Bryn Mawr Classical Review 04.06.03


Henry Ansgar Kelly, Ideas and Forms of Tragedy from Aristotle to the Middle Ages. Cambridge: Cambridge University Press, 1993. Pp. xvii + 257. $54.95. ISBN 0-521-43184-0.


Reviewed by Eugene Vance, University of Washington.

Henry Ansgar Kelly has provided a well-informed study of the history of the concept of tragedy in European culture, both classical and vernacular, from Aristotle to the time of Italian humanism. The result of an enormous exploration of sources in six languages, here is a book for which a need has existed for a long time. Kelly has not written a history of "tragedy" as drama per se; rather, the semantic history of a term. Its title is a bit misleading, then: until the very last pages, at no point doe s either Kelly or his sources ever pinpoint a cogent concept of tragedy as a "form" that might be apprehended and described, nor do the works that Kelly describe in the final chapter make or break any clearcut norms or rules.

Kelly starts with the Greeks, though with Aristotle and not Plato. I find this decision odd, since so many of Plato's criticisms of tragedy and of dramatic art in the Republic become the basis of hostility toward theater in the Christian era. Platonic notions of mimesis are at the heart, for example, of Augustine's rejection of dramatic art in the Soliloquia and tragedy as a cathartic experience in the Confessions. Not being, himself, a classicist, Kelly does not bring new scholarly theories to the aporias of his classical sources; nor does he approach the art of classical tragedy itself or probe either its poetics or its function in Greek society. Rather, Kelly's purpose throughout his book is to proceed from Aristotle's inconclusive concepts of tragedy to hundreds of statements by late Roman and medieval writers about tragedy, all of which he takes at face value. By doing so, Kelly unpacks a granny's trunk of claims without a common denominator, many of them bizarre, that have accumulated around the term "tragedy" during some 15 centuries of pre-modern culture. It goes without saying that Kelly resists the modern temptation to essentialize the notion of tragedy, as did, for example, Northrop Frye; nor does Kelly harbor any romantic illusions about tragedy as some exalted state of consciousness embodying the fulfillment of Selbstbewusstsein. To the contrary, Kelly's patient and persistent semantic history of the term "tragedy" points up the inaugural fuzziness of an ancient term that endures in the modern critical lexicon. If we concede that an equivalent fuzziness might inhabit conventional terms for many other genres (among them, epic, romance, roman, novel, allegory, and parody), we cannot escape the worry that genre-based "literary criticism" is a discourse shot through with conventional terms that do not bear definition.

Kelly approaches Aristotle as a skeptic: "It is clear that many self-proclaimed Aristotelians not only read Aristotle selectively, but also employ a double standard in their use of the term 'tragedy.' They seem willing enough to use a simple generic meaning of the word when speaking of ancient tragedies: that is, they accept as tragedy any play designated as tragedy. But when it comes to postclassical works, they restrict the term to mean 'great tragedy,' in accordance with their own definition of tragedy" (xv). Kelly illustrates this claim as he swiftly evokes (in a mere 4 pp.) the apparent inconsistencies in descriptions of tragedy in Aristotle, proceeding then (in 9 pp.) to the Romans. Thereafter, he fans out into the various vernaculars of the European middle ages, including those of England and of the Romance languages.

Thus, without entertaining any modern attempts to reconstruct what Aristotle really means in all of the ellipses of the Poetics, Kelly simply deduces from what is said about "tragedy" or "the tragic" in the Poetics that no clear discursive boundaries existed in Aristotle's mind (for instance, epic is "tragic"). Nor did Aristotle evoke any adequate formal criteria for tragedy: pitiable, fearful or pathetic events and a catastrophic ending are not necessary to make a play tragic: "Aristotle could not avoid the fact that more than a third of Euripides's tragedies ended happily. And no matter how he rated their effectiveness, he could not deny them the name of tragedy" (4). Nor, finally, is the social function of tragedy very clear, for, as Kelly notes (3), the notion of catharsis is never developed by Aristotle beyond its first brief mention.

If the Greeks left a legacy of shifting sands in their critical thinking about about the poetics of tragedy, we could not have expected the Romans to shore things up. Kelly shows how Horace and Ovid diverted thoughts about tragedy into criteria of poetic style that had nothing to do with the form, subject matter, or social purpose of tragedy. The Roman link of tragedy with the high style of eloquence was not illuminating, since the most abundant thought about style was Cicero's, and Ciceronian notions of style involved only forensic debate and not the artistic premisses of tragedy. Kelly infers that the only norm that may have persisted in both Greek and Roman thoughts about tragedy involved the "spudean" -- that is, the element of the moral superiority of the tragic hero (as opposed to the "phaulic," or contemptible, nature of the comic character). The Romans did, however, convince themselves that tragedy is mighty sad, and perhaps on that basis it could be dismissed, in the words of the late grammarian Diomedes, as "nihil aliud quam miserarium comprehensio" (Kelly, 15).

Classicists will surely not be enthralled by Kelly's fast tour of Greek and Roman theories of tragedy, but since the knowledge either of Aristotle's Poetics or of Greek tragedy itself vanished from the the medieval Latin West until the rise of Humanism, Kelly's haste in such quarters is defensible. So too, his brevity, in chapter 2, with the Romans (who were exceedingly shallow and unsystematic in their thoughts about tragedy) is no less defensible, all the more since any familiarity with Roman that tragedy endured in the medieval ages was literary, and not theatrical. Indeed, Kelly infers (16) that stage productions of tragedy were no longer prominent in the civic life of Rome by the beginning of the Christian era. Poets continued to write full-scale tragedies, and even to hope that they might be produced "in the old fashion;" but more realistically, Kelly says, they could only hope that their written compositions would be adapted to other kinds of "tragic" performances current in Imperial times. These included the pantomime ballet with a chorus, the danced tragedy (tragoedia saltata) and the "citharedy," in which "a performer sang a tragic aria and accompanied himself on the lyre..." (17). "St. Augustine reports [De doctrina christiana 2. 38. 97, brackets mine] that in Carthage, before his time, a herald would describe to the audience the story that the pantomimist was about to enact, whereas in his own day, the spectator was left on his own" (17). Among the performers of late classical culture were the emperor Nero and, yes, Kelly claims, Augustine himself!

Though anyone who has wandered about the Roman ruins of the Mediterranean basin and who has seen how many late Roman theaters were built throughout the empire will wonder what might have been the chief performances given there, Kelly believes that full-scale tragedy was not among them, but that tragedies could be given in the abbreviated forms mentioned above, or else recited. "As with other kinds of literature, the circumstances of such readings of tragedies could range from an informal gathering of friends to the assembling of a large audience in an auditorium. We can readily imagine tragedies being read as after-dinner entertainments, perhaps with the various roles taken by different speakers. Finally, they could be recited in the theater, since even nondramatic poems were so presented" (19). By erasing the boundaries between drama and literature, Kelly implicitly highlights a theatrical dimension of late classical poetry itself, and such a performative dimension of, say, the Aeneid, (whose intertextual ties to Greek tragedy were strong) could perhaps be further explored by critics in new ways. And was the late Roman schoolboy in Thagaste who wrote so well about the wrath of Juno and who wept for Virgil's Dido not a most promising budding tragedian?

After his very rough sketch of the conditions of performance of classical Roman tragedy, Kelly skips to Lactantius in the 4th century. I know little about Roman drama, but I am struck by the paucity of information that Kelly has garnered about theatrical performance in late Antiquity, and leave it to others to evaluate his scholarship in this regard. Since the passage he quotes from Lactantius is mainly a fulmination against the licentiousness of Roman theater as seen by a zealous Christian iconoclast, it gives us precious little practical information, and perhaps Kelly wonders rightly whether Lactantius even knew what he was talking about.

Given the unenlightening contributions of the Romans to the concept of tragedy, it comes as no surprise that as Christian culture replaced that of the late Roman, the "applications" of the word tragedy would become only more diffused. Kelly says (23-5) that such disparate events or stories as Nero's murder of his mother Aggripina, the Biblical story of Herod and his wife and children, the martyrdom of a saint, and even the incarnation of Christ were all somehow "tragedies." The looseness of the term "tragedy" in early Christian culture would make even a modern tabloid reporter guffaw. Obviously, the Church Fathers were hardly more comprehending as spokesmen of pagan theater than are modern Evangelicals of Robert Mapplethorpe: neither crowd can see beyond its own obsessions into the deeper human or moral lesson embedded in art, no matter how eloquent it might be. Nevertheless, for his own purposes Lactantius did not hesitate to employ the term "tragedy" to dignify the suffering of innocent people and of martyrs before tyrannical Roman officials (32-3). So too, Boethius alludes to Christ's incarnation as a "great tragedy" (33). Such statements show the extent to which early Christian writers could emulate the very cultural traditions of Rome that they also reviled -- and with every bit as much vagueness.

With chapter 3, which opens with Isidore of Seville, we embark at last upon the high road of Christian medieval culture. If, as Kelly claims, "theatrical activity of the ancient world had come to an end by Isidore's time," and if Isidore himself has be en shown to have been blissfully ignorant of classical tragedy, whatever could he contribute to the debate about tragedy (or comedy)? The answer? More confusion. Even more removed from classical drama than Lactantius, Isidore concocted his tragic stew out of the leftovers of patristic and late Roman writers scarcely better informed than he -- to which he added his own speculations based, perhaps, on what he might infer from seeing the Roman theaters that remained standing in his era. "We get some notion of Isidore's lack of direct contact with the ancient authors when we see him name Juvenal and Persius as authors of new comedies" (37). Nor does it help when Isidore seems to forget what he himself has written (37). In book 8 of his Etymologies he defines tragic poets as poets who sang for the prize of a goat; then they sang of "public affairs and the histories of kings," and what they sang of was sorrowful (39). In book 18, he defines tragedians (tragoedi) as "those who sang in poetry of the ancient deeds [in antiqua gesta] and sorrowful crimes of wicked kings while the people looked on" (Kelly, 46; brackets mine). However, Isidore's remoteness from classical tragedy is not merely a consequence of his ignorance: it is above all moral: "Isidore's unfavorable attitude towards tragedy in the context of the theater -- the 'scene of the crimes,' so to speak -- is encapsulated, perhaps, by his note that another meaning of theatrum is 'brothel'" (49).

Hardly daunted by the incoherence of his late classical and early Christian sources, Kelly looks ahead to the high middle ages with cheer: "Because Isidore is simply passing on information from various sources about matters of which he has little or no personal knowledge, the question of how integrated his understanding of tragedy was is not of the first importance. More important is the question of how he will be understood by others who have little else to go on except one or other of his brief discourses" (50).

Did Isidore build, then, better than he knew? Remi of Auxerre, who wrote in the more knowledgeable era of the Carolingian renaissance, hardly put the house of tragedy in order, and multiplied the errors of his early Christian forebears: tragedy was largely a matter of style; because of its (false) etymological ties with the Greek word for "shadow" the "scene" (scaena) was, for Remi, was a "shaded area or structure" where plays were performed and poems recited. More interesting is Remi's awareness of the threatening proximity of the story of Christ's begetting to the "ludicrous and monstrous things" of tragedy (52).

Since scholarship in the vein of the intellectual historian must always press forward in time, Kelly's trajectory now (57) aims at "non-Remigians, Senecans, Horations, and later Isidorians" who lie between the Carolingian and 12th century renaissances" With admirable patience, Kelly threads his way through sources that many modern medievalists themselves may never have heard of. We will remain adrift in a confused and hostile sea raked by winds from all quarters. One can only wonder at the frivolity with which Christian culture remembered the classical culture it had supplanted. However, as a study in the intellectual history of a term, Kelly's book is exempted from the challenge of relating a chronically vague term to the broader cultural context of the world that continued to bandy it about. If the success of intellectual history depends on the coherent evolution of its subject material, "tragedy" is perhaps not very promising. It is like trying to make a lens out of a jellyfish.

In chapter 4, "The twelfth-century scene," we move into one of the great cultural epochs in Western history, but what we encounter here in the term "tragedy" is only a swamp of oddities which are the residue of classical and early Christian non-thoughts about tragedy. Perhaps William of Conches had a new idea that tragedy and comedy had plot structures (70), "But Conches in no way suggests anything of the theatrical context of tragedy" (70). Thus, for him too, the scena was a "covering" or "shade" because it beclouds the hearts of men (71). For the figure known as the Heiligenkreuz Anonymous, the scena is a brothel (73). And then we are fed crumbs from the tables of Bernard of Utrecht, Conrad of Hirsau, and so on. Out of this generous inconclusiveness comes the same promise that we were given early when we marched forward in time from Isidore: "It is William of Conches's formulations that are most significant in terms of their effect on posterity." The nuances that Kelly evokes are too fragmented to enumerate here, but to speak of their "effect on posterity" seems forced to me.

A suggestive part of Kelly's book is his probing of John of Salisbury's appropriation of tragedy as an arch-metaphor for reality: "all the world's a stage" (79). More learned about his sources than other churchmen, John extends his metaphor of tragedy as a work shaped by Fortune, but one whose spectators, because they are virtuous and wise, refuse to act in it (80). Kelly surveys later metaphorical uses of "tragedy," but these continue to rest on the notion of tragedy as a style that is embodied in non-dramatic writings, including Arthurian romance (86-7). However, for Peter of Blois, shedding tears for the Arthur of fable is useless if one cannot shed tears for God in true penitence (87). Tragedy is strongly associated with lamentation and grief experienced by those who experience loss or undergo the fall of Fortune. As we advance, the names of Otto of Freising, Bernard of Cluny, Berengar of Poitiers, Orderic Vitalis, and William of Malmesbury form a procession of major thinkers who toss the word tragedy around, but without bringing depth or novelty to it. Since tragedy remains linked to mastery of the grand style, it is applicable to a non-corpus of all sorts of works that we associate, perhaps just as loosely, with other genres. Indeed, the Isagoge in theologiam (mid-12th century) lumps all poetry under the category of tragedy (92). For Matthew of Vendome, tragedy is above all a ferocious roar (93).

What I find disturbing about the project that Kelly has so pursued with such devotion is that it puts many otherwise distinguished writers in a mediocre aspect: since they know virtually nothing about tragedy, they tend to wallow in the same empty formulas. "Tragedy," then, is one of several things: a story about a fall from a high place or a disaster, a contest for a goat, a poetic work in the grave style, a bombastic bellowing. and so on.

Even such fecund rhetoricians as Geoffrey of Vinsauf and John of Garland have nothing to add to the notion of "tragedy." For Geoffrey, tragedy deals with "grave persons" and their misfortunes, it begins in joy and ends in mourning, and teaches contempt of Fortune (99). As for John, writes Kelly (100), "He says simply that only two tragedies had ever been written in Latin: Ovid's Medea, now lost, and one of his own composition." How far into the abyss can the intellectual history of tragedy carry us ? Well, here is Kelly's summary of John's gift to the history of tragedy: "It tells the sordid story of a washerwoman and her soldier-lover, who are killed out of envy by a second washerwoman; the murderess then lets the enemy into the city in order to conceal her crime, and in the ensuing massacre all of the defending soldiers, including her own brother, are killed. The three properties of this tragedy, he says, are as follows: it is written (described) in a grave style; it sets forth shameful and criminal deeds; and it begins in joy and ends in tears" (100-101).

As we move into the Aristotelian "high middle ages" of the 13th century in chapter 5, we are confronted right away by the heading, "Aristotle: A lost opportunity" (111). Kelly is right. Albert the Great, Thomas Aquinas, Robert Grosseteste -- as well as their Arab sources, Averroes and Avicenna -- bring nothing new or substantial to the concept of tragedy. For Avicenna, tragedy deals with the misfortunes of the virtuous and deserving" (123). Though the Poetics was translated from Arabic in to Latin in 1278, it had no effect, even on its commentator, Albertino Mussato of Padua, "the only known user of Moerbeke's translation in the Middle Ages" (117).

So too, the discovery of Seneca's Book of Tragedies in the 13th century might have led to some speculations among the medievals as to what the notion of tragedy might have implied, but did not. Seneca was mined above all for moral wisdom, especially by Nicholas Trevet of Oxford, "who brought the French moralistic interest in Seneca Tragicus to its culmination" (126). Trevet's first contact with the notion of tragedy comes through Boethius, and this "provided him with a knowledge of tragedy from nonclassical traditions which he was to bring to his analysis of Seneca's plays" (127). Thus, Lactantius, Isidore of Seville and William of Conches furnish the parameters of his notions of tragedy as tales of deeds and crimes that are punished and end in grief (130-2), as well as his belief that tragedies were mimes performed to the recitation of the tragic story by a poet (134).

What Kelly has to say about Dante will perhaps be among the more useful parts of his book, not necessarily because he offers a basis for profound interpretations of the Comedy, rather as a brake for ungrounded speculation about a term that remains vague. In his De vulgari eloquentia, Dante merely equates tragedy with the noble style, elevated construction, and excellence of vocabulary, all of which are proper to the canzone (145). Its subjects are salus, amor and virtus, and it does not require either high crimes or a plot movement toward grief. Since Kelly does not believe that the Epistle to Cangrande is (with the possible exception of the opening paragraphs) authentic, Kelly does not deal here with Dante's alleged association of tragedy with fetid goat songs, or with his notion of the tragic plot as catastrophic. Kelly's refusal to read anything into Dante's very vague concept of tragedy and comedy will not, of course, staunch the speculations of those who (like, most recently, Robert Hollander) do argue for the authenticity of the Epistola.

Another potential occasion for a revival of thought about "tragedy" was the translation of Boethius's Consolatio into French in the 14th century. Once again, we encounter the impermeability of medieval culture -- that is, its lack of historical perspective and a readiness either to ignore the triumphs of classical culture or shunt their prestige into unrelated medieval cultural forms. Thus, despite the abundance of translations of Boethius into French, "in none of the versions is there a clear notion either that tragedies were a special genre of literature in Boethius's day, or that they are a surviving special genre still practiced [sic] by minstrels. The most common view seems to be that tragedy is an obsolete synonym for chanson de geste, and that chansons de geste in Boethius's day, at least according to Fortune, constantly bewailed her overthrow of happy kingdoms" (164).

After such a humble chronicle of medieval thinking about "tragedy" in Italy and France, will our hardy professor of English let some insular Cheshire cat out of the box to delight us? If not, will Spain disclose some pre-Columbian treasure to save continental culture from its doldrums?

The answer to the first question must wait: "The most important development occurred in England, with Chaucer and his fifteenth-century disciples, Lydgate and Henryson. These authors are so significant in the history of tragedy that I shall devote a separate study to them: "Chaucerian Tragedy" (169). In the meantime, the brief summary of that future study deals only superficially with the expected works, the Monk's Tale and the Troilus. Though the latter was perhaps the best "tragedy" of any stripe (and certainly the funniest) in any language since the Greeks, we will find out why when "Chaucerian Tragedy" comes to the light of day.

In the meantime, Kelly nourishes us with the bread and water of Humanist revivals of Senecan tragedy in Italy, which are "paltry," and destined not for theater but for recitation (193). Kelly does not, strangely, delve into Italian critical debate about tragedy, comedy and tragicomedy evoked so well a generation ago by Bernard Weinberg. Spain, to the contrary, brings us Celestina, which, as Kelly recognizes, is a genial rule-breaker, yet one that leaves me uncertain as to whether I should laugh or cry. My embarras may have been shared, given the confused "horizon of expectations," to use Jauss's term, that prevailed, according to Kelly in Rojas's culture: "In comparison with his Italian contemporaries and his Spanish critics, ... Rojas was quite limited in his knowledge of literary traditions, and this uninformed tradition won out in Spanish, at least as far as the word comedia is concerned: it came to be used regularly to designate any dramatic work, whether humorous or serious -- or, in our mode rn parlance, "comic" or "tragic" (202). Indeed, each of the three Spanish tragedies that Kelly describes in this chapter is, in his words, sui generis.

The value of Kelly's book will stand as that of a useful reference work in the domain of Rezeptionzgeschichte, though it will not lead far into the study of medieval "theater," much less, into the study of medieval culture in its broader parameters. Indeed, by carrying his study of "tragedy" through so many great epochs and authors of classical and medieval cultures, Kelly might encourage some to judge badly the rank incoherence of their thinkers, rather than to see their ignorance of "tragedy" as a precondition of medieval "theatricality" in all of its originality and wealth. Royal entries, coronations, tournaments, feasts, liturgy, executions and canonizations were all occasions where important persons assumed and acted out elaborate roles before audiences. Medieval "theatricality" would be the subject of a different kind of book.