SOURCE: Haidu, Peter. “Funerary Rituals in the Chanson de Roland.” In Continuations: Essays on Medieval French Literature and Language: In Honor of John L. Grigsby, edited by Norris J. Lacy and Gloria Torrini-Roblin, pp. 187-202. Birmingham: Summa Publications, Inc., 1989.

[In the following essay, Haidu explores how Charles breaks with tradition in his reaction to Roland's death.]

The Baligant episode interrupts, in its various segments, narrative developments whose coherence is defined by a common theme: not only do they come after Roland's death, these narrative syntagms are concerned with problems posed by that death. Death is a problem for the living, as is well known, not the dead. Anthropologically speaking, the crucial problem is the reintegration of the members of the social group into its unity after the disappearance of one of its crucial members. Vis-à-vis the fact of death, the social group performs its rituals, the means by which the absentification of the dead is acknowledged and the re-presentification of the living achieved. The values of the group define the individual who dies as a hero, and his actions as the idealized narrative representation of at least one ideological potential in the group's semiotic universe. Issues of responsibility and culpability, problems of the figuration of value in a mode transcendent of the quotidian, the radical incommensurability between the survivors and the human gap facing them, are social and ideological fissures obsessing the survivors, not those who are gone. The group's task is to knit again the web of its social unity, to reintegrate its members into the collectivity.

While portraying the generality of the surviving Frankish troops as grieving their losses, the narrative focuses on the figure of Charles as survivor. It is Charles who is cast into the textual figures of the consciousness of loss. The rhetorical charge of this aspect of the figure is not to be denied. It is perhaps that aspect of the Roland-text that remains the most accessible to the poetic taste and values of the typical reader of our time: an academic reader, haunted by the nostalgia for an imagined past of wholeness and integration. The continued effectivity of this emotional charge, however, must not be allowed to mask the fact that the poem is here performing an essential operation. The section in question changes things, it performs a transformation upon those materials that, at its beginning, are givens because they are the product of the earlier sections of the poem. In doing so, of course, it functions like any narrative syntagm: it constitutes a transformation of its narrative materials.

The essential Moments of this transformation are three: the lament for the dead, the revenge upon the retreating Saracens, and the discovery, collection, evisceration, and burial of the heroes' bodies. These three moments imply successive locative displacements: Roncevaux, the River Ibre, Roncevaux again. They also are constituted by different types of textualizations: as implied by the noun that designates it, the lament is a purely verbal textualization; the revenge, though it includes some dialogue, is primarily a narrative moment; the complex “burial” of the heroes is composed of a particular combination of the verbal and the narrative.

The lament takes the form, somewhat surprising, of a classical topos in an oral text: it is the first use of the ubi-sunt theme in vernacular French poetry. The appearance of this poetic theme from the classical and written tradition in the oral text of the Chanson presents clear evidence of the mixed character of the Chanson de Roland. It is “mixed” not only in its ontological status, but also in its socio-historical and ideological content. Its form as a rhetorical question—Where are the dead?—is “empty” in the sense that the question is not a real one: the answer is not open to question. In this case, the inevitable answer to the rhetorical question is, after each naming of a hero, that he is dead, dead on the field of battle […]¹

Although “empty” in this sense, the lamentation in the form of ubi sunt does perform essential narrative structural roles. It presents itself as the expression of emotions, a theme textualized a number of times in this syntagm. This textualization is a necessary precondition of the following narrative program. It is only once the emotions of Charles and his men have been given socio-textual existence that they provide the content of a qualifying statement justifying the revenge obtained against the fleeing Saracens. Duke Naimes intervenes upon the expression of the emperor's great grief, pointing to the fleeing Saracens: “Look, two leagues ahead of us, See the great dusty roads, There is the mass of the pagans: Ride! Avenge this pain!” (vv. 2425-28). Vengez ceste dulor!² provides the link between the emotion and the succeeding action. The grief is experienced as an impermissible trespass upon the survivor, a damage done him for which legal reparation can be claimed. That reparation is the basic narrative stuff of the chanson de geste as a whole, in particular that which is known as the rebellious barons cycle: there, revenge is taken upon the feudal lord whose treatment of the vassal is unjust. Here, it is revenge against those who by definition have no moral or juridical status before a strictly Christian law, the nonbelievers. In any case, the survivor experiences his grief as an injury done him, for which the appropriate mode of treatment is the exaction of vengeance, conceived as a legal mode of compensation. The expression of grief here transforms what a twentieth-century consciousness would consider pure interiority into a social and semiotic fact that is the basis for narrative action.

That narrative action is collective: “tuit en sunt cummunel” (v. 2446). In fact, it is so collective that it requires the intervention of God in a new mode. In the Roland-text, God intervenes by saintly interposition, either cognitively (as in the dreams brought to Charles by the angel Gabriel) or pragmatically (to collect Roland's soul as his ultimate Destinator). Here, however, God is presented as directly interfering with the operations of Nature: at Charles's prayer, God stops the flow of time to allow the Franks to catch up with the Saracens. It is thus a miracle, a type of grant of a magic Adjuvant that had not been awarded Roland. Charles is endowed by the text with the power not only to receive privileged communication from God in the form of proleptic dreams—a power he already had in the Roland-section, and which he retains in this second major half of the Chanson: Charles will once again be visited by annunciatory dreams. His communicative competence with the divinity extends to the point of being able to request and obtain a change in the fundamental assumptions, not only of human life, but of narrative itself. The suspension of the flow of time, the ability to perform narrative actions that would otherwise be prevented by something as fundamental as the passage of time, changes the very conditions of possibility of narrative itself. What would be impossible for any normal/human/actor becomes possible for Charles. The miraculous interference of the divinity in the workings of Nature foreshadows the perhaps equally miraculous solution that will be achieved in the final trial scene of the epic.

At this occasion, “religion” makes a striking entry into a text that is anything but religious or Christian in its fundamental content. In view of this limited import, the form this religiosity takes should be noted. It constitutes a direct grant of power to Charles as King and Emperor, utterly disregarding (as the poem does in general) the historical institutions of religion: the medieval Church in any of its forms. It may well be that the Pope and Emperor are mutually defining binary opposites in the historical, extra-textual domain: in this concrete text, the Pope and the institution he heads simply do not exist. All the ideological power associated with the Pope and the Church in contemporary cultural codes is attributed to the figure of the King of France who is also the Emperor of Christendom. Within the economy of the text, this “extra” ideological valorization is of the same order and function as the Baligant episode: the extraordinary investment of ideological value in this particular figure is functionalized by the need to enable him to perform certain narrative acts that might otherwise prove too difficult for an ordinary human agent. The character of these acts, the specific isotopy on which they will occur is indicated by the third Moment.

Having despatched the fleeing Saracens by drowning and by the sword, Charles and his men return to Roncevaux. This a/b/a structure marks the revenge wrought upon the Saracens as necessary on the one hand, but as a necessary interruption of the basic business of this section, namely the enactment of the rituals of the dead. Before their performance, however, the weary Franks take their rest and sleep for the night. Charles lies down as well, but does not sleep immediately:

Carles se gist, mais doel ad de Rollant
E d'Oliver li peseit mult forment,
Des.XII. pers e de la franceise gent
Qu'en Rencesvals ad laiset morz sanglenz.
Ne poet muer n'en plurt e nes dement
E priet Deu qu'as anmes seit guarent.

(vv. 2513-18)

As at the moment of Roland's death, when he surrenders his glove to God as a sign of final and ultimate fealty, so here as well the relation of the individual to his God is figured by the feudal Metaphor, the divine assumption of dead souls and hope for their salvation being figured by the relationship of feudal protection […]

When Charles, wearied by grief, does fall asleep, it is to encounter annunciatory dreams that parallel those of the first part. In terms of the twentieth-century reader, these proleptic dreams “make sense,” that is, carry specific meaning, only on the basis of a retrospective reading: one must have read the Chanson to the end at least once before in order to understand that these two successive dreams refer first to the impending battle with Baligant, and secondly to the trial scene at Aix, with an allegorical representation or pro-presentation of Ganelon, his thirty relatives, and both Thierry and Pinabel. Reading, for us, is a punctual, datable event, whose cognitive function is clearly delimitable. But the cognitive conditions of the medieval audience were quite different from ours. For the medieval audience, there is likely never to have been a “first” time: the Chanson de Roland, from what we know of it, was a popular text, and is likely to have been performed by jongleurs repeatedly, albeit in different forms. As a result, the events of the poem have an identity under the specific variations characteristic of any individual performance: they exist in some field of permanence external to the particular performance. Functionally speaking, they exist in a sphere that is analogous to the neo-Platonism typical of medieval aesthetics descending from Philo and St. Augustine. As a result, the meaning of these proleptic dreams, definable for the modern reader only on a linear dimension of temporality, are extra-temporal for the medieval audience of the oral performance and immediately accessible. That audience, even illiterate, will always already have known the story and its outcome: even though this was obviously not true in any one particular case (even a medieval individual biography had to contain a “first” time hearing of the text), it is nevertheless an ontological condition of the oral medieval text that its basic narrative pattern has to be construed as always already known.

Thus it is that the proleptic dreams that come to Charles, ambiguous as they are if taken in a unidirectional successive temporal sequence that permits them to be semantically informed only by a retrospective look upon the narrative, must be considered as endowed with the semantic content that, for the modern interpretant, devolves from a forward look to the later narrative events. The first dream sets Charles in a battle accompanied by multiple “natural” signs of terror—thunder, winds, and ice, storms and tempests, fires and flames falling upon his army; lances and shield take fire, weapons and armor twist, and the army is in great distress. It is attacked by bears, leopards, serpents, vipers, dragons, and demons—a plethora of the supernatural, to which are added thirty million griffins! The battle is general, until Charles himself is attacked by a great lion filled with rage, pride, and courage. There is little reason to hesitate in identifying the proleptic narrative referent of this laisse as the later battle between the Frankish army, led by Charles, and the troops of Baligant, a battle that will find its conclusion in the individual combat between the two emperors. The second dream, localized at Aix, sets into the dream scene one chained bear, thirty more claiming to be his relatives and demanding that he be returned to them. At this point, a hunting dog arrives, a greyhound to be exact, who attacks the largest of the thirty bears. The emperor witnesses a miraculous combat, but, as in the preceding dream, does not know in his dream who wins. Again, there is little reason to hesitate in juxtaposing this dream with the combat between Pinabel, representing the thirty relatives of Ganelon at his trial who are also his hostages, and Thierry, the lean and unheroic knight who will, in fact, win against Pinabel.

The structural import of these dream syntagms is not their allegorical significance, which is so patent as not to require “interpretation” in any serious sense at all, but rather their function within a narrative structure. They occur at the hinge point between two larger syntagms: the narrative of the funereal rituals of the survivors, and the beginning of the Baligant episode. Whether they are to be considered a part of the former or of the latter is not an issue for the present reading, which accepts the presence of the Baligant material in spite of a conviction of its later and additive nature. For these proleptic dreams, and the meaning they bear and which is demonstrated to Charles by the angel Gabriel who brings him the dreams themselves (senefiance l' en demustrat mult gref, v. 2531), those dreams rejoin a general semantic level of the text. They demonstrate the connection of the present narrative moment with a later one, by “re-telling” the later and narratively “real” event ahead of time, before its actual place in the outplay of the narrative. The establishment of the semantic connection produces the meaning that both the funereal rituals and the Baligant episode have to be seen in relation to the later syntagms they announce. In particular, the funereal rituals and the reintegration they operate, lead to the ultimate narrative sequence of Ganelon's trial. They lead to it, in that they perform a function that is prerequisite to the later narrative syntagm. The reintegration of the living must be performed before the final judgment of the cause and responsibility for the dead can be passed. Ganelon's trial could not occur unless its actors retrieve their full standing as Subjects within the social group.

After the Baligant episode stages the encounter of Baligant's messengers and Marsile, and then between Baligant himself and his vassal, the text shifts back to Charles at Roncevaux. The third Moment of the funereal syntagm is the essential one. It is constituted by a compound of verbal and narrative subsections. In a battlefield covered by the twenty thousand bodies of the rear-guard, the problem is that of finding the one body Charles wants above all, Roland's. He finds it by a speech that recalls and re-presentifies the extraordinarily moving moments of Roland's death. At a solemn festival observance at Aix, Roland had once boasted that he would not die in a foreign land unless he had gone farther forward than his men and his peers; he would have his head turned toward the foreigners' land, and the hero would end as a victor. In order to find his nephew, Charles restates that aspect of him that led to his death, the combination of heroic virtue and boastful supererogation of knightly prowess that we “mean” by the name “Roland.” Charles recalls the traits that made the subordinate knightly group so problematical for their lords and for the society at large.

In other words, the mere recall of the actor's words, at a party sometime before the current war, also implies, because of the earlier narrative context, the political isotopy, which is doubled, here, by the identification of the familial bond between Charles and Roland (v. 2859, 2870, 2876). The political isotopy is reiterated. Finding the body of his nephew among the red flowers and green grass, under the two trees, near the marks of the blows Roland struck on the rocks in trying to destroy Durendal his sword before dying, Charles faints dead away. Coming to, he begins to “regret” Roland, that is, to mourn him in lamentation. At this level the verbal syntagm is identical with the earlier plainte in the ubi sunt form. But in looking more specifically at the content of this particular lament, a major difference is found. If the ubi sunt theme looks backward and stresses the absence of the dead, Charles's lamentation now does quite the opposite. The absence of the dead is presented in a proleptic framework, looking forward to a future point in time and presentifying the effects that Roland's death and absence will have then. Recommending Roland to God's mercy, Charles touches on his uniqueness as a warrior and concludes the first brief speech:

“La meie honor est turnet en declin.”

(v. 2890)

Literally: my “Honor” is turned toward its decline. But “honor” is a strange word. What is restricted to the valorial field for us moderns—“honor” for us is strictly a matter of opinion, either internal or that which others hold of an individual—is simultaneously material in the medieval language. “Honor” can indeed have the same meaning it has for us, the credit or esteem in which a given individual is held by his fellows; but it also has the meaning of the complex of land and social organization that produces wealth and power and that often takes the specific form of a fief. Indeed, the ambiguity of the word (an ambiguity that is apparent only to us, since we—unlike the medievals—separate the material base from its ideological value) suggests the interdependence of the two semantic components: one is not “honorable” unless one has a certain material base of power and wealth; and possessing such a base produces the credit and esteem of one's fellows, unless one acts in such a way as to lose that respect. In certain ways, the Middle Ages were far more materialistic than we are; or perhaps their materialism was simply more self-evident. Charles's honor is then both his reputation and that complex of geographical territory and socio-political organization upon which that reputation is based: his fief, his kingdom, his empire. All three are profoundly weakened by Roland's death […]

The necessary dualism of funereal ritual—“letting go” and reaffirming the unity of the social group, assertions of absence and presence as simulaneities—is thus laid over, by Charles's own complaint, with the specifically political form that is the ultimate isotopy of the chanson de geste in general. That dualism is restated in the most primitive moment of the ritual, in the burial of the great majority of the dead, the extraction of the hearts of the heroes—Roland, Oliver, and Turpin—and their enclosure in marble sarcophagi, well washed with spices and wine. The hearts are to retain for the survivors the strength of the dead, even as their (ultimate) burial is to recognize their departure. Presence and absence both, the anthropological ritual has a complexity that the political does not. Where the funereal ritualization of loss affirms a duality against the brutal fact of death, the political isotopy allies itself with that brutality. There is nothing in this plainte about the inspirational value of the legend and chanson to be told and sung about Roland and his peers. In Charles's political evaluation of the meaning of Roland's loss, there is room only for loss, for the resulting political and military weakness of his reign. His personal grief is doubled by a political issue that overrides the personal. Indeed, the personal must be expressed, indulged in perhaps, in order to allow for the movement toward the political. For the personal loss is not resolved, but met, by the funerary rituals of the text; the political loss, and its implications, must yet be met and resolved, more directly and more brutally.

Finally, the syntagm of funereal ritual has as narrative function the (re)statement of this dual valorization of the loss of the hero, the personal and the political. It does so with an internal organization of closure. I have pointed out that the first Roncevaux syntagm is verbal, the Ebre syntagm is narrative, and that the second Roncevaux syntagm is both verbal and narrative. That second Roncevaux syntagm, however, is internally organized in a pattern that is recursive of the overall organization of the ritual syntagm. Its first “move” is to discover Roland: as we have seen, that takes a predominantly verbal form as Charles recounts Roland's earlier boast at Aix. That is followed by another verbal syntagm, Charles's long lamentation. These two verbal syntagms are followed by the concluding syntagm, narrative in nature, in which the dead are buried, the heroes have their hearts extracted and are both shrouded and encased in sarcophagi. This recursivity, recapitulating in the final narrative subsection the pattern of the whole, thereby gives closure to the whole. That sense of closure marks the fact that one moment of the narrative has ended, another begins.

The major burden of most of the text after the conclusion of the Baligant episode is to deal finally with the issue of culpability for Roland's death, as well as the socio-political development of a way to prevent the recurrence of what led to the disaster at Roncevaux. In the Oxford Roland, the syntagm concerned with these issues—Ganelon's trial—is preceded by a brief, two-laisse passage that returns to the isotopy of the funereal rituals. It also initiates another isotopy, that of exchange: the brief episode is a narrative isotopic connector between the funereal rites and the trial.

The emperor has returned to Aix. Aude, Roland's betrothed, comes to the court to ask for her man. She asks, and the manner in which she asks, and the manner in which her demand is received, indicate that what she asks for is her right. Charles acknowledges Roland's death; he offers instead his son Louis as substitute. The text asserts the advantageous character of this offer: Louis would inherit Charles's reign. Nevertheless, Aude rejects the offer, and does so radically: she dies on the spot. Her corpse is turned over to nuns in a convent, who bury her properly, alongside an altar:

Mult grant honur i ad li reis dunee.

(v. 3733)

The syntax suggests what the text does not specify, that the “honor” is financial value.

The syntagm is framed by two lines that seem to announce something else. The two laisses of the Aude syntagm are laisses 268 and 269. The last line of laisse 267—preceding the Aude syntagm—announces a different topic:

Des ore cumencet le plait de Guenelun.

(v. 3704)

Then, in laisse 270, the theme of the return to Aix is restated (v. 3734) and developed, and in the following laisse, the beginning of Ganelon's trial is announced once again:

Des ore cumencet le plait et les noveles
De Guenelen, ki traïsun ad faite.

(v. 3747 f.)

Thus, both before and after the Aude syntagm, Ganelon's trial is announced. On the one hand, it is patent that the Aude syntagm is not Ganelon's trial, or any part thereof. Nevertheless, the repeated assertion of the text suggests a connection between Aude's death and Ganelon's trial.

The likeliest explanation, for both the discrepancy between the topic announced in v. 3704 and the succeeding syntagm, and the isotopic continuity between the earlier section dealing with the funeral rituals of the dead of Roncevaux and the Aude scene at Aix, is fairly obvious. In an earlier stage of the Chanson de Roland, the episode of funereal ritual and the death of Aude were two subsidiary parts of the same narrative syntagm, whose topic was the collective and individual griefs, and the rituals of reintegration and exclusion that follow upon the death of the culture-hero, the icon of the warrior society. It is a syntagm whose topic is the reconstitution of the collectivity after a disaster that calls its very values into question. The present state of the evolution of the oral epic contains an additional syntagm, inserted between two subsidiary parts of the earlier stage, in which an outsider and his troops come and interrupt the collective ritual of grief and reconstitution. The Baligant episode, in other words, interrupts a narrative unit that had a unified production of meaning dealing with social emotions after a great disaster, intervening between the narration of the great defeat itself and the legal and political fall-out that succeeds it. The new episode interrupts the sequence: (loss) + (mourning) + (adjustment), each of these terms occurring on both the individual and the collective isotopies. The present syntagm adds new elements to the textual production of signification; it also introduces a new cultural and valorial logic of coherence into the world of textuality. Insofar as social contracts determine the exchange of social value, Charles does more than merely discharge his responsibility as a social agent: his offer contains a valorial supplement, a surplus value that might have been expected to determine, even to overdetermine, an acceptance by the Recipient of the communicative exchange. The surplus value here is the difference between the value of a great noble, even a culture hero, and the king's son. More determining than cultural values, the political value and the implicit territorial and financial values are expected to assuage the damage to her own value that Aude has undergone in her grief and the loss of her betrothed. Aude's refusal is all the more striking then, since it not only rejects the proposed exchange and the supplementary surplus value it contains: in her death, she also refuses any further social intercourse and verbal exchange with the Destinator.

To the modern reader, the notion of the substitution Charles proposes seems absurd: how shall one man be accepted as the substitute for another in a romantic relationship that consists precisely in the election of one individual as the unique cathexis of certain emotions? Here again, twentieth-century codes are inappropriate. Marriage, far from an expression of the romantic election of another person as the primary cathexis of the individual and the announcement to the collectivity of the couple's desire to make their union permanent (perhaps, at this point in 1989, already a somewhat old-fashioned view of marriage!), marriage in the Middle Ages, at the social level with which we are concerned, is a dynastic and diplomatic affair generally more concerned with the aggrandizement of dominion over lands and other forms of property than with tender emotions. Not that tender emotions do not exist in this text, or in the epic in general: they do, and this particular syntagm is only one example of such emotions in the Chanson de Roland. Others are the not infrequent references to wives and women who wait for the warriors in dulce France, to whom they are eager to return; the equivalent is stated of the Saracen warriors. The text recognizes such attachments as normative and refers to them without hesitation. But the text is a warrior epic, and such topics are not the topics that are developed within its frame; in this respect, the Old French epic is poorer than the Greek. On the other hand, while the Greek epic may be more inclusive, the French is both more intense, and perhaps accomplishes social transformations that remain unthought of in the Iliad. While the tender emotions do exist at the edges of the Roland, however, they are not the basis of the politics of marriage, nor are they a major concern of the text. The theme of individual fidelity is touchingly deployed, here, but it masks another operation that is of more moment to the achievement of this text.

The two laisses that deal with Aude are more than a romantic interlude (the way they have most frequently been taken by traditional criticism), and more than a textual supplement incorporated into the text itself. They have a function; they perform a transformation. First of all, the narrative syntagm provides an isotopic connector, incorporating the theme of death, and of Roland's death in particular, within a textual space that also places the action at Aix, where the trial takes place. It does more than merely state these two themes of death and the trial simultaneously, however. The funerary theme, as we saw, was meant to achieve the reintegration of society as a whole after a crucial loss, and in particular the reintegration of the primary grievant, Charles. That reintegration is not achieved yet, however. Charles's effort to set things right, as feudal overlord, as monarch, as uncle, is unsuccessful: the issue of compensation for Roland's death is still unsettled.

That issue is as foreign to the modern reader as that of the values actually involved in noble marriage. Whether because of a religious heritage that can be taken to assert the sanctity of human life, or because of a tradition of individual subjectivity that represents each person as a unique phenomenon, our ideology does not allow for the principle of compensation to be recognized overtly in cases of death. Both the Judaic tradition and the Germanic are more sensible. Acknowledging that perhaps there are emotional components in such a loss for which compensation is literally unimaginable, both traditions recognize that the dead person was also a social value, and that compensation for that aspect of the person can be arranged. That is to say that a system of exchange, in which the social being of the individual is recognized as a social value for which other forms of social value can be exchanged, is elaborated in both traditions. It is probably the Germanic tradition that was most determinative in shaping medieval attitudes. The tradition of a fixed tariff of equivalencies—the wergeld—of compensation even in the case of the unemendable morth (murder), was a characteristic of archaic German law that continued into the period of the Chanson de Roland. Such legal compensation is an extension of the principle of exchange into the domain of law, even as law governs the most extreme disruptions of social relations. The principle of exchange is at work, then, in Charles's offer, which is neither unfeeling nor crude: it is a recognition both of the reality of loss and of the limited nature of the compensation that society, as an entity, is capable of providing its members. The offer is an entirely appropriate enactment of social and conventional codes and bears with it a surplus of valorial supplementation that signifies generosity or, in medieval terms, largesse.

Nevertheless, the offer is absurd, even in Charles's own terms. Charles's offer refers to Roland as an hume mort. In a line already quoted, he says:

Jo t'en durai mult esforcet eschange.

(v. 3714)

Aude's response, by its very rhyme within an assonanced text, stresses that it is the notion of exchange she is rejecting:

Alde respunt: “Cest mot mei est estrange”

(v. 3717)

What is estrange (L. extraneum; foreign; the term belongs to the same semantic family as the Greek barbaroi) is exchange. Aude takes the position that no exchange is possible for Roland: it is an anomalous position in the codes of her society. It is also a costly position to take: it costs her her life. The beloved's death implies an absolute cathexis on one individual. What may strike the post-romantic reader as a literary cliché is, in the textual and cultural context of the end of the eleventh and the beginning of the twelfth century, quite remarkable. That anomalous character, rather than the romantic associations, is what is textually most functional.

Oddly enough, Aude's refusal of the principle of exchange in connection with Roland had been prefigured in our text, and prefigured by none other than Charles himself. It was Charles who, speaking of Roland to Naimes, said earlier:

“Deus! se jol pert, ja n'en avrai escange.”

(v. 840)

It was Charles himself, then, who established the principle of nonsubstitutability in Roland's case, well before he offered a substitute for Roland to Aude. Aude's refusal is merely a reiteration to Charles of Charles's own understanding: it is his own cognition that comes back to haunt him. No man can be a substitute for Roland. Roland was unique, but as soon as that is said, it is crucial to specify what that adjective comports, lest our conceptions of the uniqueness of the individual flood into the receptacle of the waiting signifier.

Roland's uniqueness, at the actorial level, is specified so frequently that it is hardly necessary to cite those bits of the text in question. It is military, and hence political. Not only is there no trace of a unique interiority in Roland—one might make a better argument for Oliver, Charles, or Ganelon in this regard—his uniqueness is entirely constituted by the social qualities that are those of his social class carried to extremes: both his warring abilities, and their characterological implications, are those of the knight as a social type. Simply, he has a far greater allotment of the more specific traits than the ordinary knight. In terms of the inner-outer dichotomy, Roland's uniqueness, like his individualism, is entirely “external.” What counts is not some hidden and unique subjectivity, but the fact that, as a Subject, Roland is capable of undertaking narrative programs—and carrying them out successfully—that no one else can.

Aude's refusal of Charles's offer and the implication of Roland's uniqueness that it bears have as their meaning a reassertion of the momentous loss that is Roland's death. It reasserts that the issues implicit in that death, and in the narrative syntagms that lead up to it, have not yet been resolved, especially as they attach themselves to Charles's person. Not only is the sequence of funerary rituals undergone ineffective in reintegrating Charles as survivor into the social group: insofar as Aude is in her rights in demanding of Charles her betrothed, his inability to produce Roland causes her death. […]

Not only is the ideal hero of the society dead […]: the basic principle of social organization—that of exchange—has been at least interrupted and suspended. This is the ultimate significance of Aude's refusal. If the normal pattern of exchanges encoded in the laws and conventions of the society no longer hold, if the damage to the social fabric is so grievous that its system of compensatory awards is refused by those whom it should benefit, then the very principle of sociality has been suspended. What is at stake is not a romantic attachment, nor even the justice to be accorded heroism, betrayal, and contractual responsibility: because of the characteristics of the textual actors involved, what is at stake is the continuation of society. […] The stakes that are set into play by the Chanson de Roland—as by any great work, from the Iliad and the Greek tragedians to Samuel Beckett—are the ultimate values of the society in which it is embedded. That “setting into play” is also a “setting at risk”: each time textuality “plays” with the values of its social structure, it not only takes the risks of an aesthetic adventure, it also risks the survival of the social text, the social fabric, the social body.

Notes

1. Major excisions from my longer essay are indicated by ellipsis marks within square brackets.

2. All citations are from Joseph Bédier's edition of the Chanson de Roland (Paris: Piazza, 1921).

The present text, excerpted from a longer discussion of the Chanson de Roland, seems apt in honoring the memory of a colleague who was unique in his particular combination of respect for the concrete text, attention to an inherited tradition of medievalism, and attentiveness to the paradigms of contemporary theory.