No linked pair of poets has, over the centuries, been considered as so significantly related a pair of poet-prophets as Dante and Virgil. Not only does each of them present himself in the role of the vates or prophet, but their works contain solemn moments of explicitly prophetic utterance, promising us that we shall behold a novam progeniem of one sort or another. It further seems clear that Dante's presentation of himself as prophetic poet is at least importantly joined with his sense of Virgil's own assumption of that role.

      Among the past generation of Dante's readers it has become increasingly germane to place his role as poet into relation with his self-presentation as Judeo-Christian prophet. The «Italian school», in which the most significant name in our century is probably that of Bruno Nardi, and the «scuola arnericana», led by Charles Singleton, have in common, for all their many desperate differences, an awareness of Dante's appropriations of the vestments of such as David, Jeremiah, Isaiah, John the Baptist, St. Paul, and of John's vision on Patmos. Surely no one who reads even cursorily in the dantology is innocent of the notion of «Dante theologus-poeta». Whether or not this putative reader admires or rejects a theological formulation of the Comedy's essential stance, all can see that the second half of our century has seen the direction of Dante studies move away from aestheticism and toward theology.

      The questions which such a perception of Dante's theological purposes in his poem necessarily enjoin most dramatically concern Virgil. Just as we should never forget to be amazed at Dante's choice of Virgil as guide and master in this vigorously Christian poem, neither should we cease to be pestered by associated doubts: How can the prophetic enterprise of Virgil be assimilated to the specifically Christian purposes of the Comedy? How can Dante make himself at once a champion of Virgil, loyally following in the master's footsteps (as at the conclusion of Inferno XXIII), while he simultaneously condemns the master to hell? Why does Dante, who treats Virgil's texts with more deference (even with veneration) than did any Christian writer before him, at the same time subject these very texts to a hostile scrutiny which we would be less surprised to find in the pages of St. Augustine? To approach such questions, we do well to have some sense of the «problem of Virgil» as this has been met and formulated by Dante's readers. The problem has an interesting history, and the answers offered since the earliest days of Dante criticism, reveal the troublesome nature of the questions themselves.

      It is a disconcertingly widespread assumption, from the earliest commentators onward, that Virgil is not to be understood as a historical figure in Dante's poem, but rather as an allegorical expression, either of Reason in general or as the rational capacity within every man (or at least in this man, Dante Alighieri). This formulation, in whatever version, was born, I would suggest, of the union of Surprise and Embarrassment («Virgil? Dante must have meant something else» ^__ we can hear even a Pietro di Dante or a Boccaccio murmuring). It remains the most popular view of Dante's Virgil, even today. In it, both Roman poet and Roman poem are «de-historicized», and thus cleansed of their offensive pagan coloration. A good dozen of the most intelligent of the commentators during the first quarter millennium of interpretation of the poem greet Virgil's miraculous apparition in Inferno I with the perhaps reassuring (but certainly impoverishing) gloss, «id est rationalis philosophia», or «ratio naturalis». The phenomenon occurs just as readily outside of Italy. Most of us have grown up with like-minded commentators at our elbows (at random I have opened to the gloss in The Temple Classics edition of verse 63: «Virgil, who stands for Worldly Wisdom», it reads). Allegory may be ^__ when applied, as Pope Gregory the Great intended, to Scripture ^__ «the wonderful machine»; this kind of allegoresis, it seems to me, is more correctly recognized as «the dismal science». Its thudding flattening out of the problematic presence of Virgil into a one-for-one simplistic equation has an understandable appeal for its practitioners. It rids them of the need to encounter a complex and disturbing issue, with all its ramifications. For if «Virgil» stands for something else, the problem represented by his presence in the Christian poem simply disappears.

      It is thus that we tend to glide past the most disturbing moments in the text. For instance, when we observe that in the series of interrelated events in Inferno XXI Virgil is neither wise nor rational, do we argue that «the allegory is temporarily suspended»? In these scenes we observe the following reactions on Virgil's part: He is fooled by Malacoda's dropping of his weapon into believing that the demon fears his heaven-lent powers enough to intend no further harm (XXI, 79-90) he believes that Malacoda's information concerning the condition of the next bridge out of this bolgia is truthful (XXI, 106-111); he further considers that Dante's desire to proceed without demonic escort, out of his fear of potential harm, is ill-considered (XXI, 127-135). By the time we reach the twenty-third canto we discover that, on each of these counts, Virgil has been absolutely incorrect. At the close of that canto, chagrined, he is mocked by the Jovial Friar Catalano for not realizing that devils are capable of masking their intentions. When Virgil strides angrily away, toward the next bolgia, Attilio Momigliano has only this to offer in response: Virgil's behavior momentarily (there's our «momentarily suspended allegory») «takes off Virgil's back the somewhat burdensome and monotonous mantle of guidance» ^__ as though that were all that is here at stake. Yet Momigliano's gloss is only typical. Indeed, if Virgil «equals» Reason, such interpretations are only necessary, given the evident discrepancy between text and gloss.

      Only four or five of the early commentators have the wisdom to identify Virgil «historically», that is, as «a Roman poet». Yet even most of these, sooner or later, end up treating him «allegorically». It is something of a commonplace, encouraged by such commentators as the relatively astute Benvenuto da Imola, for the earlier glossators to assert that Dante's Virgil is to be considered both «allegorical» and «historical» (a condition which they also find pertinent to Beatrice). This second conception of Dante's Virgil is surely an improvement upon the first. Yet it too frequently resorts, when discussing Virgil's actions within Dante's poem, to the simple allegorical equation ^__ if with some happy intermittence. Let us briefly examine the key text for all commentators of this stripe.

      The champions of this allegorical reading of Virgil as character in the Commedia mass their troops behind the single text in the poem which does in fact associate Virgil with reason (and Beatrice with faith): Purgatorio XVIII, 46-48, where Virgil specifically delimits his powers to those of reason. This apparently convenient handle is more slippery than its maneuverers tend to appreciate. As I have pointed out in a forthcoming essay, each of Beatrice's nominal presences in Purgatorio (VI, XV, and here in XVIII) associates her in turn with one of the theological virtues: hope, love, and now faith. At the same time, while excluding himself and his fellows in Limbo from these virtues, Virgil does identify them and himself with the four cardinal ones (Purgatorio VII, 34-36). Thus his remark here may be seen to refer to a quality of pagan knowledge, prudence, if you will, while Beatrice is accorded capacity in one of the three theological virtues. To turn Virgil into «Reason» and Beatrice into «Faith» on the authority of these three verses is to extrapolate quickly and to disregard a wider context, in short, to take parts for wholes.

      The allegorized Virgil, as personification of «Reason», in whatever form and with whatever intermittency, has two aspects which recommend such a formulation: He helps a convinced reader to deal with Dante's poem as though it were to be conceived as a conventional fable (that is, it helps avoid the patently «historical» claims made by the text itself, through which Dante claims to be a poet only on condition that we take him first as historian, one who actually saw what he sets down), and it makes Dante's Virgil a mere appurtenance of this larger fiction.

      It was only with Filippo Villani's strange commentary to Inferno I that both these interpretive assumptions were strongly challenged ^__ if by a weaker mind than one might have hoped to have as champion of a better understanding. At the turn of the fifteenth century Filippo was the first commentator to attempt to accomplish two new and important things. First, he was the first to say explicitly (for an earlier tacit acknowledgment of Dante's paternity see Guido da Pisa's commentary page 290 in the Cioffari edition) that the Letter to Cangrande, with its claim for a Scripture-based allegoresis, was written by Dante (and to insist, in the track of the Epistle, that, consequently, the poem was a divinely inspired vision and not a mere fable); second, Villani not only claimed that the Aeneid was, unbeknownst to its author, a prophetic poem which pointed explicitly to the mysteria of the Christian religion, but that Dante himself had been the first reader ever to understand the pagan text aright. I have no desire to second these outrageous overstatements, but only to claim for them a movement in a better direction. Dante himself, I believe, would have been amused by Filippo's over-exertions (if he would have been dispirited by the dismal responses of his other commentators on these two major points). His rejoinder would have been, I think, that Filippo had failed to see the hedge which Dante had set between himself and his own claims for inspiration and the similar distance which he has established between himself and Virgil's claims for prophetic status. What I find especially instructive is that hardly anyone has taken Filippo seriously enough even to argue with him. He may be so sloppy a thinker as hopelessly to confound his Virgil-vates with the allegorized Virgil of the school of Fulgentius and (the pseudo-?) Bernard Silvester (something he should not logically have done) and to treat the Aeneid (and the Commedia) both as figural/historical and as fabulous (which he ought to have done in a more sophisticated way, perhaps). Nonetheless, his misperceptions may lead a dantista of our day to a more fruitful confrontation of the puzzling presentation of Virgil put forth in the Commedia.

      The Virgil whom I find now nearly everywhere in the poem is treated as prophet ^__ but as failed prophet. In place of Villani's syncretistic enthusiasm I would put a tragic paradox. For in my current readings of the Roman poet's presence in the Comedy I find the following startling and moving harsh moments for Dante's Virgil. I advance them here without argument or evidence, if with the hope that, even described briefly, they reveal that our traditional «comfortable» understanding of Dante's Virgil needs to be revised.

      Dante's Virgil first appears as a figure of John the Baptist, his near contemporary. (Within a generation of Virgil's death John would be preaching the Word made flesh.) Dante's Virgil, like John, appears «nel gran diserto», but we do not hear his voice «crying in the wilderness», but see someone who is «weak from long silence» (Inferno I). Virgil's fourth Eclogue may have converted Statius to Christ, but it failed to have anything like this effect on its author (Purgatorio XXII). Anchises' (and thus Virgil's) prophecy of the woeful early death of Marcellus finds its way to the mouths of the angels who surround Beatrice at her advent («Manibus, oh, date lilïa plenis» they call out), but now the sad vision found in the sixth book of the Aeneid yields a better meaning as a song of Resurrection, Roman funerary lilies become the Christian flower of Easter (Purgatorio XXX). And in the same canto, as Daniello was the first to point out, Virgil's fourth Georgic, with its double tragedy of Orpheus's loss of Eurydice and of his own life, is reformulated to serve as Dante's farewell to Virgil, when he perceives that his «sweet father» is no longer with him, the words «Virgilio, Virgilio, Virgilio» paralleling Orpheus's three apostrophizing cries, «Eurydice, Eurydice, Eurydice».

      The picture of Dante's Virgil which emerges from these and other considerations of his presence, both as character and as author, in the text of Dante's poem is one of a poet-vates, but of failed prophet. Dante's Virgil seems to me to have been considered a failure both in his tragic Aeneid (and surely Dante so considered the «genre» of the Aeneid, no matter how few of us recognize this important judgment), with its uncertain view of the future of the empire, a view that is countered by Dante's imperial Comedy, as well as in the fourth Eclogue, a closed book to its own author, if it illumined Statius to the truth:

In that sense, then, for Dante, Virgil is a light that failed. Dante's Christian Statius is, in my understanding, a fabrication, his conversion invented by Dante, entirely on his own authority, in order to allow us to infer that he himself had become again a Christian (having lapsed «nel mezzo del cammin»?) by agency of Virgil's text. And such a fabrication, along with the presences in the text of three other saved pagans (Cato, Trajan, Ripheus), serves more to blame Virgil than to praise him.

      All of the early cantos of Purgatorio show us a Virgil befuddled by the ways of Faith. The simile which opens the sixth canto, with its winning and losing gamblers, alludes unmistakably ^__ to me and to Margherita Frankel, whose current work considers this text closely ^__ to Dante as winner and to Virgil as loser. And like the losing gambler, we can easily imagine Virgil going back over the «plays» which he has badly made (Purgatorio VI, 1-3). That losing Virgil is not forgotten even in Paradiso, where the enigmas of the Aeneid are referred to in unflattering ways when they are compared to the Christian soldier Cacciaguida's plain Italian speech:

The author of the Aeneid may have done more than anyone else to help create the Comedy. Yet having done so does not gain him heaven ^__ or Dante's unconditional affection. The return of Virgil to Limbo, the necessity that puts an increasing burden of sadness on both character and reader as we move up the mount of Purgatory, will not be described in Dante's text. At the same time, our imagination is easily able to recreate the scene, partly because we have seen a version of it once before.

      Perhaps no passage better embodies the doubleness in Dante's judgment of Virgil which we have been examining than the words which he has issue from Homer's mouth in Inferno IV. Returning to the Limbus, Virgil is greeted by the «eagle» of epic poetry: «Onorate l'altissimo poeta». We sense the depth of Dante's love for Virgil. The next verse, as Elizabeth Statmore, then a student at Princeton, suggested some years ago, should hold our attention as well; it may not appear as significant, but it is so: «his shade, having departed, is now returning». Whether or not we are supposed to consider their possible reaction, we are nevertheless compelled by these words to wonder what the members of the bella scola thought when, only hours earlier, Beatrice appeared to lead Virgil out of Limbo. May they not have believed that, like Trajan and Cato, he was being elevated to heaven by a latter-day divine intervention? But no, here he is again, back where he belongs. We can sense in Dante's mind the appreciation of the rightness of Homer's words as his own version of Virgil's epitaph.*

ROBERT HOLLANDER

Princeton University

*This paper, delivered as a talk at a conference concerning «Poetry and Prophecy» held at Yale University on 28 April 1985, concerns a topic to which I have devoted a good deal of my recent work and to which I am returning for a more comprehensive treatment in the next few years. I have made little or no attempt to indicate the large bibliography of works which concern one of the central concerns of Dante studies, but have chosen to offer some of the more salient points which I have already made. The studies from which these are drawn refer to some of the pertinent bibliography; they are as follows: «The Tragedy of Divination in Inferno XX», in my Studies in Dante, Ravenna, Longo, 1980, pp. 131-218; Il Virgilio dantesco: tragedia nella «Commedia», Florence, Olschki, 1983; «Tragedy in Dante's Comedy», Sewanee Review, XCI (1983), pp. 240-260; «Virgil and Dante as Mind-Readers (Inferno XXI and XXIII)», Medioevo romanzo, IX (1984), pp. 85-100; «A Note on Dante's Missing Musaeus», Quaderni d'italianistica, V (1984), pp. 217-221; «Dante's Pagan Past: Notes on Inferno XIV and XVIII», Stanford Italian Review, V (1985), pp. 23-36; «Dante's Commedia and the Classical Tradition: The Case of Virgil», in Proceedings of the International Dante Symposium (Hunter College, November 13-16, 1983), forthcoming; «Dante's Georgic (Inferno XXIV, 1-18)», Dante Studies, CII (1984), pp. 111-121; «Inferno II: Commentary», Lectura Dantis Californiana, forthcoming.