create similar tales even without direct lines of transmission, provides a second (if contradictory) support for creole privilege: even cut off from France, Bourbon can produce literature that resembles metropolitan aesthetics. Finally, to the extent that Bédier’s insight derives from colonial memory and a formative experience shaped by colonial economies (all the reasons that people of such vastly different origins found themselves traveling on a ship across the Indian Ocean), global imperialism shaped the historiography of medieval France.

Tristan et Iseut

Even before Bédier began his thesis on the fabliaux, he published a comparison of French and German versions of the story of Tristan and Iseult—lovers destined for unhappiness and death after Iseult’s marriage to King Mark. Bédier’s 1886 article actually makes two points identical to those developed in the Fabliaux: it is useless to identify the “oldest” version of a story; two authors can independently invent the same details.65 These critical principles lay the groundwork for Bédier’s “nationalization” of the Tristan romance. While for the fabliaux this process involves dismantling theories of Indian origins, for the Tristan materials it involves the denial of Celtic influences, along with the recuperation of the French sources behind texts in other languages. Bédier’s dealings with the Tristan materials exploit philology’s full potential to comfort national identity through the construction of coherent memory fictions. They also fall under the persistent shadow of creole longing.

Bédier worked primarily to discredit the widely held idea that the surviving Tristan materials derived from Celtic sources. His main argument presages his approach to nationalizing the Roland: the French Tristan et Iseut has a single source—the French author of genius who first conceived it; all evidence to the contrary represents corruption by less talented writers.66 He argued further that passionate love did not exist in Celtic culture.67 Much later, he published a popular article whose title condenses his position succinctly: “The legend of Tristan and Iseult is essentially French” [La légende de Tristan et Iseut est essentiellement française]. By insulating the French materials from “foreign” influences, Bédier established a centrist and monolingual literary history. When he elaborated this approach in the Histoire de la littérature française illustrée, Breton reviewers objected strenuously.68 For Bédier, regional autonomy threatened France’s sacred unity; local traditions (be they Breton, Provençal, or any other) contributed nothing substantial to national literature. This approach extends the broader “colonialist” aggressions of republican reforms: Brittany had sustained one of France’s strongest regional cultures, and lost the most to the consolidations of the Third Republic.69 Paradoxically, creole claims to imperial privilege rested partly on the promotion of Celtic heritage (“a Bourbonnais with blond hair and blue eyes,” as Bédier said).70 Bédier’s embrace of colonial Celticism thus purifies creole lineages, while his resistance to medieval Celticism purifies national origins. Both moves connect republican France to an idealized history that “forgets” the many ways in which métissage has shaped the nation.

Bédier imposes a similar homogeneity on the fragmented corpus of Tristan manuscripts. Contrary to his attack on single origins in the Fabliaux, Bédier argues for a single “ur-Tristan” romance written prior to the surviving text from “Thomas” (itself the source of all other surviving versions, according to Bédier). To convey the original form, Bédier produced an admittedly “composite” text.71 The result includes both edited manuscript fragments and conjectural reconstructions in modern French (the “edition” actually contains proportionally few pages of medieval French).72 Having “restored” the romance, Bédier measures the surviving texts against it, concluding that even Thomas’s text falls far below the aesthetic perfection of the French original, his Anglo–Norman dialect blocking access to the author’s own language (which Bédier considered comparable to Racine’s).73 Obliged to make due with a “best” manuscript in Anglo-Norman, Bédier strives to purify its forms: Thomas’s phonetics are “remarkably pure,” his inflections “remarkably regular”; the language is actually quite close to continental French: “his phonetics contains hardly any colorings or traces of Anglo-Norman.”74 These linguistic arguments support Bédier’s broader conclusion that Anglo–Saxon culture had no influence on Thomas: he may have resided in England but there is no reason to consider him English or for German critics to place him in the “Germanic” tradition.75 In a similar vein, Bédier praised the French culture of his own English translator, Hilaire Belloc: Bédier characterized him as an exemplary French patriot because he lauded the Roland, the Crusades, and contemporary French colonialism in North Africa; Belloc, like Fustel de Coulanges, had rid his nation of the “legend” of its debts to Germanic invasions.76 These terms of praise legitimate Belloc as a voice for France, even in the English language. Bédier thus nationalizes the Tristan materials by casting linguistic differences (medieval and modern) as incidental to cultural identity.

At the same time that Bédier worked on the editions that brought his arguments to scholarly audiences, he also nationalized the story of Tristan and Iseult through his own modern French narrative. The Roman de Tristan et Iseut (1900) offered the French public a “native” version to rival the popularity of Wagner’s German opera Tristan und Isolde (which premiered in Paris in 1899).77 In the process, Bédier turned the surviving manuscript fragments into a coherent narrative of aesthetically integrated episodes—in a style that became a model of elegant French.78 Throughout, he suppresses elements of savagery and sexuality that might indicate “primitive” influences, opting instead for a classicizing style that enhances courtly manners, reduces ambiguities, and unifies tone.79 These methods are meant to “recover” the lost original. Indeed, the early editions of Bédier’s romance variously characterize it as a “restoration,” “renewal,” and “reconstitution” [restauré, renouvelé, reconstitué]. All of these terms cast the text as a repair for the distortions wrought by the accidents of history.

Bédier accomplishes this transhistorical performance of a French national literary aesthetic through deep identification with the medieval sources. In a draft of his preface, he writes that he sought “historical and critical sympathy,” “impregnating” himself with the authors’ color and style.80 Gaston Paris ratifies this performance of historical sympathy in his own preface: Bédier has composed a twelfth-century poem at the end of the nineteenth century.81 Bédier seals his sympathetic voicing of historical feeling by concluding the romance in the first person:

Lords, the good poets of yesteryear, Béroul and Thomas, and my lord Eilhart and master Gottfried, told this tale for those who love, not for anyone else. They send you, through me, their greetings.82

Bédier makes himself the conduit of a transhistorical telepathy—connecting himself to medieval sources just as he later envisioned himself connected to all creoles (see chapter 3). Speaking literally on behalf of medieval authors (and treating the German sources as identical to the French ones), Bédier places himself in a continuous line of transmission, much like the Mauritian jongleur—heir to a lengthy tradition and also an autonomous inventor of a tale that will have a life of its own.

Bédier’s historical “sympathy” engendered a philology tied more to his sense of absolute “truth” than to the actual evidence of documents. Indeed, Bédier replaced a short passage in his Tristan et Iseut that he had actually translated with an extensive dialogue of his own invention because “to my mind, it’s the original form of the episode, even though no text conserves it; it alone satisfies the spirit.”83 The “true” text here is not the one documented in surviving manuscripts, but the one that best satisfies Bédier’s aesthetic sensibility. Bédier embraces “truth” above philological accuracy in other contexts as well, from situations of relative insignificance (Gaston Paris’s lecture notes) to publications of profound ethical impact (documents related to German war crimes): he averred that his texts could be both false and authentic.84 Bédier returns to the idea of sympathetic truth when asked, on the occasion of his entry into the Académie Française, how he had composed Tristan et Iseut:

I believe that old texts have a soul and that it’s useless to waste one’s time deciphering them if one does not feel one’s soul in sympathy with them . . . there should not be any difference between the work of the scholar and that of the novelist.85

In startlingly frank terms, Bédier conflates objectivity and subjectivity, analysis and imagination. Indeed, his “edition” of Thomas’s Tristan, with its long passages of modern French, deserves to be called a “novel” almost as much as Tristan et Iseut. Bédier’s belief in a lost “original” Tristan romance led him, in the end, to a reconstructive method even more radical than the critical edition.

Bédier’s ability to identify with the Tristan story derives not only from his historical imagination but also from his colonial experience. In addition to the general ways in which creole chivalry made Tristan’s story recognizable, Bédier claimed specific inspiration from his life in Saint-Denis. He reportedly told the Mauritian poet Robert Edward Hart in 1930 that he never would have written certain pages of Tristan et Iseut “if I hadn’t once been in love, at age fifteen, with my cousin.”86 Bédier here claims a creole origin for the very language of his romance. His legend of creole love lends credence to Corbellari’s speculation that Réunion inspired Bédier’s enigmatic reference to a “Pays Fortuné” (a term commonly applied to idealized colonial dominions).87 Bédier’s comment on his youthful love also places him in the company of Bourbon’s prominent poets—from Parny to Leconte de Lisle to Dierx—who founded their creativity on the doleful remnants of impossible love stories, usually with their cousins (Hart explicitly compares Bédier to Leconte de Lisle). By basing his “medieval” text on this kind of creole love story, Bédier imports the racialism of creole chivalry into the romance of knightly valor and sublime love. These common tropes of medieval literature thereby reveal their own exclusionary effects. Bédier’s personal identification with Tristan’s disappointed love and exile thus mobilizes colonial experience as an explanation for medieval narrative. And since idealized medievalism contributed to creole culture, it becomes difficult to disentangle the lines of transmission that run simultaneously from the Middle Ages and the Indian Ocean.

Bédier’s personal story of creole love opens Tristan et Iseut toward the colonial imaginary—and specifically toward its counterpart in tragic love, Bernardin de Saint-Pierre’s Paul et Virginie (1788). Although set on Mauritius (Ile de France) and written by a metropolitan, Réunion creoles embraced the novel’s idealized depictions of creole life as their own. The Leblonds and others considered Bernardin’s narrative the prototype for the modern colonial novel.88 Indeed, what Chris Bongie calls Bernardin’s “neopastoral fantasy of global harmony” accords quite directly with the Leblonds’ racialist ideologies.89 Bernardin tells the story of two young children raised in pastoral bliss; colonial and class prejudices disrupt their pure affections and send Virginie into exile in France. Returning to the island after being repudiated by her opportunistic metropolitan family, she perishes in a shipwreck within sight of the shore; Paul dies of heartbreak soon after. This spare summary of the complex narrative suggests its basic affinities with the themes of “impossible love” and exile in Tristan et Iseut. Virginie, like Iseult, is praised as a “faithful lover” (183, 187); Paul, like Tristan, wanders the woods heartbroken (192–93).90 Indeed, many of the epithets that Gaston Paris applies to Tristan et Iseut could easily describe Paul and Virginie: they illustrate the “fated love,” which endures even when “battered by all kinds of storms”; their emotions and speech seem “half medieval, half modern.”91

Bédier’s own connections to Paul et Virginie go beyond those of the average French reader, or even the average creole. Bédier’s father Adolphe had asserted (in the book of family history he wrote for his children) that Bernardin’s models for Paul and Virginie were Paul Thuault de Villarmoy and Virginie Caillot—great-great-uncle and cousin respectively of Adolphe Bédier. According to Adolphe, the two young cousins drowned in the historical shipwreck that inspired Bernardin’s novel.92 He reminds his children that the family dishes bear the arms of Thuault de Villarmoy and that Paul’s portrait hangs at the house of their great-aunt. Although in Adolphe’s version Paul and Virginie are not in love with each other, a heartrending love story does surround Virginie’s demise: a young Frenchman on board the ship, “whose name escapes me today, who was going to the Colonies” fell violently in love with Virginie [il se sentit épris pour elle d’un violent amour]. As the ship began to sink, he offered to save her but she refused to remove her clothes; he reached the shore but swam back to try once again to persuade her. Terrified by certain death, she agreed to remove “her most cumbersome clothes” but the next day the sea washed up both of their bodies.93 Thus, according to Adolphe, did Bernardin encounter the core of the drama that concludes his novel: Virginie refuses to remove her clothes in front of a sailor and so drowns; her body is found alone the following day.

For Adolphe, the story of Paul Thuault de Villarmoy and Virginie Caillot linked the Bédier family to the origins of what had become the most famous creole narrative of the nineteenth century (among both metropolitans and islanders). Indeed, he enhances the story’s “Bourbonnais” connections in several ways. First, his story lacks the class differences that contribute to Virginie’s death in the novel: Virginie Caillot exhibits even more sensitive virtue than her fictional counterpart, for she refuses to remove her clothing in front of a sympathetic social equal rather than recoiling from an already naked sailor.94 Paul Thuault de Villarmoy and Virginie Caillot appear, moreover, as faithful migrant creoles, just like the Bédiers: both spent a substantial amount of time in France without ever altering their attachments to their island home.

Adolphe’s story, like the novel’s, ends at the gravesite on Ile de France:

They were each placed in a separate tomb: the two tombstones stand a half-league from the scene of the shipwreck, on the banks of a small stream, one on the right bank, the other on the left bank, in the middle of tufts of bamboo. They are facing one another, in the same form, of the same height, separated by the water that is not even twenty feet wide: no names on the tombstones. This is because the property where they stand belonged, undoubtedly, to Virginie’s family, who had no need to engrave her name on the stone to know who lay there. You will find in my desk two small fragments that I detached from each of the tombstones. The memory of this place will never leave my mind.95

Adolphe relies here as much on imagined speculation as a novelist . . . or a “restorer” of medieval sources: the tombs remain nameless because they “undoubtedly” lie on the property of the Caillot family; Paul Thuault de Villarmoy “must” have been mentioned more often to Bernardin than Virginie’s would-be rescuer. Instead of identifying Paul and Virginie, the tombstones record the marks of numerous visitors, including Adolphe, who took the liberty of breaking off pieces. The tombs thus serve several narrative functions: they mark “real” history (contrasted with the novel’s “fiction,” which leaves the graves unmarked even by tombstones);96 they materialize the transfer of creole memory to France; they ground creole identity in a dual geography that begins (rather than ends) overseas. Adolphe’s tombstone fragments reference memories both broken and repaired, a phantom history in which the choice between two homelands never arose.

Adolphe’s story, with which Bédier was intimately familiar, relates to Tristan et Iseut differently than to Paul et Virginie. Most superficially, Adolphe’s love story unfolds on a ship, like Tristan and Iseult’s (and unlike Bernardin’s). Alongside Bédier’s memory of his own beloved cousin, Bédier lived with the memory of the real and imagined histories of Adolphe’s “Paul” and “Virginie” (perhaps even inheriting the tombstone fragments). Adolphe’s narrative of tragic young love thus provides another source for Bédier’s “sympathetic imagination”:97 the love story of “cousine Virginie” compliments the imaginative riches of the medieval sources. Finally, Adolphe’s book documents the primacy of individual memory in the transmission of cultural knowledge. Adolphe turns fragments of inherited stories into a coherent written narrative—much like Bédier filling in the “missing” links between fragments of the medieval Tristan materials. Adolphe’s singular document, like Thomas’s and Bédier’s, serves as a nodal point between past and future identifications.

Even without Bédier’s anecdotal, and belated, linking of Tristan et Iseut to colonial experience, the romance engages Bourbon through its dedication—addressed to Bédier’s stepfather: “A mon cher Du Tertre. Hommage filial. Joseph Bédier” [To my dear Du Tertre. Filial homage. Joseph Bédier]. According to Bédier, Du Tertre provided his stepsons with a model of exemplary citizenship: “the simplicity of devotion, the virile and joyful accomplishment of daily duty, a simple and serious understanding of life.”98 The dedication thus signals an attachment to father and fatherland. The emphatic printing of Bédier’s full name, however, reveals a genealogical rupture—filial devotion to a man with a different name. In this case, the rupture coincides with continuity, for Du Tertre is also Bédier’s mother’s maiden name (the stepfather is also a cousin). The doubly continuous and discontinuous family relations implied in the dedication capture anew the dilemmas of always already displaced creole identities.

On the island, Du Tertre represented traditionalist creole interests. Elected mayor for three different terms between 1900 and 1914, Du Tertre participated actively in a tumultuous period of island politics. Electioneering on the island almost always generated violence, vote corruption, and partisan manipulations, to which Du Tertre was no stranger.99 His allegiances were primarily to the island’s elite Catholic landowners. Under the influence of his relative De Mahy, he formed an alliance with the métis journalist and colonialist lawyer Lucien Gasparin (Du Tertre’s son and Bédier’s half brother, Maurice, once seconded Gasparin in a duel); after De Mahy’s death, the two became rivals.100 Before his death in 1926, Du Tertre witnessed Gasparin’s election as deputy (in tandem with Auguste Brunet, 1924) on a platform of solid “creole” values that rested on their joint claim to native identity.

Bédier’s personal attachment to Du Tertre has little to do with these political connections. Yet the dedication does link Tristan et Iseut to the aspirations of creole republicanism—published in 1900 at the zenith of De Mahy’s career and just as Du Tertre became mayor of Saint-Denis. At this time, Bédier could have honored any number of influential mentors, from Brunetière (to whom he intended to dedicate the first volume of the Légendes épiques101) to Hermann Suchier (to whom he did dedicate that volume). In naming Du Tertre, Bédier instead underscores his closeness to Bourbon, as well as Bourbon’s reassuringly close ties to “ancient” France. Indeed, the dedication forged indelible associations among Bourbon, the Middle Ages, and Bédier: one of his obituaries on the island focused entirely on Tristan et Iseut and Du Tertre.102

Bédier’s affirmation of creole filiation (homage filial) also connects Tristan et Iseut to Réunion’s outsized reputation within the national canon of poetry: with his elegant “novel,” Bédier took his place in the pantheon of writers hailing from the “island of poets”—Parny, Leconte de Lisle, Dierx (see Introduction). As Gaston Paris writes in the first lines of the preface, “It is truly a poem, even though it is written in beautiful and simple prose”; Bédier writes as the “worthy continuator” of medieval poets.103 Ultimately, Tristan et Iseut constitutes a literary form of creole medievalism, grounded in the chivalric thinking that shaped so much of elite creole culture and extending the island’s vaunted poetic tradition into the Middle Ages. Tristan et Iseut participates in Bédier’s reflections on the persistence of colonial memory, the primacy of creole identity, and the global reach of medieval France.

Tristan et Iseut defined Bédier as a major writer. The romance, along with his war record, formed the primary basis for his election to the Académie Française (1920). To celebrate this honor, his compatriots gave him an ornate desk at which he vowed to work “for France, as a good creole,” for the rest of his life. Made from Bourbonnais woods, the desk featured two large illustrated medallions on the front—one depicting Tristan and Iseult and the other the dramatic peaks of Salazie, “one of the marvels of Bourbon” (and an area where Bédier vacationed in his youth) (Figure 33).104 The desk cannily, if inadvertently, ties together the medieval romance, the creole origins of Bédier’s version, and national import of this fusion (which catapulted Bédier to the highest levels of French prestige). The desk embodies nothing less than the materials that wrought creole medievalism—and that literally sustained it in Bédier’s later years.

Throughout these years, Bédier remained engaged with Tristan et Iseut. His dramatic adaptation, begun soon after the romance’s publication in collaboration with his cousin Louis Artus, played twice in 1929 and again in 1934; he also claimed to have planned a film version.105 The modernized narrative has inspired numerous creative artists, from Barrès to John Updike.106 And the romance’s conception of passionate love had a substantial impact on popular and scholarly treatments of the psychology of desire in the European literary tradition.107 By 1980, Bédier’s romance had reached its 576th French edition; translations remain available in numerous languages.108 Only since 1989 has another translation of the medieval sources, based on fragments rather than reconstructions, become available for students.109 Thus, until quite recently, French students read the story of Tristan and Iseult only in Bédier’s version; many still do,110 while English translations of Bédier appear regularly on college-level reading lists in the United States. In my own early experiences of medieval literature, I recall vividly the unnerving discovery that I had written more than one student paper based on Bédier’s text, in complete ignorance of its composite genealogy. The fact that Bédier’s romance still represents “medieval French literature” in various ways, and that it originated partly in a creole love story of the late nineteenth century, testifies to the broad and enduring reach of colonial formations in academic and popular culture.

Les légendes épiques

Bédier dated his manuscript of Tristan et Iseut 29 December 1896, claiming to have written the romance in two months while taking a break from editing the medieval sources.111 Earlier, in August, he said that he was working on the first volume of Recherches sur la formation des chansons de gestes—subtitle of the future Légendes épiques; his course that year was on the epic, and he had already taught the Roland in 1893.112 The fact that Bédier’s Tristan work precedes the Fabliaux (with an article in 1886) and interrupts the Légendes épiques (not published until 1908) suggests the importance of viewing his creative and critical activities as continuous reflections on national literary history. Indeed, taking into account the protracted process of bringing Tristan et Iseut to the stage, Bédier engaged the romance throughout his life and alongside all his other work (Corbellari considers it the “nerve center” of his scholarship).113 Bédier never gave up on “reconstituting its mutilated original” (contradicting all of his other philological work).114 Despite profoundly different goals, then, all three of Bédier’s major projects (romance, fabliaux, epic) address foundational moments in a nationalized literary history—and in each case creole memories underlie his convictions.

The epic genre represents the first founding moment of French literature (predating romance and fabliaux), and as such occupied a privileged place in the historiography of national identity (see chapter 1). Bédier offered the most ambitious arguments yet for rooting the genre firmly in French soil. Together, his Légendes épiques and edition of Roland deploy all the methods already applied to fabliaux and romance: Bédier hypothesizes elite authors of genius, rejects “foreign” influences, recuperates the Anglo–Norman dialect, and rewrites through translation. He thus extends to the epic the same kind of arguments for cultural homogeneity that he made for the fabliaux and the Tristan romance. Since the epic genre carried greater cultural weight than either fabliaux or romance, the Légendes épiques received great popular as well as academic attention, further reinforcing the genre’s political significance.

Bédier’s nationalization of the epic targets the dominant theory that the genre originated in oral songs preserved through generations until collected in the texts that remain. This theory of ancient, popular creation implied Germanic influence because the epic would have emerged before the development of distinct “French” and “German” cultures. By contrast, Bédier posited that the surviving epics originated much later with individual poets along pilgrimage routes emanating from Paris. This theory made the epic a solely French creation: “this poetry is entirely ours; it has nothing Germanic; it has nothing but French.”115 Bédier summarized his conclusions succinctly almost two decades after publishing the first volume of the Légendes épiques: “I opposed the theory of [the epics’] ancient, popular, Germanic origins with a theory of their recent, aristocratic, and entirely French origins.”116 For Bédier, the epic marks a founding break with “mere orality” that promises durable national memory.

Bédier solidified his belief in elite creation under the influence of Édouard Chavannes, his friend and eventual colleague at the Collège de France. Bédier praised Chavannes, a sinologist, for opposing folklorist doctrines and reestablishing the dignity of literary invention: Chavannes’s ideas, first published in 1894, influenced the Légendes épiques by making Bédier see that “popular art” was simply “an impoverishment of art itself.”117 In retrospect, Bédier claimed to have always believed in the unique role of elites in artistic creation:

As of my doctoral thesis, devoted to the fabliaux, I focused all my efforts, all my research on historical facts or legends, on national narratives become popular, but which, according to me, and contrary to the romantic theory, emerged originally from the elite.118

Bédier’s rejection of the role of the “people” in the formation of French national literature coincides with his rejection of scholarly romanticism, guilty not only of applying excessively rigid interpretive “systems” to historical materials but of granting too much creative power to the folk. The nation—like Bourbon—relies instead on the creative force of the aristocracy.

The formative role that Bédier assigns to pilgrimage also contributes to his nationalization of the epic. His idea that religious sanctuaries sponsored the invention of epics to attract visitors to their shrines distributes “national” sensibility throughout the provinces.119 The map of epic origins, in other words, is already a map of “France.” Medieval France thereby presages the unifying goals of the Third Republic, with regional differences minimized in favor of shared sentiments. The authors of France’s foundational literature have already forgotten, in the late eleventh century, the differences that separate them; even as they promote their local interests [la petite patrie], they contribute to national cohesion [la grande patrie].

Bédier himself traversed the provinces in his research, interviewing peasants and developing a living sociology of ancient national feeling. He suggests in a letter to Philip Becker that “scholarly” and “popular” traditions might not be so different as they usually appear.120 Modern pilgrimage furthered the epic’s nationalization, for the nationalist Catholic revival after 1870 popularized pilgrimage tourism; Pope Pius IX assigned the pilgrims the red cross, symbol of the First Crusade. As Bédier constructed the medieval epic’s pilgrimage geography, villages all over France were producing articles for the popular magazine Le Pèlerin, telling the story of their local shrines and famous medieval pilgrims.121 These modern storytellers performed the same function that Bédier attributed to clerical authors and jongleurs: they strove to attract pilgrims to their shrines. They thus reanimated what Bédier understood as France’s geopolitical and literary origins in the eleventh century.

The content and timing of Bédier’s research ensured its broad political appeal. Both before and after the war of 1914–18, the Légendes épiques received extensive praise for its rejection of every aspect of Germanic involvement in the French epic. In awarding Bédier the Prix Gobert in 1913, the Académie Française cited specifically his extension of “our national genius.” Partisans of the xenophobic right (Pierre Lasserre, Léon Daudet, Robert Brasillach) praised Bédier effusively, and triumphant anti-German notices of Bédier’s book appeared in the reactionary periodicals Action française and Le Gaulois, among others.122 But Bédier’s admirers crossed the political spectrum, with one of the more virulently anti-German appreciations appearing in La démocratie. And he received warm congratulations from Jaurès for his penetration of “the secret of France’s creative activity at one of its most noble moments.”123 Bédier’s election to the Académie Française in 1920 inspired a new round of appreciations for the Légendes épiques; the intervening war had only heightened the value of the anti-Germanic thesis. Pierre Champion extolled Bédier’s “conquest” of French philology; Henri Lognon his victory over “intellectual barbarism”; Grappe the triumphant spectacle of Becker’s admiration of Bédier’s conclusions; Hanotaux his repression of “the German invasion of this rich domain where it had installed itself so brutally.”124 Receiving Bédier at the Académie, Barthou confirmed the patriotic status of the Légendes épiques, in terms that could apply to Alsace: “you have returned to France what belongs to France.” Indeed, in a posthumous assessment (published in 1945), Karl Michaëlsson considered Bédier’s “degermanization” of French medieval studies his greatest achievement.125 Bédier himself connected his thesis on the epics’ genesis to their enduring impact on national thinking: “they still serve, after so many centuries, to fortify within us our national feeling, Golden Legend of our country.”126 For a number of different reasons, then, Bédier’s book became an emblem of national identity on multiple levels: it purified medieval origins, it refuted German scholars, and it fortified a culture shaken by military defeat.

Bédier’s nationalization of the epic also bears a deeply creole logic. Bédier’s Réunionnais compatriots of course appreciated his book for the same nationalist reasons as continental critics, adding distinctly colonial inflections: the German science that Bédier corrected was inappropriately “imperialist”; Bédier’s creole culture qualified him more than others to attack Germanism, since creoles preserved the most vigorous roots of the French race itself.127 Bédier’s defense of elite authorship, moreover, strengthened the claims that tied colonial privilege to an aristocratic ideology. And like Fustel de Coulanges, Bédier decolonizes French medieval culture itself: the epic emerges not with the Germanic invasions but with France’s own confident expansion during the Crusades. Finally, the pilgrimage thesis places purposeful migration in the service of national culture: people (like creoles) who travel toward their most “desired” destinations shape the literature most expressive of immutable French values.128 Bédier—who claimed a “pure genealogy” [sans mélange] —thus offers France and Bourbon a national epic without métissage, fit for an ancient global empire.

Bédier’s colonial experience directly supports his theory of epic composition. His argument rests on the idea that popular memory survives for only one or two generations: the events that inspired epics would have been forgotten rather than preserved. Bédier “proves” this theory by analogy with his own family memories:

A hundred years back . . . we no longer know anything of the past, except what paper has conserved for us. And how would we doubt this, I who write this, you who read me, if it is true that, without family papers, we wouldn’t even be able to name our great-grandparents, if it is true that we are incapable of identifying, on the walls of our father’s house, the old portraits or even the photos that nonetheless represent our immediate ancestors?129

With the performative “we” of shared experience, Bédier draws readers into a sense of collective loss. As he stares blankly at the anonymous creole faces that decorate his family home, he enjoins everyone to admit the fragility of memory. He then treats family genealogy as a microcosm of national history:

What remains today in oral tradition of events much more considerable [than the battle of Roncevaux in 778], of our wars in Algeria, Mexico, or Tonkin? Here and there a patriotic couplet that will be repeated as long as the tune isn’t too out of fashion, a proper name, a vignette.130

Here, Bédier posits that the events of Charlemagne’s lifetime survived as briefly as modern ones in the popular imagination—and so could not have been the direct source of texts written down centuries later. Bédier gives a colonial inflection to Renan’s theory of national identity: citizens and family members bond more over what they have all forgotten than over what they remember.131 Bédier’s colonial examples, moreover, echo his own family history:

One of my great-uncles, in 1809, had his kidney ruptured by a bullet in Madagascar, at Fort-Dauphin; one my direct ancestors served in India under La Bourdonnais;132 people with my name and blood serve today in Cochinchina, in Tonkin, in Sudan, and I know what this handful of Frenchmen, the creoles of Bourbon Island, lost beyond the seas, have given to France in good servants, soldiers, mariners, men of state, industrialist, poets.133

Bédier’s family members participated in the same colonial wars that he compares to Roncevaux. This connection ennobles modern aggression while making France’s oldest epic a celebration of imperial expansion. Since the story of Bédier’s family coincides with imperial history, the Roland looks like a personal archive—the great-uncle in Madagascar heir to Roland.

Bédier concludes his reflections on the fallibility of oral transmission with a discussion of the book written by his father—“family papers” designed to preserve memory against the ravages of time:

Sometimes in our houses an old sword, a gorget, a cross of honor retain such memories [of our wars], and that’s our domestic folklore: but everyone knows how poor it is, and it becomes more impoverished each time death carries away the elders of the house. My father bequeathed to me a book where he recounted (according to old letters, registered documents, etc.) everything he knew about his family, and I am thus informed about my paternal ancestors back to 1680. But of the maternal line I don’t know anything, not even proper names, before 1830, and my children won’t even know how to reach 1830. In vain have I told them the old things I know: they confuse them or forget them. Preserve thus a little while a repository of anecdotes, soon reduced to insignificance, that’s all that oral tradition, left on its own, can do.134

The objects of Bédier’s “domestic folklore” are all military and implicitly colonizing, the memories they evoke bellicose. These memories are to the nation what genealogy is to the family—its source of identity. Even in writing, however, memory remains only partial: it only goes so far (1680, 1830), it holds only so many names (more fathers than mothers).135 Bédier seems to imagine, moreover, that the book won’t help his children. He himself can’t recall the name of a Revolutionary ancestor [je ne sais plus bien lequel].136 Even written documents furnish only fragile barriers to progressive amnesia. In the end, Bédier’s colonial “proof” of epics’ written origin carries as much contradiction as his story of the Mauritian jongleur in the Fabliaux. Both anecdotes reveal a concerted effort to maintain distinctions (oral versus written, French versus foreign) in the face of more nuanced realities. The analogy itself between medieval and colonial experience weakens the historical and cultural differences that Bédier so cherishes.

In presenting his father’s book, Bédier minimizes the role of oral transmission in order to emphasize the continuity guaranteed by written communication. While he later states that he “believes” [je crois] that parts of Adolphe’s book are based on “mere oral tradition,”137 in the Légendes épiques he suppresses the role of orality when he describes the book as based on “old letters, registered documents, etc.” Here, “etc.” covers for the damage of forgetfulness by affirming that all possible sources were used. Adolphe’s own narrative contains similar slippages between certainty and supposition. After stating that he has relied on papers, letters, and conversations with his parents, he writes only that he “believes” [je crois] that the oldest known family ancestor was compromised in the Cellamare conspiracy: this key founding event, repeated by Bédier and others with such assurance, turns out to be only conjecture. Adolphe goes on to state that the family’s aristocratic title is recorded “on an old parchment that can be found at one of the notaries of Bourbon,” but then admits that he is uncertain how to spell the noble name since he has never seen it written.138 Not only has Adolphe not seen the parchment, he does not name its actual location. His history thus relies far more on oral memory than on written materials. And since he wrote the book in Paris (far from any Bourbonnais notaries), it documents exilic imagination as much as colonial experience.

Adolphe’s book illustrates both the ephemeral nature of generational memory (supporting Bédier’s thesis in the Légendes épiques) and its vivifying effects (contradicting Bédier’s thesis). Adolphe characterizes memory itself as a desirable personal legacy:

Happy is he who has left such memories: he does not perish so long as those who knew him think of him and love him and establish thereby through death a kind of communication beyond the grave.139

Memory can establish continuity even in the face of death—an occult telepathy. Imagining his own death, Adolphe urges his children to seek “faith in immortality” by returning often to their memories and regrets.140 Bédier’s attitudes toward death and memory resemble Adolphe’s quite closely. He writes to Gaston Paris’s widow that “we live with our dead,” and to the Marquise that “our dead only survive in our memory: but there, at least, they truly live; it’s up to us that they have there a new and beautiful life.”141 Much later, Bédier vaunted the perfection of his own memory: he recalled every detail of his childhood on Bourbon:

And, by a strange spell, the painful things of my childhood and youth, even fever, even my crutch, even my old sorrows, seem to me penetrated with the same charm as my memories of happy events: all my joys, all my sufferings, equally purified, I cherish them with the same tenderness, at once distant and durable.142

Although the idealization of memory with age is a commonplace, in Bédier’s case it engages the larger forces of a national history that “forgets” its own foundational traumas, including the colonial violence that brought people like the Bédiers to Bourbon. The mutually reinforcing effects of memory and amnesia described here, moreover, drive the philological reasoning that underlies the Légendes épiques and Bédier’s entire approach to literary transmission.

Bédier’s discussion of his father’s book in the Légendes épiques ultimately bears witness to processes of oral and written transmission far more complex than he admits.143 Like the anecdote from aboard the Yarra told in the Fabliaux, Bédier’s colonial tale in the Légendes épiques reveals its own unreliability. The two anecdotes, moreover, contradict each other: while one illustrates how memory survives across centuries, the other suggests that memories die out quickly; whereas the Mauritian fabliau sustains a fantasy of permanent popular memory, the family book purports to show the power of writing. In both cases, however, Bédier presents his colonial sources as proof of the permanence of French culture since the Middle Ages. His apparent blindness to the fact that they also evidence impermanence and contingency reveals the powerful work of colonialist ideology. In the face of the disruptions wrought by geography, history, and experience, Bédier constructs a (romantic) vision of continuous national memory from the eleventh century to the present.

Bédier’s thesis inaugurated a defining split between scholars who accept epics’ written genesis (“individualists”) and those who argue for more diffuse processes of oral elaboration (“traditionalists” and “neotraditionalists”).144 While many specific details of Bédier’s pilgrimage thesis have appeared overly imaginative to subsequent scholars, the basic idea of decisive clerical intervention in the latter years of the eleventh century has remained an obligatory point of departure for discussions of origins. The broad effects of Bédier’s creole medievalism have thus contributed to shaping scholarly investigation of medieval literature. For Bédier, any modification of his theory meant “giving away” the Roland to the Germans:

I will not agree without good reasons that the epics are of Germanic origins, and, knowing nothing to support this hypothesis except reasons of force, I will only surrender our Song of Roland to the Germanics when the Germans will have already surrendered to the Scythes their Nibelungen.145

Bédier refuses to trace the nation beyond its written record—while positing the absolute stability of that record’s meaning once created. The very structure of dichotomy that Bédier posits—French versus German, written versus oral—derives from a conception of medieval literary history wholly dependent on modern national formations.

Bédier’s belief in impermeable categories rests partly on the ideology of continuity that underwrote creole desires for imperial privilege. By claiming to preserve French culture overseas, nationalist creoles invoked a “magic” fusion of continuity and rupture, intimacy and distance. A decolonized approach to epics requires not only looking beyond the question of origins but dismantling the work of categorization itself by pursuing, for example, the sophistication of oral literature and the generic “impurity” of the written tradition itself.146 Even when not explicitly engaged with colonialist and nationalist legacies, these kinds of approaches challenge their hold on the past, encouraging us to imagine a medieval world shaped by conceptions other than our own.

La chanson de Roland

Bédier put the arguments of the Légendes épiques into practice when he produced his edition of the Roland. He also extended to the epic the nationalizing methods used earlier on the Tristan materials (his publisher reissued Tristan et Iseut alongside the first Roland edition in 1922). Indeed, Bédier’s approach to France’s oldest epic combined the methods of all his previous scholarship: he critiques prevailing editorial methods, recuperates the Anglo–Norman dialect of the “best” surviving manuscript, and produces his own modern French version of the poem. The national, in this case, functions in tandem with the colonial because the Roland motivated Bédier’s desires for the Middle Ages even before he left Bourbon. It later became the vehicle for particularly strong declarations of faithful creole identity.

Having finished the manuscript of the Légendes épiques, Bédier seemed at a loss for a new topic for his courses. He wrote to the Marquise in September 1911, and again a year later, that he dreaded the beginning of the upcoming school years because he had no idea what to teach: in both cases, he ended up announcing courses on Roland, as he continued to do during and after the war.147 Indeed, students complained that even when the course titles promised other subjects, Bédier invariably returned to Roland.148 Eventually, he published five revised editions, no two alike, along with a volume of commentaries (1927) and several lengthy essays on editorial history (the last installment appeared in the same journal volume as his obituary).149 It is thus no exaggeration to characterize Bédier’s relationship with Roland as permanent and even obsessive—“an old love story.”150 Throughout his life, he worked in one way or another on France’s foundational epic. Returning repeatedly to Roland, he also returns to the past—his own and the nation’s. The poem’s double status as a personal and public archive enables imaginative conflations of temporal and geographical distances that respond to the homesick desires that philology was supposed to cure. Roland serves ultimately as both the source and destiny of Bédier’s creole medievalism.

Bédier’s relationship with Roland begins in Saint-Denis. There, in 1878, under the mango tree that shaded the garden of his family home, he read the new book given to him as a school prize.151 In terms of form, Bédier’s 1922 edition returns to this founding colonial scene. Whether Bédier’s prize was Gautier’s hefty luxury edition of 1872, or one of the later editions designed for schools (4th and 5th ed. 1875, 6th ed. 1876), his own Roland serves as a literal double: all are based on the Oxford manuscript (Bodleian Library, Digby 23); all present the medieval text and its modern translation on facing pages. Bédier’s editions of 1922 and 1924 reinforce the equivalency of Old and Modern French with a decorative border that links Bédier’s voice with that of the medieval poet: the border’s inner edges are open half circles, such that together the pages form a continuous line.152 Bédier’s actual translation also seems influenced by Gautier’s. Corbellari suggests, for example, that Bédier followed Gautier in not translating the term amirafle/amurafle.153 Most revealingly, Bédier initially perpetuated one of Gautier’s errors—translating “ewe” as “jument” [mare] instead of “cours d’eau” [stream].154 Bédier thus produces an edition that resembles very closely the book he himself first read on Bourbon.

Like other cases of colonial memory and philology, Bédier’s relation with Gautier’s editions turns out to be profoundly contradictory. On the one hand, he imitates Gautier and so authorizes his own formative reading experience. On the other, he often ignores Gautier’s early editions when discussing editorial history, referring almost exclusively to editions produced after 1880—when Gautier had substantially changed his editorial practice. In Gautier’s editions of the 1870s (one of which Bédier read in Saint-Denis), he treated the Oxford manuscript as the primary authority (with subordinate use of other manuscripts). These editions thus follow the same manuscript classification adopted by Bédier.155 As of 1880, however, Gautier gave equal weight to three manuscript families, producing the composite text that shaped students’ experience of Roland for generations.156 When Bédier published on Roland’s manuscript tradition in 1912, he wrote as if Gautier had only ever followed this “critical method”; he repeated this tendentious editorial historiography in his famous final study.157 Only in the much less influential Commentaires (1927) did Bédier engage the full scope of Gautier’s editorial practice (while still using the composite edition as his sole basis for specific comparisons).158 Overall, Bédier’s editorial historiography simplifies Gautier’s practice, occluding public recognition of his own debts to Gautier’s precedents.

The classification of Roland manuscripts lies at the center of nationalist conflicts over philology. The composite editions that Bédier so reviled followed German theories: Wendelin Foerster and Edmund Stengel gave the Oxford manuscript a “brother” and granted equal weight to two other families (Gautier cites Foerster as the inspiration for his change of method). Bédier’s arguments in favor of the Oxford manuscript’s unique authority and “preeminent dignity” displaced the Germans in favor of a purely French philology inaugurated by Francisque Michel’s “heroic” first transcription in 1837.159 Bédier writes as if Michel represented the only precedent for his work (his preservation of manuscript forms does resemble Bédier’s philosophy).160 By tying his arguments to Michel, Bédier sidelines another German philologist, Theodor Müller: his 1851 edition is actually based on Michel’s, such that his prioritization of the Oxford manuscript merely ratifies Michel’s foundational work.161 By overlooking the formative role of the early Gautier editions, which followed Müller, Bédier purifies his own editorial genealogy, reconnecting with an origin prior to the attentions of German philology. This nationalist argument simultaneously elides and recovers colonial memory: Bédier ignores the Gautier edition he read on Bourbon, while producing a text quite similar to Gautier’s. This double relation translates the recursive contradictions of creole ideology.

The dialect of the Oxford manuscript posed another kind of problem: how could a poem in Anglo-Norman represent French national origins? Previous editors had addressed this embarrassment according to their respective theories of origin: Gaston Paris amended the text to reflect the Francien dialect, while Gautier sought to restore the “purity” of the Norman dialect (liberating it from the “vices” and “dishonor” of Anglo–Norman “dust”).162 Bédier rejected these kinds of interventions while remaining as committed as Gautier and Paris to French origins. He “purifies” the poem’s form in several ways. First, the original French author guarantees basic unity.163 The poem’s universality and “classical” style (akin to Racine’s), moreover, transcend local particularities: the manuscript’s graphic forms do not in themselves orient the poem toward Britain.164 Bédier concedes that the manuscript offers a “pure” representation of the French spoken in England around 1170. This language, however, shares a number of dialectical features with the continental “French” of 1100, and so does not actually differ very much from the poem’s original language.165 Bédier reasons further that the poet himself could have used inconsistent forms: since we do not know what “rules” he followed, we cannot make the copyist responsible for everything that looks like an “error”; correcting anything could change the poet’s own words.166 In effect, Bédier argues that the manuscript’s Anglo-Norman may as well be French, for all anyone knows about the nature of French in 1100. While scrupulously preserving the “foreign” dialect, Bédier renders it utterly meaningless. His edition thereby claims a double continuity—simultaneously faithful to the lost original poem (in French) and the surviving manuscript (in Anglo-Norman).

Bédier’s editorial arguments reinforced the national recuperation of Roland begun by Lenient, Paris, and Gautier during the Franco-Prussian War of 1870 (see chapter 1). Bédier lends philological support, for example, to Lenient’s declaration that the Roland was first written “in true French.”167 And as I have mentioned previously, Bédier radicalized his predecessors’ anti-Germanism in reaction to new political pressures. As he began working on his edition in 1913, he wrote to the Marquise that it was his most urgent project because “it is deplorable that one reads [the poem] today in German editions, detestable besides.”168 His Roland could replace these “deplorable” editions (Müller, Stengel) as well as any number of French editions based on Stengel’s “corrupt” critical method. The Roland had of course been embroiled in national rivalries since its first publication in 1837.169 Bédier, for his part, expressed philological chauvinism from his earliest publications: in his 1894 discussion of the work of the Société des Anciens Textes Français, he declared it “humiliating” and “deplorable” that Germans were recovering French literature before the French.170 Bédier’s Roland thus reanimates the longstanding political dynamics of nationalist philology, renewing their visibility in the wake of the treaty (1919) that restored to France the lands it had lost to Germany in 1870.171

The ultimate destiny of Bédier’s edition is not only national but colonial: he sends the “restored” Roland back to Bourbon with a dedication that publicizes his creole identity: “A L’ILE BOURBON. DIIS PATRIIS. J.B.” This dedication captures the dualities of creole identity: the book is addressed to (A) Bourbon, in the same terms that would locate it on (A) Bourbon. The dedication thus deftly places “J.B.” both at a distance from the colony and within it, permitting a double gesture that reproduces Bédier’s own migration. In some editions, this colonial commemoration appears before the title, placing “Bourbon” literally prior to “Roland,” the colonial before the medieval.172 The very name Bourbon amplifies these temporal and geographic tensions, for it usurps the island’s republican name, Réunion, in order to recall the deposed monarchy that first claimed the island. Bourbon vivifies a dynastic memory that reaches all the way to the Middle Ages while resisting metropolitan impositions on local culture. As an expression of creole medievalism, the dedication denies the historical and geographic discontinuities that would necessitate a “ré-union.” “A l’île Bourbon” thus phrases a desire for unity while also containing personal and political ruptures.

The Latin phrase that completes the dedication introduces further multiplicities. Latin itself draws Bourbon into the prestige of ancient imperialism, establishing a cultural continuum from Rome to France to the Indian Ocean. To be “at home” in all of these times and places requires a transcendent sense of imperial nationalism. The actual phrase, “diis patriis” [beloved fatherland], points toward transcendence by naming a geographical site with a genealogical principle. Through lineage, Bourbon and France are both “ancestral lands” [la etite patrie, la grande patrie], Bédier a colonial migrant with two homes. These ambiguities turn the little island far away into an outsized concept always close at hand. This spatial and temporal conflation defines creole identity as a struggle to reconcile the fractures of a divided “home.” The risk of failure remains equally close at hand: for one metropolitan memorialist, Bédier’s “beloved Bourbon” signals his inveterate alienation: “This tropical atavism, the Parisian of recent vintage wanted to remember it in the dedication of one of his works.”173 From this perspective, Bédier shares fully in the migratory disorientations of the Réunionnais factory worker I cited at the beginning of this book. The shifting “time” of national belonging means that migrants never fully reach “France” or the civilized present. “J.B.” makes every effort to counter this threat of “savagery” with reassuring fictions of eternal unity.

Bédier’s Roland dedication emerged from a nexus of creole conversations that reveal the intensity of his desire to make creole identity a source of national cohesion, just like the epic itself. Bédier himself explained his choice in eloquent terms to his compatriots at their celebration for his Académie election, citing a more explicit version: “A L’ILE BOURBON DS PATRIIS. Aux Vertus de chez nous, Protectrices de nos Cases de Bardeaux” [To the Virtues of our homeland, Protectors of our Shingled Houses].174 Case references Bourbon’s “traditional” European architecture, a form both faithful to “old France” and adapted to the tropics. In a nearly contemporary letter thanking supporters on the island, Bédier underscores the word’s status as Creole rather than French: “mon la case,” set off in quotes.175 The case of the longer dedication thus anchors Bédier’s Roland in a creole landscape, where he remains always “at home,” protected by native spirits. He underscores the intimate relation between the epic and creole identity in his further explanation of the dedication:

I have put in this book and in the studies that accompany it or will accompany it, the best of myself; it seemed to me that I could allow myself to use it to render homage to our island, and that this was the only means I sought to show it something of my gratefulness.176

In a chivalric spirit, Bédier casts himself as supplicant to his homeland, his work on the epic the direct result of his creole formation. In both this speech and the letter, Bédier grounds his medievalism in faithful creole memory.

Later references to the dedication place Bédier in contact with the younger generation of creole writers intent on creating their own “colonial literature.” In one account, Bédier decided to dedicate his Roland to Bourbon while talking with the Leblonds:

When he published his Song of Roland, he said to the Leblonds: “I wonder how to demonstrate my love and my gratitude to my little country . . . Do you think they will be pleased there, if I write in the dedication something like: To Bourbon Island, my beloved country?”177

This anecdote, recounted by Hippolyte Foucque at the 1964 centenary celebration of Bédier’s birth, places Bédier in close relation with the “blond” architects of a creole theory of Celtic supremacy and chivalric racism (see chapter 3). The Leblonds’ possible role in Bédier’s dedication makes the book’s decorative cover more than ornamental (even if Bédier did not choose the designs) (Figure 34). These characteristically Celtic decorations deploy circular structures of continuity, where beginning and end turn upon each other in perpetual recursion—a self-enclosure that suggests isolation from outside influences. The medallion and the braid on the cover, along with the borders that run across the inside pages, suggest a racialized colonial frame for the epic’s portrayal of Christian Frankish victory. When philology and colonial Celticism combine to send Roland’s sanctified body from Roncevaux to Bourbon, Bédier’s creole medievalism reaches an imperial climax.

Bédier’s connection with the Leblonds went back more than two decades by the time he published Roland. He had family connections to both men: the Merlo family of Ary were distant cousins, and Marius’s uncle Lionel Potier was friendly with Bédier during his student days (they graduated the same year from the Lycée de Saint-Denis).178 Bédier was among the first of Marius’s diaspora acquaintances: when Marius arrived in Paris in 1896, he showed his first manuscript to Bédier, who reportedly encouraged him.179 Decades later, Bédier inscribed a copy of his acceptance speech to the Académie to both Marius and Ary, “très amicalement.”180 Just a few years after Roland appeared, the Leblonds thanked Bédier at length during a celebration for their novel Ulysse, cafre.181 In October 1931, the day after a party marking the closure of the Réunion pavilion at the colonial exposition, Bédier wrote to the Leblonds inviting them to visit him any morning at the Collège de France.182 A few years later, Bédier attended the celebration of Marius’s entry into the Légion d’Honneur.183 The Leblonds repeatedly cast Bédier as one of Réunion’s most important writers, as they sought to define colonial value through literature (see Introduction). Bédier’s dedication of Roland to Bourbon, then, ratifies the place in imperial culture created for him by the Leblonds.

Roland’s colonial connections extend to other members of the creole diaspora. Auguste Brunet (Réunion’s deputy beginning in 1924 and closely connected to the Leblonds) reported that Bédier sent a copy of Roland to his lycée, inscribed with a handwritten dedication to “the protective gods of our shingled houses” [les dieux protecteurs de nos cases de bardeaux].184 Echoing the phrase from Bédier’s speech to his compatriots on the occasion of his Académie election, this dedication once again evokes Creole and the tropical landscape, with the added implication of syncretic religious culture (“gods” in the plural). This possible recognition of Réunion’s multiculturalism offers a glimpse of the epic’s own potential for pluralistic readings (Bédier’s nationalist claims notwithstanding). This copy of Bédier’s edition would have arrived at the lycée around the same time as his photo signed in Latin and Creole and the letter in which he claimed to speak Creole better than French (see chapters 3 and 6). All these documents testify to durable colonial memory, and to multiplicities that contradict the homogenizing interpretations that Bédier imposes on the epic itself. This tension between national cohesion and colonial pluralism traverses all of Bédier’s medievalism, revealing once again the contradictory values that inhabit “creole.”

Some years after Bédier’s comments about his creole case, another Réunionnais partisan of empire, Jean D’Esme, remembered his own case in nearly identical terms.185 Even more important, D’Esme provided Foucque with the anecdote of Bédier’s childhood reading of Roland on Bourbon:

Jean D’Esme in fact once told me of the confidence that the professor at the Collège de France made to him one day: at a distribution of prizes at the lycée of Saint-Denis, he had received an edition of the Song of Roland (undoubtedly Gautier’s, which had appeared a few years earlier). That same evening he plunged into the reading of the beautiful poem and his admiring surprise was such that he would readily locate in that evening the awakening of his vocation for the Middle Ages.186

Foucque’s report implies that inherited creole chivalry needed only the catalyst of contact with the historical Middle Ages to develop into a nationally persuasive medievalism. D’Esme himself cultivated an image of chivalric imperialism. He was a prolific journalist, more widely read in the 1920s and 1930s than the Leblonds or the canonical poets; Brunet named him (along with Bédier) one of the most important representatives of creole culture. Along with numerous travel narratives and procolonial journal articles, D’Esme published Les chevaliers sans éperons [Knights without spurs] (1940), perpetuating the idea of chivalric colonialism with visions of heroic deeds in Mauritanie. His earlier successful novel, Les Dieux Rouges [The red godes], appeared serially in Bédier’s Revue de France (1921–22); Bédier also solicited him to write journalistic reports on the colonies. The two met soon after Bédier’s Académie election and around the time that he founded the review—that is, just as Bédier was completing his Roland. In a 1920 letter to Mareschal de Bièvre, Bédier says that he knows well the growing reputation of “Jean d’Esménard”: having been his father’s classmate and friend, he would be delighted to meet the son (D’Esme was married to Mareschal de Bièvre’s daughter Marie-Andrée).187 Bédier’s contacts with D’Esme attach Roland directly to the creole diaspora. They also suggest, alongside the Leblond conversations, that Bédier readily accepted his compatriots’ desires that he serve as an icon of creole prestige.

One last encounter solidifies Roland’s insertion in a network of creole memory—one that extends from Réunion to politically charged Alsace. In the months surrounding the completion of the edition, Bédier delivered a series of lectures (based on the Légendes épiques) to the newly repatriated students and faculty of the University of Strasbourg. At the time, Bédier’s student Gustave Cohen was among the newly appointed French professors. When Cohen later published his account of Bédier’s comments on the meaning of being a blond, blue-eyed creole, he noted in passing that Bédier initiated the reflection on creole identity “in regards to Maugain.”188 Gabriel Maugain, like Cohen, arrived at the University of Strasbourg in 1919; he became one of the founders of Italian studies and an important figure in the university through the 1940s.189 Most significantly for Bédier, Maugain shared a colonial heritage: prior to Bédier’s arrival in Strasbourg, Maugain invited him to tea as both a fellow Romanist and a fellow creole: “if I did not have the good fortune to be born on Réunion, at least I have Guadeloupe as my ‘little homeland.’”190 The meeting with Maugain apparently sparked Bédier’s desire to make visible the meaning of his blond hair as the sign of a racial identity tempered by imperial histories that stretched from the Indian Ocean to the Caribbean.

Bédier’s embrace of creole identity in Strasbourg, on decolonized soil redolent with memories of German occupation, recalls the political affinities between Réunionnais creoles and displaced Alsatians. The Leblonds had continued the connection forged by De Mahy in 1871, publishing essays in praise of Alsace-Lorraine before, during, and after the war.191 Bédier himself imbued his journey to Strasbourg with lyrically patriotic reflections inspired by the Alsatian landscape.192 On the pages of a calendar from 1918, Bédier writes of the faithfulness of the Alsatian people and of the emotional impact of Alsace’s return “for a Frenchman of my age.” Bédier casts the national reunion in the family metaphor so common in national discourse since 1870, writing of brothers and sisters “rediscovered” and of the “renewal” of France itself. This renewal harks all the way back to the nation’s origins as Bédier connects French patriotism in Alsace to Roland (in the process of being printed at that very moment): “love of France having become cult of the ancestors. terre majur.”193 Terre majur is Roland’s term for Charlemagne’s lands, a term whose meaning Bédier gives insistently as “land of the ancestors” (see chapter 5). Bédier thus grants Alsatian patriotism ancient medieval origins, securing the province’s rightful place in the French nation. In Alsace, moreover, terre majur also evokes Barrès’s “La Terre et les Morts” and Fustel de Coulanges’s “terre des ancêtres” (see chapters 3 and 5). Ancestor worship thus grounds national belonging in both Roland and modern Alsace.

Bédier elaborates on the transcendent nationalism of regional identities in further notes from his 1920 trip:

France, one and indivisible—what does that mean? It means Alsace for the Alsatians, like Brittany for the Bretons, like Béarn for the Béarnais, like Dauphiné for Dauphinois; at the same time all of France for the Dauphinois and Béarnais, all of France for the Normans, Flemings, all of France for the Alsatians, all of France for all children, so that they are all enriched by its diversity and by its love.194

These reflections reveal the creole’s deep sense of both regional and national integrity, and the primacy of emotional belonging over geographic limits (even though Bédier does not name any petites patries from overseas). The heartland of republican nationalism can be found anywhere. In Alsace, moreover, the survival of French patriotism proves that colonial occupation cannot deform French culture. And just as Alsatians preserve ideals born of the Middle Ages, Bédier’s “creole brother” Maugain serves as a living reminder of the great distances encompassed by the current “terre majur.” National, colonial, and historical unity coincide at tea in Maugain’s Alsatian home. And Bédier’s own Roland secures their ancient and enduring bonds.

Just a month after returning from Alsace, Bédier learned of his election to the Académie Française (June 1920). Among numerous congratulations, he received a letter from a gentleman who had attended his lectures in Strasbourg: he and his wife appreciated Bédier’s election all the more because he had enabled them to connect the epic to the “new nationalism.”195 They understood, in other words, that Roland and other epic poems could support modern patriotism and Alsace’s reintegration. Roland’s deep entanglements with Bédier’s colonial memories and creole compatriots lend the “new nationalism” an imperial dimension.

Bédier in fact intensified his claims on creole identity as he ascended to the pinnacle of national culture at the Académie: between his May 1920 trip to Strasbourg and the 1922 publication of Roland, he repeatedly defined the meaning of creole—to Cohen (May 1920), to Mareschal de Bièvre (October 1920), to Maurice des Rieux (October 1920), to the Académie (November 1921), to the Réunionnais of Paris (February 1922), to the Réunionnais of Saint-Denis (photo and edition, 1921–22). During this entire period, the Roland dedicated to Bourbon worked its way toward the public. Inspired by colonial memory, it emerged amid a network of creole associations activated by Bédier as he defined his public persona as a new Academician.

Bédier’s creole memories integrate Bourbon into a nationalist vision directly supported by Roland. His literary and philological arguments purified French culture, projecting the nation’s continuous prestige all the way back to the eleventh century and across the globe to Bourbon. The result is a double cultural translation that turns the epic into a representation of the “other” congenial to Bédier’s own notion of “greater France”—a unified entity that originated in the eleventh century and endured in the twentieth, undisturbed by geographic, ethnic, or temporal ruptures. From the perspective of Bourbon, Bédier’s famous summary of the Légendes épiques takes on new resonance: “In the beginning was the road [route].”196 For Bédier, migration generated France’s first “rooted” text. His own journey to epic origins began en route from Paris to Bourbon—and also from Bourbon to Paris. The beginning, then, is a state of colonial suspension between two homes, the route a destiny and a history. This journey with two beginnings shapes Roland—which itself makes a double journey to and from the colony. Bédier’s nearly continuous work on this text is part of a lifelong effort to resolve the tensions of a colonial identity split between two homelands.

Bédier’s Roland oriented scholarly and popular understanding of the epic for much of the twentieth century. The edition remained widely available into the 1980s, serving as a durable reference despite the many other editions produced since 1922.197 Indeed, as with Tristan et Iseut, I read it myself as a student; its dedication to the enigmatic “Bourbon” (when I finally noticed it) launched this book—inspiring new contacts between the legacies of medievalism and colonialism. Even when students and scholars have used other editions, they have felt Bédier’s pervasive influence: most editors have focused exclusively on the Oxford manuscript, with ever greater respect for its exact forms. Arguably, even Cesare Segre’s influential revival of the stemmatic approach to the manuscript corpus (1971, 1989) remains fundamentally Bédierist: Segre relies on the Müller stemma that Bédier defended, changing the Oxford text much less than the stemmatic method authorizes; in his second edition he relegates interpolated passages to the notes.198

The prevalence of Bédier’s ideas in Roland’s twentieth-century editorial history reveals the “sympathetic” chord touched by what Ian Short calls Bédier’s “unarticulated political imperatives.”199 As I have shown, these imperatives were not so unarticulated after all. Bédier’s theories of authorship and language support not only the French nation, but the general conceptions of cohesion that underlie most forms of national belonging. In Bédier’s case, these conceptions owe a great deal to colonial experience. Bédier’s edition speaks to colonial formations in some very specific ways, from his first reading to his self-fashioning as a creole subject in the 1920s to his enduring reputation as a creole medievalist.200 For all of these reasons, creole medievalism has circulated broadly in French medieval studies since 1922.

Today, the Oxford Roland seems liberated from the strictures of Bédier’s conclusions. In France, the French translation of Segre’s “critical” edition has become the new authority (required for the agrégation of 2004); students not trained in Old French are likely to read Ian Short’s version in the Lettres gothiques series, which introduces numerous changes to the Oxford text in an effort to account for scribal knowledge of oral composition;201 scholars now have access to the full corpus of French-language Roland manuscripts through the new editions overseen by Joseph Duggan. Duggan’s project shifts editorial focus from questions of origins and authorship to the dynamics of reception and rewriting. The new editions encourage attention to the Oxford manuscript’s twelfth-century context, to the significance of manuscript artifacts themselves, and to later medieval thinking on the renewable significance of Roland’s story.202 They thus help move French medieval studies beyond the imperial pressures that have shaped so much Roland scholarship. They can also facilitate a decolonization of the epic, since their very plurality challenges Bédier’s privileging of the Oxford manuscript, as well as his theories of aristocratic authorship and linguistic purity (themselves informed by colonial racialism and classism). Formal changes, of course, do not in themselves sever scholarship or teaching from nationalist habits—nor has Bédier’s text precluded resistant readings, as the rich and varied body of twentieth-century scholarship attests. As I argue in the next chapter, Roland can speak to imperial ideologies, while also revealing fissures within ideology itself.

Bourbon grounds Bédier’s valuation of pure forms, despite the contradictions that traverse many of his colonial memories. While he defined France as an ancient repository of superior aesthetic genius, creole propagandists defined Bourbon as the faithful guardian of France’s ancient traditions. Bédier’s notion of literary history thus appeals to creole longings for national integration by incorporating historical and geographic ruptures into a seamless imperial whole. Through Bédier’s vibrant sympathy with the Middle Ages, distant origins cede ground to perennial continuities: as he declares in one of his late publications: “I have no sense of discontinuity.”203 A creole migrant of multiple origins, Bédier developed literary histories that excluded the recovery of lost origins: he defended multiple origins for the fabliaux, individual creativity as the beginning of both romance and epic, and “best manuscript” editions that eschewed reconstruction. Thus he endeavored to resolve his own fractured relationship with “greater France” into a permanent belonging, undisturbed by colonial migrations and the sense of “foreignness at home.” In this process, the periphery becomes the center’s echo, even its foundation. The persistence of Bédier’s recuperative effort, however, suggests that integration remains incomplete, the dislocations wrought by migration unresolved. Despite Bédier’s declared desire to forget Bourbon, creole memories consistently haunt his philology as he lives the homesickness of exile. These memories shaped his Middle Ages, leaving deep imprints on French literary history.