publisher colophon
  • Irony and Cognitive Empathy in Chrétien de Troyes's Gettier Problem

Irony often separates a work's author (or reader) from its characters. But irony can also engage our universal, adaptive capacity for perspective taking. In his twelfth-century romance Cligès, Chrétien de Troyes creates such an irony. Though medieval, it takes the form of a Gettier problem, a thought experiment from twentieth-century epistemology. Read closely, this irony can pump our intuition that we share the characters' ignorance. This intuition allows us to empathize with them, not emotionally but cognitively, taking on their worldview and realizing that we know things, or not, just as they do.

The relations of comparison and contrast among the viewpoints of characters and between the viewpoints of authors and characters is one of the most important dimensions of meaning in literary texts. . . . It is for this reason that the analysis of irony, as a central tonal medium for registering differences in point of view, occupies a position of singular importance in the interpretation of literary meaning.

—Joseph Carroll, Literary Darwinism1 [End Page 169]

Irony stings. Among friends it might be playful, its target joining in the fun. But it is hard to see irony figuring into any how-to guide for making friends in the first place. Indeed, the word is absent from Dale Carnegie's How to Win Friends and Influence People. More often irony is divisive, driving a wedge between ironist and target. The ironist is hip, aloof, cool; the target, mired in some mistake or other, is not their equal.

But must irony always drive us apart? Or can it inspire empathy? Not just sympathy for those at the receiving end of its sting—poor souls! Not some "there but for the grace of God go I," which smacks of sanctimoniousness. Can irony inspire instead a recognition that "there go I"? The twelfth-century French author Chrétien de Troyes achieves such a recognition by making not only his characters but also his readers the butt of his irony. The example explored here comes from Chrétien's second Arthurian romance, Cligès, and concerns whether we know what we think we know. When the question of knowledge is raised in literature, we might expect the author to give us a situation in which we know more than the characters. This would be a typical, divisive irony, allowing us readers to share in the author's knowledge while the characters remain ignorant. In Cligès, however, Chrétien forces us to share in his characters' ignorance, creating through irony a kind of cognitive empathy, throwing us all in the same epistemic boat: "there go we."

Curiously, my example comes from medieval literature—curious because for some time people thought of the Middle Ages as a pious, sincere era immune to irony. When scholars in the mid-twentieth century recuperated the Middle Ages as an irony-laden age, perhaps overzealously, they did so not by recognizing irony as a human universal but by making medieval authors the equals of the great, aloof ironists of modernity. How sophisticated, how modern, how . . . worthy of study! But the sophistication disinterred has too often been the easy one in which medieval authors draw fine distinctions between themselves (with their readers) and their characters. And so when, nearly half a century ago, Peter Haidu cemented the critical association of irony with Cligès, his book's title highlighted his medieval author's aloofness: Aesthetic Distance in Chrétien de Troyes.2

One irony in Cligès, however, at lines 2048–89, serves as a humbling reminder of the humanity we are supposed to share with the welldeveloped characters of great literature. This particular irony had, until recently, escaped critical notice.3 In it, some of the romance's characters find themselves in a situation akin to a Gettier problem in philosophy in which an individual may have a justified true belief and yet still lack [End Page 170] knowledge. Intuitively, justifiably believing something that is true seems sufficient to say we know it.

Suffice it to say, for now, that Gettier problems challenge such an intuition by showing how we might not really know what we think we know even when it's true. Chrétien puts his characters into a Gettier problem: they don't really know what they think they know, even though it's true. As readers, however, we know that they don't know. This is a typical, divisive irony, maintaining our "aesthetic distance" from the characters. But then, as I demonstrate below, Chrétien does the same to us: just as we are laughing at the expense of his characters, we come to realize that we actually don't know what we thought we knew.

Explaining just how this happens restores Chrétien's intended irony to his text and allows us to see, pace Haidu, that not all of his ironies create an "aesthetic distance." Instead, Chrétien achieves what Kenneth Burke long ago described as "true irony," which breaks our bias of epistemic egocentrism.4 Rather than distant readers who can only sympathize with the romance's characters, we are thrown together with them, left to empathize and reflect on our shared fallibility.5 Whereas Haidu maintains "that the reader never shares the ignorance which betrays the characters,"6 the romance's manuscript tradition and critical reception (including its editions and translations) prove otherwise. Scribes, editors, translators, and critics—all of whom are readers like us—have missed the true irony of this passage, and, as a result, they have introduced variants into the text or failed to correct them, further obfuscating Chrétien's intention. Such variants betray an ignorance shared by character and reader alike, and the claim that "the reader never shares the ignorance which betrays the characters" becomes itself unintentionally ironic in its arrogance. Restoring Chrétien's intended irony reintroduces a certain humility to both Chrétien's irony and our readings of his texts.


In 1969, the world of analytic philosophy was shaken to its epistemological core, or so the story goes, by the publication of a mere two-and a-half-page essay.7 Its author, Edmund Gettier, proposed two scenarios in which an individual can have a justified true belief without having knowledge. A somewhat unnuanced history of philosophy might say that knowledge had been defined as justified true belief at least since Plato, who puts just such a definition into the mouth of Theaetetus, who in turn had heard it from somebody else.8 While Socrates initially seems to [End Page 171] favor this definition of knowledge as "true judgment with an account," which is to say knowledge as justified true belief, by the end of this aporetic dialogue both he and Theaetetus abandon it; Socrates himself calls it a wind-egg "not worthy of bringing up" (210a–b). Nevertheless, as the centuries became millennia, despite the aporia already present at its origin, the standard definition of knowledge as justified true belief was thought to be secure. And so in 1969, Gettier did more than just add a footnote to Plato: he forced a rewrite of the entire chapter on epistemology by overturning the standard account of knowledge.

A Gettier problem invented by Alvin Goldman gives us an idea of what is going on in the passage from Cligès.9 Although every detail in any narrative deployed in analytic philosophy can play a role in the argument (the narrative's "knob settings," as Douglas Hofstadter and Daniel Dennett call them10), the following summary should meet our needs: A man—here named Henry—is driving in the countryside, sees a barn, and believes it to be a barn; indeed, it is a barn. Henry then claims, "That's a barn." At this point, we are likely to say that Henry knows it's a barn. Goldman wonderfully qualifies this point by adding, "so long as we were not in a certain philosophical frame of mind" ("DPK," p. 772). Even if the definition of knowledge as justified true belief fails to satisfy the rigorous requirements of philosophy, it may nevertheless capture our intuitive, "folk" understanding of what knowledge is.

Goldman, however, is not done with his tale: "Unknown to Henry, the district he has just entered is full of papier-mâché facsimiles of barns. These facsimiles look from the road exactly like barns" ("DPK," p. 773). Again, prior to this last bit of information, we would say that Henry knows that the barn is a barn. But knowing that he would have believed a facsimile to be a barn, we now "would be strongly inclined to withdraw the claim that Henry knows the object is a barn" (p. 773). As Robert Fogelin has pointed out, an element of "epistemic luck" makes us "refuse to acknowledge that his grounds establish the truth of the claim."11 At the start of Goldman's narrative Henry seems to know that what he sees is a barn, but by the end we realize that he does not, despite his believing, correctly, that it is a barn. Our intuitions flip from "Henry knows it's a barn" to "Henry doesn't know it's a barn," and so the traditional account of knowledge now seems wrong or at least incomplete. [End Page 172]


The episode at lines 2048–89 of Chrétien de Troyes's Cligès is structured like such a Gettier problem. In this passage, some knights come to believe that their lord and his cohort have been killed. Their belief is justified and, in part, true, but true only by luck. Chrétien uses this plot twist to further his keen, ironic view of human claims to knowledge, which may remain fallible even when their underlying belief is true. Literature here shows once again its power to contain insights only later stripped down and laid bare by philosophy. But readers have missed this irony, perhaps because of the complexity of a literary text, as opposed to the clarity of a no-less-carefully crafted philosophical thought experiment. Chrétien's presentation of the Gettier problem is less straightforward than Goldman's intuitive version and mingles justified true beliefs with justified false beliefs.

Compounding this complexity is the instability typical of medieval texts. Due to the vagaries of scribes, tastes, and even the elements, the content of a medieval text may change from manuscript to manuscript or indeed become lost altogether. Nevertheless, even though only seven manuscripts of the complete text of Cligès are known to us (with five others containing fragments), and even though these manuscripts don't agree on some important details, we can reconstruct a Gettier problem from the text. To do so, however, we must set the scene.

The first half of Cligès tells the story not of its eponymous hero but of his father, the Greek prince Alexander. The romance opens with Alexander deciding to leave his home at Constantinople and head west with a dozen men to King Arthur's court. Once there, Alexander and his twelve companions find themselves fighting for King Arthur against a rebellious baron, Angrès, who had been left in charge of Britain while Arthur was in Brittany. Arthur, now aided by his Greek guests, returns to Britain and lays siege to London, where Angrès has dug in. The rebel army attempts a night attack on Arthur's camp but, after a bloody battle, they are defeated and rush back to the protection of their castle and its keep.

At this point, the romance risks stalling, returning to the status quo ante of a prolonged siege. But Alexander, in a ruse worthy of a Greek hero, has six of his men disguise themselves by swapping shields and lances with those stripped from dead enemies killed during the night attack. Alexander's deception works: the disguised group penetrates the enemy castle keep and defeats Angrès. [End Page 173]

But the deception works too well, for it fools not only Angrès's sentries but also some of Alexander's own men who were not part of the scheme:

The Greeks who were still outside [the castle] knew nothing of [Alexander's victory]. In the morning, after the battle was over, they found their companions' shields lying among the dead and wrongly presumed them to have been slain. When they recognized their lord's shield, the Greeks were in such anguish over his loss that they fell in a faint upon his shield, proclaiming that they had lived too long. Cornix and Nerius fainted, and on recovering regretted they were still alive. From Torin's and Acoriondes' eyes there flowed a torrent of tears down over their breasts; their lives and their happiness seem detestable to them. And more than all the others, Parmenides tore at his hair and pulled it out. These five grieved more deeply for their lord than can be imagined. But their grief was groundless: instead of Alexander, whom they thought they were bearing off, they had another. The other shields, which they thought marked the bodies of their companions, likewise caused them great sorrow. They wept and fell in a faint upon them; but they were deceived by these shields, too, for only one of their companions, Neriolis, had been slain. They might rightly have borne off his body, had they known; but they were in as much distress for these others as for him. So they took them all and bore them away, though they were mistaken about all but one. The shields made them take appearance for reality, like a man who dreams and takes a lie for the truth. By the shields they were deceived.

(C, lines 2048–89)12

There are two ironies in this passage. Haidu identified the first, calling attention to the grieving Greeks' confusion of "reality and illusion," which renders their grief "meaningless" (AD, p. 87). They mourn Alexander and his six men, who are not dead on the field but still in the castle. This is a typical, disjunctive irony that creates an "aesthetic distance" between the characters, who hold justified false beliefs, and us readers, who know the truth. Here, five of Alexander's men believe that he and their other companions are dead because their shields lie among the corpses on the battlefield, and the author lets us know that they are wrong: "their grief was groundless." The five Greeks are thus the targets of the author's use of situational irony, and their paroxysms of grief become comedic to us because we know what they do not.

But calling their grief "meaningless," as Haidu does, is a mistake. Recognizing the meaningfulness of their grief leads us to see a second irony at work here, one not disinterred by Haidu. Although six shields and lances belong to their companions who are in the castle, at least [End Page 174] one set does not, for Neriolis is truly dead, and his armored body is among those over whom they are grieving (C, line 2078). Given that they grieve truly for a dead companion on the field, we are forced to reconsider our reaction to the first irony.

Had the five companions come across solely Neriolis's shield and lance next to a dead body in armor, they would have formed a justified true belief that he was dead, and we would not hesitate to say that they knew he was dead. Today, we might think there is a problem of justification here, for surely they should have looked more closely before making such a heavy inference. But their belief is justified by the standard means of identifying a body on a late-twelfth-century battlefield. The romance itself establishes such means of identification: Alexander makes this clear at lines 1827–28, when he equates his men's identifying markings (their cognizances, conuissances) with shields and lances. In this context, shields and lances provide all the epistemic justification one needs for knowledge.

By coming across both Neriolis's equipment, which signifies truly, and the equipment of the companions who are part of the ruse, the five grieving Greeks stand among the cognizances like Henry among the barns: just as Henry would have believed the facsimiles to be real barns, the five mourners do believe the switched lances and shields to be signifiers of their dead friends. Their grief for their living lord and companions has the same justification as their grief for Neriolis. The former grief, though justified, is false; the latter belief, also justified, is true. We come to realize that only a certain "epistemic luck" distinguishes these cases: they cannot, then, know that Neriolis is dead. Chrétien's use of the imperfect subjunctive at line 2080 makes this clear: "They might rightly have borne off his body, had they known (se le voir an seüssent)." They do not know the truth (le voir) that they believe.


Chrétien almost certainly did not write this passage to challenge any formal justified true belief theory of knowledge. He would not, at least, have read the Theaetetus. Nevertheless, his irony here is in keeping with the irony we often find at work in philosophical thought experiments like Gettier problems. One of the most effective ways of defeating an argument is to derive from it either a formal contradiction (a reductio ad absurdum) or an intuition that it just cannot be so. We may ask our readers to temporarily adopt a belief, or we point out that they already [End Page 175] hold a certain belief, and then we construct a narrative intended to solicit some intuitive response. Dennett has described such "intuition pumps" as "little stories designed to provoke a heartfelt, table-thumping intuition—'Yes, of course, it has to be so!13 In what we could call an "ironic thought experiment," that intuition gets expressed as, "No, it cannot be so!" Ironic thought experiments thereby produce what Fogelin has identified as the "correcting judgment" characteristic of all irony.14

Because thought experiments can be ironic, we might expect them to have analogues in literary texts. Once found, such analogous texts do not have to be mere curiosities, but can have critical, pedagogical, or even philosophical importance. Literary examples of philosophical thought experiments may show that an author is engaging in contemporaneous philosophical debates. In the Middle Ages, when strict divisions between disciplines were still inchoate, philosophical inquiry took place in many different forms, including literature. Literary thought experiments may also provide more memorable examples to students learning the argument in a philosophy class. Most interesting, perhaps, they may even provide new complications, new knobs to turn on the thought experiment, advancing—and not just exemplifying—philosophical analysis.

So what, then, is Chrétien's literary contribution to the Gettier thought experiment? If his point is not to challenge a medieval version of the justified true belief theory of knowledge, then so what? After all, we could point to Plautus's Menaechmi as an earlier example, or to that play's Shakespearean version, The Comedy of Errors, for an even messier jumble of justified true and false beliefs. Chrétien's poetic genius takes us instead in a different direction. Rather than doubling characters to multiply mistaken identities, Chrétien intuits the general structure of Gettier problems and gives us not one, but two. We have seen the first: the five grieving Greeks on the battlefield justifiably believe that Neriolis is dead without knowing that he is dead. To see the second, we must consider what makes us not aesthetically distant from the characters but cognitively and epistemically similar to them.


As we read along, just before we are told that Neriolis is dead—before, that is, the Gettier reveal—we might have asked, why only five Greeks? Remembering that Alexander has brought twelve men, we could realize that, if five are grieving on the field and six are in the keep, then one is missing in action. Before we learn what happened to Neriolis, we might have realized we have no account of him. In theory, we should have been keeping track of Alexander's cohort as follows: [End Page 176] Reading carefully, we should have already recognized a problem before the big reveal. Chrétien's use of numbers is precise everywhere in Cligès, and so, when he names only five Greeks, he is also careful to match that list with the proper count: "These five grieve for their lord" ("Cist .v. font duel de lor seignor," line 2067). There is no way to get around it: someone is missing. We know six companions are in the castle; we know that these six have left their cognizances on the field next to dead bodies; we know that five companions are grieving. But 6 + 5 = 11, not 12. Even if we are not "in a certain philosophical frame of mind," we should sense something is amiss.

Alexander's Twelve Companions
Name Status Cognizances
1 Licoridés in the castle on the field
2 Nabunal in the castle on the field
3 Ferolin in the castle on the field
4 Francagel in the castle on the field
5 Pinabel in the castle on the field
6 Calcedor in the castle on the field
7 Cornix grieving with him
8 Nerius grieving with him
9 Torin grieving with him
10 Acoriondes grieving with him
11 Parmenides grieving with him
12 Neriolis ? ?

As readers we have no way to get around this absence. We might have passed over it, but the hole is there if we look closely. Provided, that is, that we are reading Chrétien's text as he intends it. The failure of his readers to appreciate the ins and outs of his irony in this passage, however, may result from changes to the text of Cligès made by medieval scribes (and sometimes maintained by modern editors). At line 2067, for example, the scribe of manuscript BnF fr. 1420 writes that there are ".vi.," six Greeks who mourn their lost lord, despite having faithfully copied only five names (C, lines 2059–65). Scribal errors are not uncommon, and perhaps this was a typical scribal slip, not meaning anything in particular. Note, however, that it is an error of addition rather than omission: the scribe had to add a mark to give us ".vi." (six) instead of ".v." (five). It is as though the scribe had done the math and realized that there should be six grieving Greeks to match the six in the castle and so keep the total to twelve companions.

Copying out the text word by word or line by line, the scribe may very well have not yet read on, transcribing the text before learning that Neriolis is dead. Before knowing the fate of this twelfth man, any reader would recognize the disjunctive irony that the mourning Greeks are weeping meaninglessly for those in the castle. If their grief truly is meaningless, and since six companions are already accounted for in the castle, the other six must be here on the field grieving even if [End Page 177] only five are named. Perhaps the sixth griever on the field isn't worth naming because he just wasn't doing anything worth writing about. He neither faints, like Cornix or Nerius, nor cries copiously, like Torin or Acoriondes, nor tears out his hair, like Parmenides. Still, he must be there, mustn't he? Perhaps, rather than a slip of the quill, this scribe has made an active, though mistaken, change to the text. By not recognizing Chrétien's mathematical clue that someone is missing, the scribe may have prematurely grouped Neriolis with the other five and so come to see only the first irony of mere mistaken belief at work in this passage—a medieval instance of the inattention to the other irony in this passage also shown by contemporary scholars.

Still, only when Chrétien finally reveals that Neriolis is dead do we understand the significance of this missing companion. Only then do we realize why we could not have completed the above table. At that point, we should recognize that the grief on display is not entirely meaningless. We should, but this is not what readers have done. We have noted how Haidu calls the grief "meaningless." But he is only following a long line of readers inattentive to the significance of Neriolis's death. Again, we can see this inattention in the text's manuscript tradition, for when Chrétien completes the twist of this Gettier problem, we are met with still more scribal errors: Neriolis has trouble not just being counted but even being named!

Although Neriolis is his name in the translation quoted above, my reference edition actually reads "Leriolis" at line 2078. Indeed, five of the seven manuscripts that contain this passage give some variant of Leriolis. The others give either Ricreolis or Nereolis (an acceptable variant for Neriolis given the instability of Old French spelling). Something intractable seems to hover about Neriolis, raising fundamental questions about how we read this romance and its irony. The actual name Chrétien intends matters only a little. What matters more is that four of the seven manuscripts—not to mention at least one modern edition15—are inconsistent and give a name that does not match the name they give earlier when the Greeks are first introduced. Surely Chrétien intended for his character to have the same name throughout the text. Like the scribal counting error, we are left to wonder why scribes would be inconsistent in their naming (let alone why modern editors and translators would follow them in the error).

A reader of any version of Cligès where the names do not match would have to ask: Who is this thirteenth man? Where did Leriolis come from, or where did Neriolis go (or vice versa, depending on the manuscript [End Page 178] or edition)? Notice that this discrepancy has no bearing on the Gettier problem the characters find themselves in, for the dead individual, whatever his name, is still clearly a companion whom they justifiably believe to be dead. Still, the mismatched naming is symptomatic of readers' inattention to these characters, which makes the complex and multiple ironies of this passage ever harder to see. Although the Gettier irony of this passage survives such inattention, critics may have missed it because they, like the medieval scribes, have not paid enough attention to the romance's characters. What wonder, then, that an "aesthetic distance" separates us from the characters if we do not care who or how many they are! We begin to see that any such aesthetic distance is, at least in part, our own creation. It is also contrary to the author's aim of creating cognitive empathy between reader and character.

As we have seen, Chrétien precedes his Gettier problem with a more typical irony that indeed creates an "aesthetic distance" because we know things the characters do not. And, as Haidu notes throughout his study, such a situation normally does not inspire much empathy. But our experience as readers, of moving from the first irony of this passage to its second, just might. We make that transition either prepared by Chrétien's subtle mathematical hint or totally unawares. Chrétien seems to have intended the former, preparing us for the Gettier moment. The latter would be more like Goldman's philosophical version; but from a literary perspective, having the rug pulled out from under us like that is somewhat cheap. In great works of art, such a plot twist should be prepared for, not seem like the non sequitur punch line of a shaggydog story.

However we encounter the plot twist of Neriolis's death, it provides important nuance to the idea of aesthetic distance, for while we now know more than the characters do, we begin to understand how they can view the world as they do. True, we do not experience any emotional empathy: we do not grieve their friends with them. Nevertheless, we can begin to adopt their epistemic point of view. We can experience a kind of cognitive empathy, a perspective taking in which we do not just understand that the grieving Greeks are mistaken but also experience how they come to be so mistaken.

We are not able to do this for the first, disjunctive irony based on their justified false belief that their lord Alexander is dead; we know ahead of time that this grief is meaningless. But the situation is different with Neriolis. We know that he is dead—Chrétien has told us so, and he is no unreliable narrator. Our belief is true and justified by the [End Page 179] text. But shouldn't we have known or at least suspected that Neriolis is dead before Chrétien reveals this to us? If we missed the mathematical hint, our superior knowledge seems ill gotten, a bit too lucky, in a way similar to the epistemic luck of a Gettier problem. We can either continue to ignore Chrétien's hints and blame him for a cheap trick, or we can retrospectively recognize our ignorance and understand how the grieving Greeks fell into their Gettier problem. Either way, a sense of communion with his characters should start to grow. Moreover, when we see our fellow readers—medieval scribes and modern editors and translators alike—getting the numbers or the names wrong, we might sense that maybe we too are just as fallible as the characters: "there go we."

But Chrétien is not done with us yet. His final innovation is to put us readers in our own Gettier problem.


The Gettier problem we have been looking at—involving the justified true belief that Neriolis is dead—relies upon the justified false belief that Alexander and his six companions in the castle's keep are also dead. The five Greeks mourn meaninglessly for these seven because they are alive and well. But a reader of Cligès should have known that this is not quite so, for earlier in the story (C, lines 1886–91) we read: "Count Angrés charged Calcedor and struck him such a blow upon his golden shield that he knocked him to the ground dead for all to see. Alexander was sorely distressed to see his companion slain" (AR, p. 146). Although Calcedor's dead body is not present on the battlefield like Neriolis's, his cognizances are, and they do truly refer to a dead companion: [End Page 180] The five grieving Greeks would be mistaken about the identity of the body in armor lying next to Calcedor's shield and lance, but their belief that he is dead, justified by these cognizances, would still be true. Chrétien has thus given us two Gettier problems, one for Neriolis and one for Calcedor, each of which has unique knob settings involving the nature of justification and reference. Exploring such knob settings further would, perhaps, allow literature to advance philosophy by exposing new variations on the Gettier problem.

Alexander's Twelve Companions
Name Status Cognizances
1 Licoridés alive in the castle on the field
2 Nabunal alive in the castle on the field
3 Ferolin alive in the castle on the field
4 Francagel alive in the castle on the field
5 Pinabel alive in the castle on the field
6 Calcedor dead in the castle on the field
7 Cornix grieving with him
8 Nerius grieving with him
9 Torin grieving with him
10 Acoriondes grieving with him
11 Parmenides grieving with him
12 Neriolis dead on the field on the field

In any event, Chrétien has preceded the Neriolis-Gettier problem with a Calcedor-Gettier problem, and yet we somehow miss it. I certainly did until the Neriolis incident forced me to pay more attention to Chrétien's use of names and numbers, going back to see what else I missed. If you have not read Cligès, you might feel that I am being unfair in assuming that you too would have missed this earlier Gettier problem. Perhaps, but here again I can rely both on the absence of modern critical discussion of this textual problem and on evidence from the manuscript tradition. I know other readers too have missed it, since at line 1888, where Calcedor's death is mentioned, the manuscripts are divided, with some saying that a certain Macedor has been killed. Again we find a textual inconsistency because Macedor is a new name, not present in the original list of companions. While some modern editions and translations emend the text to make sure there are no supernumerary Greeks, not all do.16 Just as the trouble over Neriolis's name does not annul his Gettier problem, whether it is Calcedor or Macedor who is killed does not matter, for we still see another companion who is truly dead. The five grieving Greeks in the field thus have a second justified true belief: they believe truly both that Neriolis is dead and that Calcedor is dead. But because these true beliefs are as justified as their false beliefs about Alexander and the others, these two justified true beliefs fail to count as knowledge.

This doubling poses a problem with Chrétien's Gettier problem: Why kill two Greeks? Calcedor's death was sufficient to motivate the Gettierproblem irony. The five grieving Greeks outside the castle could have been joined by a very much alive Neriolis to mourn Alexander and their six other companions. Attentive readers would have known that this belief is both false with respect to Alexander and five others but true with respect to Calcedor. Or, similarly, we could ask: Why kill Calcedor? Neriolis's death would have been sufficient. Did Chrétien forget about Calcedor, just as his copiers, editors, translators, and critics would later do? Surely, even Chrétien nods [End Page 181] .

Chrétien's motivation for the Neriolis-Gettier problem seems clear enough. Putting Neriolis's dead body on the battlefield is akin to placing a real barn smack dab among facsimiles. Doing so offers a clear Gettier problem, unburdened by the complications of reference found in the Calcedor case. Goldman likewise created his story about Henry to avoid some of the potentially objectionable moves that had been made in Gettier's original story about a man named Smith, such as responsibly inferring a truth from a nevertheless false premise. Similar objections might be made to the Calcedor version: since his body is not present, his cognizances signify differently than Neriolis's do. Moreover, while the five Greeks have a justified true belief about Calcedor and truly grieve his death, they carry away the wrong body—a more typical irony of mistaken identity. Neriolis instead affords them both true beliefs and true actions: they grieve Neriolis and correctly, though unknowingly, carry his body away.

The difficulty we have as readers of Chrétien's Cligès—whether as critic, translator, editor, or even scribe—in coming to terms with his Gettier problems has ultimately left us in a position similar to those five mourning Greeks. We may come to believe that there is a deeper irony in this passage, beyond the simple difference between our knowledge and the characters' knowledge. But do we truly know that there is such an irony, given the luck of the manuscript tradition? Or, more to the point, do we truly know that there is irony in the text, given the luck of Chrétien giving us the Neriolis-Gettier problem, which allows us to go back and realize that a Calcedor-Gettier problem was there the whole time? The characters are lucky in their grief, and we are lucky in our recognition of irony. These are the beginning intimations that we experience ignorance very much like the characters do. Stepping back for one final look reveals that they are our true equals.

Sure, Chrétien is having a bit of fun with his characters and is letting us in on it. Their paroxysms of grief on the battlefield are comically exaggerated. But Chrétien is having a bit of fun with us as well. We stand among Chrétien's ironies like Henry among the barns or like the Greeks among the cognizances. We identify a disjunctive irony, and indeed it is there. We think we are looking down knowingly on the characters in their ignorance. But to either side are true ironies, facsimiles of the disjunctive one, that reveal our ignorance as well. So: imagine now that Henry is an avid reader of Chrétien de Troyes. He picks up Cligès and comes to the passage at lines 2048–89. He then declares: "I know the irony here: their grief is meaningless." He has a good laugh, as well he [End Page 182] should. But he keeps reading, not realizing that, by missing Chrétien's Gettier problems, he has created one himself. As Haidu writes: "Things are rarely what they seem in Cligès" (AD, p. 82). We see now that this holds for character and reader alike. Perhaps, then, a better response to the text is not aesthetic distance, but cognitive empathy.

Brian J. Reilly
Fordham University


1. Joseph Carroll, Literary Darwinism: Evolution, Human Nature, and Literature (New York: Routledge, 2004), p. 150.

2. Peter Haidu, Aesthetic Distance in Chrétien de Troyes: Irony and Comedy in "Cligès" and "Perceval" (Geneva: Droz, 1968); hereafter abbreviated AD.

3. Brian J. Reilly, "Chrétien de Troyes's Gettier Problem: Cligès Lines 2048–89," Romance Notes 53, no. 1 (2013): 21–25. My references to Cligès and its variants come from Chrétien de Troyes, Cligès, ed. Stewart Gregory and Claude Luttrell (Cambridge: D. S. Brewer, 1993); hereafter abbreviated C. Translations are mine unless noted otherwise.

4. Kenneth Burke, "Four Master Tropes," The Kenyon Review 3, no. 4 (1941): 421–38.

5. Compare with Haidu: "The ironic incongruity . . . is especially gentle and sympathetic here, without for a moment allowing us to move from sympathy to empathy" (AD, pp. 43–44). Haidu insists throughout his text that we only ever sympathize, never empathize, with the characters. He means emotionally here, but it is also clear that he thinks we never move to their epistemic position either. See note 6 below.

6. AD, p. 87.

7. Edmund L. Gettier, "Is Justified True Belief Knowledge?" Analysis 23, no. 6 (1963): 121–23.

8. Plato, Theaetetus, trans. M. J. Levett, rev. Myles Burnyeat (Indianapolis: Hackett, 1992).

9. Alvin I. Goldman, "Discrimination and Perceptual Knowledge," The Journal of Philosophy 73, no. 20 (1976): 771–91; hereafter abbreviated "DPK."

10. Douglas R. Hofstadter and Daniel C. Dennett, The Mind's I: Fantasies and Reflections on Self and Soul (1981; repr., New York: Basic Books, 2000), pp. 375–77.

11. Robert J. Fogelin, Pyrrhonian Reflections on Knowledge and Justification (New York: Oxford University Press, 1994), pp. 23, 26.

12. Chrétien de Troyes, Arthurian Romances, trans. William W. Kibler (New York: Penguin Books, 1991), p. 148; hereafter abbreviated AR.

13. Daniel C. Dennett, Intuition Pumps (New York: Norton, 2013), pp. 5–6.

14. Robert J. Fogelin, Figuratively Speaking, 2nd ed. (New York: Oxford University Press, 2011), p. 13.

15. See Chrétien de Troyes, Cligès, ed. Philippe Walter, in Œuvres complètes, ed. Daniel Poirion (Paris: Gallimard, 1994).

16. See Chrétien de Troyes, Cligès, ed. Charles Méla and Olivier Collet (Paris: Librairie Générale Française, 1994); ed. Alexandre Micha (Paris: Champion, 1957); trans. Ruth Harwood Cline (Athens: University of Georgia Press, 2000); trans. Burton Raffel (New Haven: Yale University Press, 1997).

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.