SOURCE: Pucci, Joseph. “The Dilemma of Writing: Augustine, Confessions 4.6 and Horace, Odes 1.3.¹” In Arethusa 24, no. 2 (fall 1991): 257-79.

Mirabar enim ceteros mortales vivere, quia ille, quem quasi non moriturum dilexeram, mortuus erat, et me magis quia ille alter eram, vivere illo mortuo mirabar. Bene quidam dixit de amico suo: dimidium animae meae.²

(Conf. 4.6.27-28)

For I wondered that other mortals were living, because he whom I had loved as if he would not ever die had indeed died, and I wondered more that I was living rather than he, because I was his other self. Well Horace put it of Virgil: “the other half of my soul.”

                    Sic te diva potens Cypri,
sic fratres Helenae, lucida sidera,
                    ventorumque regat pater
obstrictis aliis praeter Iapyga,
                    navis, quae tibi creditum
debes Vergilium; finibus Atticis
          reddas incolumem, precor,
et serves animae dimidium meae.

(Odes 1.3.1-8)

Let the powerful goddess of Cyprus, let the brothers of Helen (bright stars), let the father of winds (by which all things are bound before Iapyx) rule you, ship, who owe Virgil entrusted to you; to Attic shores may you bring him unharmed, I pray, and preserve the half of my soul.³

Augustine's attitudes toward classical literature have normally been cast by scholars in the archetypal terms drawn from Confessions 1.13, where Augustine describes his reading of Virgil's Aeneid. In this justly famous recollection, Augustine offers that his excursions in the Aeneid were nothing more than an enchanting dream, futile, empty, worthy of avoidance, and much—perhaps too much—has been made of this episode and of its wider implications for Augustine as a reader of classical poetry. One overlooked aspect of this recollection, for instance, with fundamental implications about Augustine's attitudes toward classical poetry, is the fact that in the telling of his story, with all its attendant antipathy and doubt about Virgil and classical poetry, Augustine inserts a Virgilian half-line (atque ipsius umbra Creusae, Aen. 2.772), seemingly contaminating his own work with a pagan tag, but also linking his own spiritual state to the figure of Creusa who, like Augustine, is lost, dead to the world, a phantom.

There are other moments in the Confessions where Augustine will allow a classical text to intrude into his own story in this way. These episodes normally are flagged by Augustine, as here by the use of meter, in order to make the reader notice them. Invariably they deepen and enrich the context of his own spiritual history when they are made to function coherently in his own narrative. Unfortunately, such moments have yet to be examined in even a cursory way in Augustinian studies.⁴ Indeed, the literary element of Augustine's creative work in general awaits substantive treatment.⁵ While one paper cannot fill the lacuna represented by such an omission in Augustinian scholarship, a more modest goal can be advanced: to make good on this error of omission in one instance by examining an episode similar to the allusion to Creusa.

Such an episode, an allusion at Confessions 4.6 to Horace Odes 1.3, is the topic of the present essay. It will be analyzed in two parts. In section one, a reading of Odes 1.3 is ventured. Then, section two considers the ways such a reading is corroborated by Confessions 4.6 before going on to determine the ways in which the two texts form a composite meaning, a meditation on the dilemma posed by the act of writing and the morality of mimesis.

1. HORACE READS VIRGIL: ODES 1.3

Odes 1.3, the so-called propempticon for Virgil, has long embarrassed commentators, who have not known very well what to do with it.⁶ Some have suggested that it is two poems.⁷ Others have held that its tono scherzoso makes it a humorous extravaganza.⁸ It has often been viewed as a genuine propempticon, but this aspect of the ode has also been rightly questioned on several grounds. Scholarly contention aside, however, there can be little doubt that the poem is about crafting, writing, and the morality of textual mimesis, and if most commentators have not stated the central issue of this ode in language quite this frank, then they nevertheless have sensed this truth regardless of how they chose ultimately to present it. The presence of Virgil in a Horatian ode makes this general statement of meaning more or less indubitable.

Indeed, the presence of Virgil in 1.3 would seem to be the fundamental interpretive datum of the poem. The problem in the literature has been to determine the tone of the ode relative to its famous addressee and in going about this task, commentators have roamed far and wide, as we shall see presently, in attempting to make Horace and Virgil the best and greatest of friends. The diversity of scholarly opinion⁹ over this ode has also been a result of the inherent tensions between the form (genre) Horace has chosen in 1.3, the propempticon, and the content of the ode, which seems to subvert that genre.¹⁰ There are two fundamentally opposed ideas at play in this ode: there is the idea, implicit within the form of the ode itself, of propempticon, which functions by definition to offer a prayer for a safe return of a loved one;¹¹ and opposed to this formal idea is the contextual idea that men have no business boldly travelling over the water, challenging nature and the gods themselves. Odes 1.3 is a classic example wherein form challenges content (and vice-versa).¹²

Much of the work done on this ode has stood aloof from the problems inherent in such a challenge. Because Virgil would seen to play a prominent role in this ode, readers have been particularly hesitant to read much that is negative in its lines, since such a view would of necessity impinge upon Horace's estimation of Virgil.¹³ As a result, the inherent tension between form and content has often meant that scholars have chosen to privilege either the ode's form or content.¹⁴ Many have been content, for example, to read the ode generically as a propempticon, in which case the negative characterization of travellers as audax has usually been viewed as ironic or playful, or, in the worst cases, entirely ignored.¹⁵ Those who have sensed the fundamental negativity of the ode toward the image of the audax traveller, those who have viewed travellers as symbols of human volition and of the evil of human progress, have generally downplayed the connection of such a negative view to Virgil and have read the ode instead as so much philosophical lecturing¹⁶—or have opted to read a positive tone into the negativity found here.¹⁷

While evidence in the poem supports all of these interpretations to a certain degree, any reading that hopes to achieve the virtue of coherency within the confines of what the text itself says must take into account three variables that heretofore have tended to be analyzed separately, as if they had little to do with each other. First, the presence of Virgil in this ode must be acknowledged as central, not ancillary to some desire on Horace's part to lecture about formal philosophy via a whimsical reference to his greatest competitor for poetic fame.¹⁸ Then, the meaning and implications of the phrase animae dimidium meae¹⁹ must be tackled, a task that explicitly will involve an analysis of the figure of Virgil in the ode. Then, finally, the function assigned by Horace to the adjective audax must be more specifically determined. The phrase animae dimidium meae is important because it is used to describe how Horace seemingly feels toward Virgil; audax is important because it is used generically to describe every figure introduced in the ode after Virgil; and the figure of Virgil himself is important because nothing in the ode makes very good sense without his presence as a model of poetic mimesis of a specific kind.

The opposition between form and content inherent in this ode is mirrored initially in its opening lines, where various mythological figures are named periphrastically. Venus, for example, is diva potens Cypri (v. 1); Castor and Pollux are fratres Helenae (v. 2); Aeolus is ventorumque pater (v. 3). The mental process required to translate the periphrasis of these names in order to determine the identities involved highlights the larger disjunction between form and content upon which the ode is predicated. Horace's opening language, in other words, points to one thing but means another, a posture the ode in general will assume toward Virgil and his poetry. Moreover, the diction and tone of these lines is curiously liminal, as if to highlight the clash between what is said and what is meant through the addition of epic diction to lyric verse.²⁰ Venus, after all, is called after her epic title here, diva potens, while lucida sidera would seem to imply not only St. Elmo's fire but also, and more importantly, the serene images of epic night.²¹ Even the phrase fratres Helenae suggests not only Helen herself, so imposing an epic figure, but also the image of the Dioscuri in their single, epic, aspect and this image has special resonances for the ode.²² The Dioscuri are twins, presumably the closest relationship possible in human terms. But the details of their mythology highlight not the closeness presumed to animate their relationship but rather the division, the separation of the twins. Though Pollux was inconsolable at Castor's death, one version of their story has it that they were allowed to be together for half of the time, while another version holds that Castor was allowed to live again, but only at the expense of the life of Pollux, so that they lived on alternate days. In both myths, the twins live, but always separate and divided from each other. The symbology of the Dioscuri, then, in addition to highlighting the dissonance of epic diction amidst lyric verse, also points to a disjunctive relationship in which two brothers, once inseparable, now live only by virtue of their separation.

The opposition between form and content is specified further in the opening lines of the ode in the all-important phrase animae dimidium meae. This phrase participates in such oppositions because it follows on the heels of the periphrastic namings that sensitize the reader to the oppositions possible between form and content and which include the joining of epic and lyric language. The phrase is used to describe Virgil and the opinio communis has it that such a phrase is evocative, warm, tender.²³ Yet, given its proximity to the liminal language of the opening lines—especially the image of the Dioscuri separated from each other—it is not possible to read this phrase as only positive. This very well may be a positive statement, but it may also imply a negative sentiment. Given its position, it is simply too early in the ode to judge the tone of the phrase except to note that it is vague.

An examination of comparable phrases in the poetry of Horace is helpful in order to determine the tone of this key phrase, and meae … partem animae from Odes 2.17.5 is exemplary in this regard. There can be no doubt as to the warm emotion and tender devotion of this phrase, used by Horace to describe Maecenas, to commend the friendship he shares with him, and to affirm Horace's own resolve to die with Maecenas rather than live without him. 2.17 is surely not the only ode that suggests the intimacy and devotion of Horace and Maecenas,²⁴ yet if there is doubt about the positive qualities of animae dimidium meae, the indubitable devotion implied in meae partem animae is a nearly perfect phrase with which to front a comparison.

The rationale for choosing one phrase to apply to Virgil in 1.3, and another to apply to Maecenas in 2.17, can be situated in a clear historical context retrievable from the oeuvre of Horace, the details of which affirm Horace's tone relative to Virgil in Odes 1.3. That context is grounded in the solid fact that Horace and Virgil knew each other. Generally it has been assumed that they were friends—even good friends.²⁵ Evidence for their “friendship,” however, comes mainly from the Satires. For example, there is no evidence in the Odes (as there is in the case of Horace and Maecenas), to suggest that they were close spiritually and emotionally during the 20s b.c., when both poets were working hard on their respective masterpieces, the Odes and the Aeneid. The passage of choice is Sat. 1.5.39 ff., where Horace is effusive in his praise, mentioning his friends Plotius and Varius in the same breath as Virgil, and describing them all as … animae qualis neque candidiores / terra tulit, before going on in the next several lines to render them in warm and intimate terms.

Yet, the evidence in the Satires is, at best, inadequate to the present task because of chronology. Satires 1.5, for example, written sometime before 35 b.c. (the date of the publication of Book 1 of the Satires) represents Horace at about age thirty, a man whose views on friends, poetry, and the craft of writing are bound to be “younger,” less mature, than those expressed in the Odes. Moreover, the Virgil recalled in the Satires is not the poet of the Aeneid (as he is in Odes 1.3), but is rather the pastoral poet of the Eclogues (published in 38 b.c.) who, at Sat. 1.10.44 ff., is explicitly praised as such in no uncertain terms: … molle atque facetum / Vergilio adnuerunt gaudentes rure Camenae (cf. 1.10.81 f.).

This specification of the role and function of the Virgil that Horace finds praiseworthy in the Satires demonstrates that Horace's view of (and praise of) Virgil in the Satires is a praise of a specific kind of Virgilian mimesis found in the Eclogues. And that praise must be considered in the context of Horace's own immaturity, for he is by any accounting a different sort of man at this point, still questioning and formulating his concepts about poetry. In fact, he goes so far in Satires 1.4 as to disclaim the title of poeta altogether, while granting it to Virgil (me illorum [poetarum] … / excerpam numero, Sat. 1.4.39-40; cf. vv. 41 ff., also).

In charting the evidence by which the relationship of Virgil and Horace might be more profitably gauged relative to Odes 1.3, chronology is important. Odes 1.3, written no later that 23 b.c., that is, as much as a decade after the Satires, and conjuring up images of the Aeneid—not the Eclogues—deals with an older Virgil, a Virgil with a different vision. Horace, by the same token, is older, too. He has become the poeta he said he was not in the Satires, the author of the Odes contemplating the author of the Aeneid. The historical situation of the two authors offers one way to begin to think about the difference implied between the phrases animae dimidium meae and partem animae meae: there must be a good reason why Horace does not use the word pars in describing Virgil but chooses instead dimidium. Attention to the etymology of pars and dimidium suggests a further difference that goes a long way toward corroborating the idea that their early friendship did not endure on the same terms, that Horace, as he matured, came to question Virgilian epic mimesis.

Dimidium is related to dimidio, meaning “to divide into two equal parts, to halve.”²⁶Pars is much more intriguing etymologically. Lewis and Short offer no verbal root in Latin, but the Latin noun portio, a word which means “a share, a part, a portion,”²⁷ is listed as kindred to pars, and the Greek poro, “to furnish, to offer, to give,”²⁸ is listed as kindred to both these nouns. The implications of “halving” and “sharing” implied in the words dimidium and pars make good sense in the context of Odes 1.3. The etymology of pars, for example, corroborates what the history of Horace's Odes already tells us about the friendship of Horace and Maecenas. Pars is a more inclusive word than dimidium and does not imply separation in the way dimidium does. Based on etymological data that imply sharing or a relation of one thing to another,²⁹ it is possible to view pars as denoting quite a different idea than dimidium.³⁰ This does not mean that the phrase of 2.17 is wholly positive and the phrase of 1.3 is entirely negative; but it does mean that the tenderness, the idea of one soul that is shared by two, the implications of a relationship that is based on a joining of selves, in short, the images and ideas suggested etymologically and historically by the word pars as used by Horace in 2.17, do not obtain in 1.3.

There, the use of dimidium implies a starker image, one based in cutting, in the division of something into two separate things, the halving of something.³¹ This image is quite apart from the inclusiveness suggested by pars. The key implication of dimidium would seem to be the idea of halving, which implies a former unity that is now changed, a wholeness cut asunder. This idea well symbolizes the relationship of Horace and Virgil in 1.3. Their souls, once joined, are now—for reasons that will become clearer as the ode progresses—halved into two separate but equal parts. The image of the Dioscuri found in the opening verses of the ode comes to mind now again, for the mythology of these twins represents an intimacy that is transformed by circumstance into situations of separation and division, precisely what seems to be suggested here by Horace in the phrase dimidium animae meae.

The extrapolation to poetry writing and textual mimesis is not difficult to make at this point, and the historical context offered by the Satires corroborates what the later evidence in the Odes suggests. Virgil and Horace were once of a kind, twin souls, as it were, like the Dioscuri themselves, but they have been separated by the different paths down which their talents have drawn them. Virgil is the “other half” of Horace's soul in the sense that he, Virgil, writes poetry of a completely different kind than Horace. And he is the “other half” in a personal sense since, as Fraenkel has noted, there was a wide disparity in the temperament of these two great Augustan poets.³² Ultimately, this becomes an aesthetic statement: Horace is not a Virgilian poet. He prefers a different kind of craft. As a human being, Horace understands Virgil's epic and tragic vision, but as a poet who has struggled long and hard for his own vision he cannot abide such a vision (if he ever could) because it is not his (and can never be).

The use of periphrasis in the opening lines of the ode is important not only because it suggests a disjunction of form and content as specified in the phrase animae dimidium meae, but also because it sensitizes readers to the first mention of Virgil who, in stark contrast to Venus, Aeolus, and Castor and Pollux, is called simply by his nomen. Such a naming, in addition to calling attention to itself in its clarity and precision, suggests on the part of Horace veneration, respect, understanding, but also an urge to isolate, fix, and wag his finger at the poet of the Aeneid, to control him at least textually and lecture him about the topic of good and bad poetry and moral and immoral mimesis by refusing to envelope him in anything other than clear language. Here clarity implies control so that something important might be said about mimesis and the contours of the lecture that will follow are implied in the division implicit within the word dimidium.

The lecture that follows consists of two blocks of material identical in length, vv. 9-24 and vv. 25-40.³³ In the first block, Horace worries over the “trip” Virgil is about to undertake by thinking about the hypothetical “first traveller” of the seas, but here again form and content collide, for Horace refuses the typical inclusion of a schetliasmos,³⁴ choosing, instead of cursing the inventor, to praise the “first traveller,” of ships:

                    illi robur et aes triplex
circa pectus erat, qui fragilem truci (10)
                    commisit pelago ratem
primus. …

to him was oak and three times bronze encircling his chest, who committed his fragile boat to the savage sea before anyone else. …

Ratis is especially noteworthy, for although it functions here in the context of fragility, it will eventually come to represent something much more negative. After framing Virgil in the context of a first traveller, a context that certainly has a literary analogue in Virgil's pathbreaking travels through Latin letters, Horace offers this catalogue:³⁵

… nec timuit praecipitem Africum
          decertantem Aquilonibus
nec tristis Hyadas nec rabiem Noti
                    quo non arbiter Hadriae (15)
maior, tollere seu ponere vult freta.
          quem mortis timuit gradum,
qui siccis oculis monstra natantia,
                    qui vidit mare turbidum et
infamis scopulos Acroceraunia? (20)
                    nequiquam deus abscidit
prudens Oceano dissociabili
          terras, si tamen impiae
non tangenda rates transiliunt vada.

… he feared not the rushing Southwest wind fighting with the North wind, nor the gloomy Hyades, nor the rage of the South wind (no judge to the Adriatic greater) whether desiring to excite or to quiet the seas. What approach of death feared he, who, with thirsty eyes, saw monsters floundering, seas troubled, and the infamous promontories of Acroceraunia? Pointlessly God chopped off in his wisdom the lands from the incompatible ocean, if perverse boats still cross the seas that ought to remain untouched.

This catalogue is hardly written to make the prospective traveller feel more comfortable. It is filled with storms, winds, monsters, promontories, and evil tidings. It suggests in its negative view of nature and of travel a literary analogue since it well describes the Aeneid itself, so full of monsters, storms, wanderings, and evil places. But there can be little doubt that the language of these lines is hardly credible as a formal propempticon, especially when taken in the context of the second block of material.

That block begins with a severe questioning of the ratis, the vessel used by the traveller in this ode and metaphorically the means by which Virgil travels through and in language. Virgil is the first traveller of Latin letters, riding out a great epic storm, but his ratis of a sudden becomes impia at v. 23 and he himself as the traveller of this ode becomes audax at v. 25. Now the disjunctions and liminalities of the opening lines of the ode are transformed. Horace crosses the threshold and begins a reasoned lecture on Virgilian epic mimesis:

          audax omnia perpeti (25)
gens humana ruit per vetitum nefas.
          audax Iapeti genus
ignem fraude mala gentibus intulit.
                    post ignem aetheria domo
subductum macies et nova febrium (30)
          terris incubuit cohors,
semotique prius tarda necessitas
          leti corripuit gradum.
expertus vacuum Daedalus aëra
          pinnis non homini datis; (35)
perrupit Acheronta Herculeus labor.
                    nil mortalibus ardui est;
caelum ipsum petimus stultitia, neque
                    per nostrum patimur scelus
iracunda Iovem ponere fulmina. (40)

Bold to allow all things, humankind runs, despite prohibition, to crime. The bold son of Iapetus brought forth fire by means of a fraud evil to all men, for after fire had been pilfered from its heavenly home, poverty and a new company of ills lied upon earth, and remote once and slow, the necessity of death rushed its pace. Daedalus tried the free air with wings not given to men; Herculean effort broke the way through to Acheron. Nothing for mortals is too steep; heaven itself we seek foolishly, nor through our sin do we suffer Jove to lay aside resentful bolts of thunder.

The transformation of ratis, from fragilis (vv. 10-11) to impia at vv. 23-24 is fundamental to the lecture Horace offers here. When Horace writes si tamen impiae / non tangenda rates transiliunt vada, it seems hard to defend such an image as somehow appropriate to the purposes of a serious propempticon. If the means by which journeys are undertaken are condemned, then the prayer for safekeeping is undermined.³⁶ When such a destabilization occurs, ratis, as a metaphor, becomes a key symbol by which Horace questions the kinds of literary journeys Virgil had undertaken (in the Aeneid). Such journeys are themselves metaphorized in the figures of Prometheus, Daedalus, and Hercules. Beyond their appropriateness as epic heroes, figures who “did it all”—or tried to—is their shared arrogance. Each was, in his own way, supremely bold in doing what he did and their audacia was in due course punished. But since nil mortalibus ardui est (v. 37), Jupiter can scarcely afford to rest the fulmen, the mechanism of punishment.

If Horace is less than convinced that Jupiter's thunderbolt will come crashing down on Virgil's head, then he clearly believes nevertheless that there are problems that exist for those who are audax and special problems for literary audacia of the kind represented by Virgil as an epic poet.³⁷ To be sure, Horace does not excuse himself from the company of those who try to do bold things with language. He makes the final verbs of the poem first person plural (petimus, patimur, vv. 38, 39) for just this reason.

But beyond that inclusion lurks the problem of degree. Poetry is written by degree, ranging the spectrum from the purely personal singing of lyric to the comprehensive song of epic. Epic is not lyric. The song of self embodied in lyric is not the same as the song of epic, always seeking a grander purview and vision, indeed always seeking caelum ipsum. To sing lyric is not to compete with the master poet Homer, either, and if Horace ever admitted to himself that he was in fact competing in his lyrics with Pindar it must be carefully noted that he always goes out of his way to say that he was not—and could not be—a Pindaric singer.³⁸

Horace seems to say that Virgil had tried too much in the same way that Daedalus, Prometheus, and Hercules had tried too much. The audacia involved in epic journeying of a Virgilian kind is fundamentally to challenge the gods at their own game, to try to recreate “everything” in the epic genre as if the poet were a god of creation himself. The desire to achieve this kind of mimesis is akin to the desire to move about the world over the water or to give fire to men or to try to fly. To attempt such acts invites, if not divine punishment, then surely psychological punishment: in beginning to feel as if he has conquered nature, for example, man can begin to substitute himself for nature, or for God(s), can begin to think that he can control nature more fully or completely simply because he has controlled it in one instance. Such a substitution is dangerous ultimately because it can lead to a sense of invulnerability which is always, as the histories of Daedalus, Hercules, or Prometheus demonstrate, associated with punishment or death or both.

Audax is the key word of the second block of material (vv. 25-40), the quality Horace challenges most strongly. It is a special word in his vocabulary, most often used in his poetry to describe figures who seek great accomplishments, with the idea that these accomplishments should not be sought in the first place. In Epistle 2.1.182, for example, it is used to describe a kind of poet (poeta audax) who is showy, who writes only to gain the attention and approbation of the audience. Although Horace claims the highest respect for such a poet, there is a sense of irony in this epistle, addressed, after all, to the emperor Augustus and prominently mentioning as it does another audax poeta, Virgil himself. At Odes 4.2.10 audax is also used of the would-be Pindaric poet, the poet whom Horace says he cannot be. In Odes 1.12.21, audax is used of Athena in a poem that emphasizes—while it seems to close—the gap that remains between what humanity can do and what it wants or seeks to do.³⁹

The linkage of these supreme examples of audacia and the figure of Virgil is stark. The special connotations of audax, when coupled with the stylistic features of the opening of 1.3 and the ways in which the ode as a propempticon fails to do what it is supposed to do, suggests that this ode is anything but a propempticon written by one poet-friend to another. In attempting to replicate in text a comprehensive and fundamentally negative view of the world, Virgil had achieved only a vision of evil and doom that highlights mankind's inability to know his world entirely or even clearly. Virgil, in a poetic sense, as an artist, had sought heaven itself, and the consequences, so Horace suggests in this ode, do not bode well for mankind in general or for Virgil in particular as a poet.

Strategic transformations and symmetries manifest and exploit the disjunction of form and content as the ode comes to a close. The opening prayer is transformed into the harsh image of Jupiter hurling his thunderbolt, so that Venus, the goddess of love (and Aeneas' mother) becomes Jupiter himself, the harsh father doling out punishment. The genre of the ode is transformed also before readers' eyes, so that the propempticon ultimately does exactly what it is not supposed to do. Virgil is also transformed, along with the craft upon which he sails, from animae dimidium meae who travels upon a fragilis ratis, to Prometheus, Daedalus, Hercules, challenging nature like the impia ratis of vv. 23-24. We begin then with prayer, fragile boats, a neutral vision of Virgil, and we end with condemnation, impious boats, and a questioning of textual mimesis as practiced by Virgil in the Aeneid.

2. AUGUSTINE READS HORACE (READING VIRGIL): CONFESSIONS 4.6

Horace's Odes 1.3 supplies a sublime distance, the allusive background against which Augustine composed the details of Confessions 4.6, one of the more poignant recollections of his autobiography. Augustine's text is, on one level, a controlled but emotionally charged memoir of the death of his youthful best friend. Increasingly, however, as the play of allusion occurs, the emotion of this memoir is transformed into a meditation on the morality of textual mimesis that mirrors Horace's apprehensions in 1.3. The same tension between form and content found in Odes 1.3 is found at Confessions 4.6.

The text opens with a deceptively touching and simple recollection, in which Augustine recalls what was surely a keenly felt loss within the context of the friendship of Pylades and Orestes. Such a recollection makes good sense and seems natural, at least initially. More than simply suggesting an appropriate context in which to frame the loss that he feels, the presence of figures from classical lore suggests that Augustine is describing the event historically, that is, as he remembers it happening in the past. It seems not at all contrived, as one reads Confessions 4.6, to think of Augustine recalling one of classical literature's great friendships, and the reverberation of that friendship in the Confessions lends a depth and subtle grace that the story otherwise would lack. Moreover, it is not hard to believe that Augustine thought of this prototypical friendship as his friend lay dead, nor is it difficult to imagine Augustine considering this friendship in such a context earlier, while his friend lived.

A few lines after this recollection, however, Augustine alludes to Odes 1.3 in a rather obvious manner (bene quidam dixit de amico suo …) as if desiring to draw close attention to the passage. When focused on these lines, such attention occasions an important shift in tone, transforming the memoir from seriousness to irreverence and mimicking the ways in which the tone of Odes 1.3 moves from seeming seriousness to something else altogether. Confessions 4.6, after all, is about a leave-taking distinguishable from Virgil's only in permanence: Virgil is travelling east in Odes 1.3 while Augustine describes in 4.6 the death of his best friend. The allusion initially strikes, then, all the wrong chords and makes Augustine seem to be about to lose control of his memoir.

Beyond irreverence, however, the allusion transforms what had been a tender recollection into a contrived literary reflection, and this contrivance sets up the textual mechanism by which Augustine continues to command his classical material as we would like. Part of the contrivance inheres in Augustine's advertisement that he is alluding in the first place (bene quidam dixit …). Beyond this, the transformation is achieved stylistically through the intrusion of the allusive material itself, since the allusion is inappropriate to the tone of the text at this point. It is asking too much of any reader to believe that Augustine would have thought of Odes 1.3 at the time of his friend's death. The line he recalls in and of itself is touching enough, but Augustine knows the entire ode, as would any audience responding to this allusion, and Horace's third ode of Book 1 is not an emotional nor even a very lyrical moment in his work.

Moreover, the Horatian allusion calls attention to the fact that the narrative, heretofore operating as if it were an historical account, is actually a contrived recollection of a past event from the purview of the present. Instead, the allusion highlights the literary quality that intrudes into the text at this point, accentuating the act of writing itself, not the memory being recalled. The recollection of these lines, in other words, moves from historical accuracy as found in the example of Pylades and Orestes (this is what happened at the time that it happened, as I remember it), to literary contrivance (this is what I am thinking about in the present, as I write about a past event). The distinction is crucial in any attempt to understand 4.6, since the movement across time implies an ongoing dialectical process wherein the historical act of recollection is questioned and examined in the literary act of writing.

Beyond this implication, however, one must consider the reason Augustine would allude at this point to Odes 1.3 at all. Under what circumstances might this allusion offer insight into his own situation of discourse? In addition to transforming the narrative pattern at this point in the text, the allusion draws attention to Augustine as a literary reader. Yet given that, why should this sensitive and prolific reader of classical texts pick this particular moment in Horace to use in the Confessions? Certainly he could have chosen other, more lyrical, moments; why did he not choose Odes 2.17, for example, and the particularly apt phrase discussed above, meae partem animae? Given the context of 4.6, the phrase from 2.17 is certainly just as apt as the phrase from 1.3 that Augustine ultimately did choose, and even more so.

If, on the other hand, Augustine means to suggest certain ideas about the rectitude of writing in general and of the kinds of writing he is pursuing in particular, then this allusion, after the fashion of a Chinese landscape,⁴⁰ offers a sublime distance against which the details of the text of the Confessions takes shape. A warm recollection of youthful friendship is cast in the wider context of the rectitude of Christian mimesis in general: just as Virgil had attempted too much in 1.3, so Augustine suggests that he himself has attempted too much in talking to God. Such a reading is particularly apt since the linkage between Confessions 4.6 and Odes 1.3 is grounded in the phrase dimidium animae meae,⁴¹ which accentuates not simply the ties between the two “friends,” but also the particular kind of “ties” implied in 1.3. The word dimidium is central to the meaning of Confessions 4.6, then, since through this key word Augustine frames his portrayal of youthful friendship.

The spiritual implications of 4.6 are such that dimidium makes sense in a way that pars never could. Ultimately, 4.6 is not about spiritual unity or close friendship as much as it is about error. Augustine's text is filled with questions grounded in moral outrage and spiritual horror. He cannot quite get over the fact that he loved his friend more than anything else. This error is compounded with the realization that, being so blind, yet he was somehow able to extricate himself from the spiritual morass that such an unfounded love implied. The death of his friend comes to symbolize an essentially evil soul, longing after a life that it knows imperfectly, fearing the death that can offer it eternal salvation, mired in this world, yet unaware of its condition. The initial question of this chapter is exemplary: quid autem ista loquor (“why do I even talk of such things?”).

The implications of that question are rhetorical and moral. Rhetorically, the question represents Augustine's attempt at self-incrimination, a way for him to suggest boldly that he has little if any business saying anything at all in text, especially of topics about which he clearly knows little. On another level, this question represents the proper moral severity necessary in a text that recounts youthful indiscretion and sin. Augustine was leading a life wholly insufficient and inadequate in Christian terms, and his response to the death of his dead friend is but one example of the exaggerated and “worldly” response to temptation. In this light, the use of dimidium well suggests the halving of his soul implicit in the act of Augustine's conversion, his separation, as it were, from a prior spiritual state. Superficially lending a tender feeling to the recollection, the word also subtly suggests the more permanent separation or “halving” that occurred between his friend and himself by virtue of Augustine's conversion. In remembering the death of his friend, Augustine could also see a severe choice being made; like Horace, Augustine had chosen one path while his friend—through death—had taken another. And like Horace, the starkness of the former path seems at once formidable and harrowing, all the more since it was part of his former soul that has since been saved.

Read with the Horatian allusion more specifically in mind, however, the question, like the text in which it is imbedded, is transferred from the rhetorical to the moral realm. The halving in Horace that is poetic is transferred to the spiritual realm in Augustine, but the separation and distance implied in the phrase animae dimidium meae are stark and real. The impropriety of Virgil seeking caelum ipsum certainly carries over here into Augustine's language, forming the moral context in which he contemplates this loaded question. More than simply shaking his head and lamenting at oats wildly sown, Augustine here wonders at the moral implications of what he is doing as a literary artist while in the very process of doing it.

This is a broad and severe questioning with several important aspects, all of which are specified in the Horatian context brought forward allusively in Augustine's text. On the one hand Augustine would seem to imply a nagging sense that to inscribe in text youthful indiscretion—even in the name of deploring it—is somehow morally to perpetrate it, to fix it and modulate it for all to see. This would seem to be one of the key problems with Virgilian epic poetry in Horace's view, which is filled with gloom, monsters, death, and larger than life evils that are somehow made real, fixed, modulated, in their inscription in language and text. Of what use is such a vision? Are there not inherent dangers involved in such a vision? For a Christian, of course, the moral implications of such a problem are cast in quite different though no less profound terms than for a pagan, but the sensitivity of the discreet and nebulous powers of language is unmistakable and tantalizing in either work.

On a larger scale, the questioning of language in a Horation context devolves ultimately onto the figure of God, with implications about the very ways in which we conceive of and talk about Him. In this sense the Horatian allusion highlights the absolute sublimity (and so the absolute absurdity) involved in the detailed writing to God about one's (God-given) life as if He were a confessor-priest. Augustine, in doing this, seeks caelum ipsum, seeks to talk to God and, more than this, succeeds in “textualizing” Him, limiting Him and constraining Him in text, modulating and fixing Him for all to see in the process.

This act, which might on one level seem ultimately submissive, is, in a larger sense, the height of audacia for a Christian, the boldest act he or she might attempt—bolder even than the audacia of Hercules, Prometheus, or Daedalus. But such a realization, as horrible and as terrorizing as it must have been for Augustine, also must have held out the real possibility of delight, else why would Augustine have dared to write the Confessions at all. Such a delight might in fact be considered a kind of temptation. To acknowledge such a fundamentally skeptical view of language, after all, would not be the same as acting upon such a view. To succumb to the overwhelming oblivion to which the imperfection of human truth leads would be to succumb to the temptation to doubt the efficacy of faith itself. But to challenge such a skepticism would be to affirm faith in a perfect God even if it were an affirmation imperfectly cast in human terms.

Such a discordancy is perfectly mirrored in the texture of the Horatian allusion, for it offers a most effective way of meditating on this problem at a distance, allusively, without being subsumed by it. As such, Augustine is able to suggest both the dilemma and the delight of the issue. Horace's presence at this point in the Confessions seems precisely the evidence that allows readers to gauge the extent of Augustine's sensitivity to the fundamental moral dilemma involved in writing. And the finely tuned literary sensibility employed to express such a dilemma shows Augustine to be caught in—and to be aware of—the paradoxical horror and delight of the twisting, imperfect, lanes of language.⁴² His sensitivity to this issue surely would have suggested to him that to write the Confessions was supremely audax, in the same sense that Prometheus or Daedalus, or even Virgil, were supremely bold.

Beginnings and endings are always important in the Confessions. Augustine begins chapter 6 with a frank question of the propriety of his talking (writing), the details of which concern the death of his youthful friend. The Horatian allusion there, lingering at a distance in the allusive play of these texts, seems calculated to confirm that question and widen its scope in a Christian context significantly. We share with Augustine the grief and tenderness of a moment in his life, but looming from beyond the humanity of this moment is a huge question concerning the morality of writing about such events in the first place (the distance of the landscape). Always in Augustine's view, to write was to fail.⁴³ To write of such topics as if God were one's confessor was to stretch human understanding of God beyond reason, to stand the concept of God on its head, to compete with God as supreme auctor, molding and shaping experience to one's own ends. For one so immersed in Christianity, the thought of writing to God might seem at first to be wonderfully attractive, but for a Christian such as Augustine only the prospect, not the reality, would have been inviting.

Notes

1. I owe thanks to my teachers, W. R. Johnson, Winthrop Wetherbee, and Frantisek Svejkovský, who helped to formulate at an early stage some of the ideas presented here; to E. A. Schmidt, who read the penultimate version with incomparable erudition, for which I am wholly grateful; to Louis J. Swift, who read earlier and lengthier versions several times amidst heavy burdens of his own; and, finally, to my students at Brown in several courses over the past two years on late antique poetry and rhetoric, for their avid interest, many corrections, fresh enthusiasms.

2. My edition of the Confessions is Skutella, Solignac, Trehorel, Bouissou 1962.

3. The translations of the Latin, which are my own, attempt to preserve the Latin word order while communicating (in the case of Horace) the fundamental negativity for which my reading of his ode argues.

4. The moments I analyze here represent my own personal interests. There are countless other moments in the Confessions, none of which have been studied. Relative to Horace, one might also analyze 9.1.9 ff. where Odes 1.18.4 is normative or 6.10.38 where Epodes 2.3 is normative. Horace exists in other places in the Confessions and a lengthy study could be devoted to them. For a partial list, see the Index Scriptores Profani in Skutella 1969.378-379. Skutella's list is far from complete. A more complete survey of the Horatian material in Augustine, something computer technology might hasten, is a present need. The last work to attempt such a survey was Keseling 1931.1278-1280.

5. Marrou 1958 is the basic work in this area; Courcelle 1963 is standard for the context in which the Confessions was written. In general, however, work on the literary content of the Confessions does not exist apart from the other more prominent concerns of Augustinian scholars. So, for example, one finds all sorts of aesthetic discussions in O'Connell 1969 but no specific discussion of literary issues. There are a few exceptions. Arthur 1941 is a very large discussion of art that includes, as one would expect, literary concerns along the way. Burke 1970 discusses the forms and functions of language in Augustine, the logical beginning point for literary studies. O'Meara 1963.252-260 treats Augustine's relationship with a classical poet. Cf. Hagendahl 1967. O'Donnell 1980.144-175 is a recent exception to the tendency not to impute much importance to literary reading in Augustine. Spence 1988 offers an up-to-date bibliography in this regard with important readings of her own relative to Christian rhetoric. There are other works that could be mentioned; the point is that there is a need for specific studies dealing with Augustine's use of the classical poetae in a literary way. There is no study that treats Augustine and Horace exclusively in any language to my knowledge.

6. The characterization is from Commager 1962.118. By way of example, Nisbet and Hubbard 1970.44-45 say this about 1.3: “[it] is an accomplished piece of versification, but little more. The poet … shows none of his usual tact and charm; there is not a hint of Virgil's poetry [sic, this is just plain wrong], and it is wrong to argue … that the ode's sombre and religious tone is directed specifically towards the recipient. The second part of the poem is equally unsatisfactory; … here the trite and unseasonable moralizing seems out of place in a poem of friendship. … Nor is the flatness of the thought redeemed by any special excellence in the writing … we miss the Horatian virtues of brevity and incisiveness.” Williams 1968.159 makes much the same comment, calling the ode “undistinguished,” and concluding that there is not “the slightest sense of a personal relationship in the greater part of the poem.”

7. So argues Prodinger 1907.165 and, with equal vigor, Verrall 1884.120 who asserts, as always with enormous erudition and persuasion, that the first eight verses of the ode are a true propempticon simply added to vv. 9 ff., which were written earlier and not as a propempticon. Most recently, Carrubba 1984.166-173 has argued for the coherency and structural integrity of this ode.

8. See Hahn 1945.xxxii-iii, who would see comedy at work in this ode.

9. For a complete bibliography, see Elder 1952.140-159, for works up to 1952, and Basto 1982.30-43, for the bibliography since then.

10. On the genre relative to Horace's use and abuse of it see now Basto 1982.30-31 and nn. 1-5.

11. Although Horace does not do this; instead, he prays for a safe passage. Cf. below, n. 36.

12. I do not hold to the view that genres in classical poetry are immutable, fixed like Platonic “forms,” but surely to change accepted norms of a genre—however fluid they may be—is to invite attention to that change. Such attention may occasion, as here, the discovery of something novel in the (newly configured) genre.

13. Thus Moore 1902.63 f. calls Horace “tactless” in linking Virgil to the idea of audacia.

14. Thus Lockyer 1967.41-45 reads vv. 9 ff. for their “positive” tone “which,” he goes on to add, “when read negatively, have often led critics astray” (p. 43). I believe we must be led astray in this way in order to make this ode coherent.

15. So Hendrickson 1907-1908.100-104 tried to explain away the tension by arguing that the first eight lines were all that Horace intended Virgil to hear (thus establishing Horace in his view as following Menander in the form of the propempticon), and that the rest of the poem was out of Virgil's earshot. Even so, how does that mitigate the negativity of what is to follow in the ode?

16. So Delatte 1935.309-336 argues that the ode condemns, under Cynic influence, the study of theology, astrology, and cosmology, not Virgilian mimesis.

17. Thus Elder 1952.140-159 is followed by Lockyer 1967.41-45 in reading this ode as a positive symbol for the poetry of Virgil. Basto 1982 follows this line of argument most recently.

18. I do not believe Horace and Virgil were enemies, but both men were enormously gifted and it seems to me indubitable that either would have felt rivalry toward the other. Cf. Duckworth 1956.281-316, and below, p. 265 f. on the developing relationship of these two poets as evidenced in the Odes and Satires.

19. The idea has an ancient provenance, possibly appearing as early as Plato (Symp. 189C ff.), and recurring in Callimachus (Epigram 41.1), Meleager and others. Yet the idea of someone being an alter idem should not be brought to bear here (cf. Aristotle, Nic. Eth. 1168b6 f., e.g.), since Horace does not say alter idem here and we owe it to him to pay attention to what he does say, not what his language suggests by way of commonly accepted cultural tags. See Nisbet and Hubbard 1970.48 and the bibliography found here.

20. I follow Basto 1982.31 here.

21. Ibid.

22. Ibid.

23. See generally Nisbet and Hubbard 1970.48.

24. See Pucci 1988.75-80.

25. Fraenkel 1957.409 is coy in noting, in a wholly ambiguous phrase, that there was a “wide difference in the constitution and temperament of the two friends. …” Just how friendly can we suppose two people of widely differing constitution and temperament to be? Just how friendly could they have been in maturity, regardless of an earlier, more youthful, connection?

26. Thus Lewis and Short 1879.581, s.v. dimidio.

27. Lewis and Short 1879.1401, s.v. portio.

28. Technically, poro is presumed to be the present form of the aorist active eporon, on which see Liddell and Scott 1940.1452, s.v. poro.

29. On this specifically see Lewis and Short 1879.1401, s.v. portio, II.

30. In fact, I would suggest that reliance upon what seems to be a Greek prototype has led to a misinterpretation of the function of this phrase. Some commentators, such as Villeneuve 1944, have felt that the phrase was too obvious to warrant comment. The opinio communis has invariably held that dimidium is to be read as a tender word. Shorey 1899.153 reads dimidium as explicitly synonymous with pars in 2.17 (cf. p. 283). Kiessling 1930.21 is not as bold in asserting the similitude but generally sees the phrase as comparable to that found in 2.17. Page 1962.142 gives a fuller history of the references to Virgil in Horace's poetry. Smith 1904.14, following earlier commentaries, notes that the phrase is “borrowed” from an old Greek definition of friendship and Diogenes Laertius 5.1.20 is seen to be exemplary: miá psuchè dúo sòmata enoikoûsa. Here is where the misinterpretation might lie, since Horace's Latin is not a translation of the Greek. It is one thing to rely verbally on a phrase from a prior text (i.e., to allude), and another thing simply to use a common idea that is part of a culture. Horace is doing the latter here and it cannot be denied that he creates a new phrase that stands out noticeably within his oeuvre and works to other than positive purposes in the meaning of this ode.

31. See Lewis and Short 1879.581, s.v. dimidius, III A.

32. Fraenkel 1957.408-410.

33. Nisbet and Hubbard 1970.40 offer a tripartite division as does Collinge 1961.83. Kiessling 1930.19 offers a four part division. I arrived independently at the same kind of division as that offered by Carrubba 1984.168-169 and agree with the ways he would see the ode structured.

34. See Basto 1982.40 n. 5.

35. There is nothing new in reading the language of the ode in this way; see, for example, Kidd 1977.97-103. Reading the negativity that follows the introduction of this metaphor as part and parcel of the metaphor itself is, however, new and something that I think must be done here.

36. It should be noted that even in this regard Horace deviates from the norm. A propempticon should pray for a safe return. Horace prays here for a safe passage, thus highlighting again the disjunction of form to content. This does imply a metaphorical, poetic level at which the poem must be read, but does it not also imply that the poet may not be returning from his travels, just as Hercules, Prometheus, and Daedalus did not, a position with which Ovid would seem to agree in his own meditations on Virgilian epic.

37. On the issue of priority and dating, I follow Basto 1982.36-37. Virgil died only four years after Horace published Odes 1-3 and, although the date of 1.3 is uncertain, it is clearly not, as some have held, one of the earlier odes of the collection. By 23 b.c., in any case, the first two-thirds of the Aeneid would have been in presentable form (all of the allusions, incidentally, to Virgil in Horace come from books 1-7), so that Horace would have a very good idea of what the Aeneid was—in his view—about.

38. The would-be Pindaric poet is specified, not coincidentally, at Odes 4.2.10 as an audax poeta. Cf. the opening images of this ode also.

39. Odes 3.4.56, 3.18.13, 3.27.28; Epistles 2.2.51; Satires 1.1.30, 2.3.165, 2.5.29, 1.10.76; Ars Poetica, v. 237.

40. The metaphor is used by Brown 1967.76 to describe the literary artistry of the Confessions.

41. Augustine has changed the word order for purposes of stylistics, since Horace's phrase had to fit into an asclepiad in 1.3.

42. The phrase is Augustine's in De Catechizandis rudibus 10.15.

43. See, for example, De Magistro 5.14, 8.22 ff., or De Catechizandis rudibus 10.15 among other examples.