‘[T]he vast majority of Horace's Odes contain little trace of narration’.Footnote 1 Yet, as Lowrie demonstrates, Horace can present ‘stories’ without actually narrating, or telling, them. This article examines two such instances in the Roman Odes (Carm. 3.1–6). In each case, the story is embedded in a larger rhetorical structure that is not narrative; and in each case the picture that emerges is elaborated and embellished over three stanzas and acquires a status of its own as a story that captures the imagination.
The first of these stories, at Carm. 3.2.1–12, presents the transformation of a boy from puer to warrior. This story unfolds within the rhetorical frame of an exhortation, which leads to the gnomic declaration – an echo of Tyrtaeus – that ‘it is sweet and seemly to die for one's country’ (dulce et decorum est pro patria mori, 3.2.13).Footnote 2 The second, contrasting story, at Carm. 3.6.21–32, details the transformation of a virgo from a young girl to a jaded wife who is pimped openly by her husband. The story is developed as an exemplum to give emotional colour to the preceding claim that generations fertile in debauchery have defiled marriage and the family (3.6.17–20). This general claim plays a causal role within a more intricate argument of causes and consequences in the ode.Footnote 3
Although the rhetorical structures – exhortation versus exemplary illustration – in which these two stories are presented can be seen therefore to differ in their aims, the stories themselves offer comparably compressed accounts of a bewilderingly rapid transformation from youth to maturity. In neither case do we see the entire life of the young person; we are given only glimpses of her or his education on the verge of adulthood, and the final result. Yet a larger life history is implied in both stories.
In the first part of this article, I present the stories. In the second part, I analyse aspects of each story's intertextual resonances, arguing that in Carm. 3.2 Horace draws on Homeric similes and alludes to Homeric and Vergilian narrative and stylistic techniques to construct his own miniature epic; the characters in his battle scene are sketched impressionistically from the characters who play prominent roles in the second half of Vergil's unfinished Aeneid. In the sixth Roman Ode, with allusion to an obscene epigram by Automedon, Horace converts his Horatian mini-Aeneid into its opposite by replacing the story of the puer's development with the startlingly contrasting story of a virgo's development and downfall. His allusion to the sleazy epigram arguably reproduces a textual corruption that has been detected and which – along with a second, previously undetected corruption – is still preserved in the epigram that has come down to us. In conclusion, I speculate that Horace's peculiar reproduction of the textual error serves as a metatextual comment on the uncurated quality of the source text itself and, by extension, on its depraved moral content. The contrast with Vergil's carefully protected and curated text and lofty morality could not be stronger. Horace's skilful variation and reconfiguration of his own material in the two Roman Odes highlights his lyric virtuosity and links these sections of two very different odes, 3.2 and 3.6, stylistically as a study in contrasts.
Before I embark on my detailed discussion, two preliminary observations must be made. First, I do not mean to suggest that the Vergilian colouring I attempt to tease out in the first three stanzas of the second Roman Ode is the only, or even the most prominent, intertextual resonance to be felt in this poem. Echoes of Tyrtaeus and other Greek archaic poets and of Simonides and Pindar have long been noted.Footnote 4 In the ode as a whole, it is possible to discern a shift of focus from Greek military elegy to hymnic lyric. I have previously argued that the rhetorical mode characteristic of the martial poetry of Tyrtaeus and Callinus – in which exhortations to bravery on the battlefield alternate regularly with gnomic declarations reinforcing those exhortations – gives way after the first four stanzas to a Pindaric-Simonidean mode, in which the relationships between gnome and theme are varied and more elusive or harder to define.Footnote 5 I concluded that ‘the juxtaposition of the worlds of military elegy and hymnic lyric motivates Horace's ode, and the movement from one to the other constitutes a question’.Footnote 6 Within this larger rhetorical dynamic, the Vergilian colouring may be felt as a minor theme that gives depth to this questioning without undermining or supplanting the main themes.
My second preliminary remark is offered in defence of my admittedly unconventional approach to the relationship between Horace's Odes and Vergil's Aeneid. The bold claim that Horace was acquainted with the basic structure, and in some cases even with the text, of Vergil's unfinished Aeneid will, on chronological grounds, be greeted by many with an understandable skepticism. Suetonius famously informs us that in 19 BC Vergil, on his deathbed, insisted in vain that the Aeneid be brought to him to be burnt when he recognised that he would not live long enough to give it its finishing touches, a labour he had anticipated would take three more years. Moreover, Suetonius reports, it was only in 23 BC – the year Horace's first three books of Odes were published – that Augustus himself was at last privileged to hear Vergil read aloud Books 2, 4, and 6 of the Aeneid, although he had long been pleading with and even demanding that Vergil let him see a ‘first draft of the poem, or any portion of it that he was willing to share’.Footnote 7 My claim, based on a careful consideration of the texts themselves, that Horace had specific knowledge of some passages of the Aeneid and of its larger plan is therefore bound to be controversial. Yet I believe that this claim can be reconciled with the anecdotes and evidence preserved by Suetonius. Against the skepticism engendered by chronological considerations, I would posit that the particular ways in which Horace engages with the unfinished Aeneid operate on several distinct but interacting levels, which may be defined as intertextual, metatextual, and – to introduce a new term – extratextual. I hope to show that these different forms of engagement with Vergil are ultimately implicated, albeit indirectly, with themes that are recognisable elsewhere in Horace's poetry and in the Roman Odes themselves: namely, Horace's withdrawal from the common crowd and his confidence in the lasting value of his poetry, with the expectation of readers far into the future.
1. The Stories
The Second Roman Ode
To put up with straitened Poverty as a friend – a boy must learn to do this, a boy made tough through strict military training; and let him harass the fierce Parthians as a horseman to be feared for his javelin; and let him lead his life under the open sky in dangerous circumstances. Looking out on him from the enemy's battlements, let the wife of the warring monarch and the girl, grown up, sigh, Ah! lest the princely betrothed, not yet trained in the line of battle, provoke the lion, savage when touched, which bloody Rage is sweeping through the very heart of the slaughter.
The second Roman Ode opens with a personification that – if it does not slip by the reader unnoticedFootnote 8 – is startling: an image of straitened Poverty treated as if she were a friend. This personification provides a link to the memorable personifications of the first Roman Ode, where Fear (Timor) and Forebodings (Minae, 3.1.37) climb as high as the owner does when he ascends to the uppermost prospect from his lofty seaside mansion, built up to jut into the water; and where dark Care (atra Cura, 40) is a constant companion on the rich man's yacht and – in a haunting image – is always sitting behind the horseman as he rides. The stylistic echo of these personifications in the figure of Poverty encapsulates the ethical concern with the folly of greed and the burdens of wealth with which the previous ode concluded. For the (re)reader of the Odes, the personification also echoes, or is echoed by, the more striking personification of Poverty developed in Carm. 3.29, a poem that presents itself as an invitation to Maecenas to put aside his political cares and the distractions of wealth and, like Horace, to cultivate an attitude of equanimity. In this prominently placed ode, the second-to-last of the collection, Horace dwells for two stanzas (3.29.49–56) on the cruel capriciousness of Fortuna. If Fortuna flaps her wings and abandons Horace, ‘I pay back what she has granted me, wrap myself up in my virtue [as if it were a cloak],Footnote 9 and go courting honest Poverty, who has no dowry’ (resigno quae dedit et mea | virtute me involvo probamque | Pauperiem sine dote quaero, 54–6). The (re)reader's perception of the structural balance between the second and penultimate odes of Horace's third book confirms, in hindsight, the philosophical implications of the first verse of Carm. 3.2 that I attempt to tease out here.
This verse thus bears a philosophical burden that may seem disproportionately heavy in an ode where such ethical concerns – an embrace of the simple life and the cultivation of tranquillity in the face of unpredictability and adversity – play no further role.Footnote 10 The abandonment of these concerns signals the rapidity of the transformation that the boy must undergo. From the image of straitened Poverty we move quickly: in the second verse we learn that the boy (puer) who emerges as the subject has been ‘made tough through strict military training’ (robustus acri militia). In both grammar and sense, the finite verb in the third verse, condiscat, looks back to the boy's ethical education (angustam amice Pauperiem pati), but his progression from philosophy to boot camp is so rapid that these successive stages of his young life are sandwiched together before the verb. The movement from his philosophical higher education (condiscat) to his participation in a foreign military campaign (et Parthos feroces …) takes place in a single line (3).Footnote 11 This third verse is grammatically completed by vexet eques metuendus hasta (4), and with the restatement of the subject (from puer to eques) a subtle shifting of identity begins in a development that culminates with leonem (11). The boy's transformation may be seen in three stages: from a human puer, he becomes an eques, a person whose function is defined in relation to an animal; finally, he becomes an animal himself (leonem). In a parallel movement, from subject (puer, eques) to object (illum, 6; leonem), he loses his agency as well as his humanity. His encounter with a second personification signals closure; we have reached the end of his story. The boy who was enjoined to make a friend of Poverty is ultimately seen as the victim of bloody Rage (cruenta … Ira, 11–12), who hurtles him (rapit, 12) through the carnage of battle.
The Sixth Roman Ode
A girl, (now) grown up, is delighted to be taught Ionian dance moves; even now she's moulded in (those) arts, and from earliest infancy she obsessively plots illicit love affairs. Soon she's looking for younger adulterers in between her husband's quaffs, and she doesn't (bother any longer to) choose one to whom she'll give forbidden pleasures, in a rush, when the lamps have been taken away; but, propositioned in the presence of everyone, and with her husband's connivance, she gets up (from her dinner couch), whether it's a travelling salesman who has put in his order (for sex), or a Spanish sea captain, a big-spending buyer of sleazy degradations.
The three stanzas of this exemplum (stanzas 6–8 in the ode) sketch the progression of a life from girlhood to maturity. In this barely narrated story, we find only two indicators of time: iam nunc (23, in stanza 6), of her earliest education in seduction, and mox (26, in stanza 7), of the depravity that is to come. This gesture towards narration is undercut by Horace's use of present-tense verbs throughout the description, a stylistic feature that communicates the rapidity of the young woman's degeneration, as if the course of her life is rushing by in a single moment. Yet, while ‘now’ and ‘soon’ seem to point to two phases of decline within this eternal (and universalising) present, in Horace's terse description there are implied a number of stages. Distinct phases of her life as a girl (virgo) are tumbled together and conflated in stanza 6. The adjective matura (22), indicating that the young woman is ready for marriage, is placed in a verbal context that refers to a period of childhood when she is still being taught lessons (doceri, 21) and her unformed character is still being shaped (fingitur, 22). In the second half of the stanza the discrepancy between her activities and her stated age is even more pronounced: de tenero … ungui (24) means ‘from earliest infancy’, but the girl cannot possibly have been scheming to have adulterous affairs since she was a baby, and in her preoccupation with her sex life she seems now to have lost her virginity.Footnote 12
In stanzas 7 and 8, a sequence of disparate events is compressed into the impression of a single dinner scene that conveys the wider contours of the story of her marriage. While matura virgo (22) remains the grammatical subject, mox (25) points to a time when the woman, now a wife, seeks ‘younger adulterous lovers’ as her husband slugs down another drink (inter mariti vina, 26). What older person is being compared to these younger men (iuniores, 25)? Most obviously, her husband; but the comparison could as well imply a string of earlier lovers. The negative formulation neque eligit cui donet impermissa … gaudia (26–8), in telling us what she does not now do, may reveal what she used to do in the past, in singling out a man at dinner with whom she would sneak in some quick sex after the lamps were taken away. This choosing is contrasted, however, with her present submissiveness. In this final stage of her degradation, she is openly summoned from the dinner couch, her husband fully complicit, to have sex with any lowly salesman who happens to be passing by, or with any rich sea captain who is willing to pay a high price (pretiosus emptor, 32) for such degradation (dedecorum).
2. Intertextual Resonances
The Second Roman Ode
Make way, you Roman writers! Make way, you Greeks! Something greater than the Iliad is coming to birth.Footnote 13
The most fascinating intertext in the story presented in the second Roman Ode may not (yet) even have been a complete text. In the years when Carm. 1–3 were being composed and at the time of their publication in 23 BC, Vergil's Aeneid was at least four years away from its publication, shortly after Vergil's death, in 19 BC. Commentators have been loath to consider the possibility of the Aeneid's influence on the Roman Odes in anything other than the most sweeping terms mainly for this reason.Footnote 14 When we consider that traditional approaches to intertextuality try first to determine the direction of textual influence and then to interpret how the source is imitated, complicated, ‘corrected’, or otherwise received, this caution is understandable in light of the uncertainty regarding the poems’ respective chronologies. And yet in the case of an epic whose composition extended over the course of (perhaps) eleven or twelve years, an epic which – as Propertius testified in 26 or 25 BC – was famously taking shape, even if it had not yet been released to the world, a more fluid approach is needed.Footnote 15 It is not necessary to suppose that Vergil had yet finalised in verse the later books of the Aeneid, which told of the war in Latium, in order to admit the possibility that Horace may nevertheless have been acquainted with the contours of its emerging plot and, perhaps, some textual details. Nor is it necessary for Horace to have had a meticulous knowledge of these later books for him to have acquired an understanding of Vergil's strategies of characterisation, his engagement with Homer, and other distinctive features of his narrative style. These techniques were already on display even in the first book of the Aeneid. From this perspective it is not a question of defining the direction of influence from Vergil to Horace or from Horace to Vergil; it is enough to posit Horace's response to elements of the larger story of Aeneas that was taking shape in Vergil's epic and to demonstrate the mutual influence on Vergil and Horace of Homeric material that is relevant to that larger story. There is an element of poetic playfulness in Horace's double engagement with Homer and Vergil: by alluding to a Roman epic that, according to Suetonius’ account, had not yet been made public through reading, even in a limited way, within Augustus’ inner circle,Footnote 16 Horace slyly usurps Vergil's partly told story and makes it his own. His artful appropriation is a tribute to the Aeneid which, precisely because no announcement is made and no names are named, presents itself, perhaps deliberately, as the opposite of Propertius’ heraldic prediction.
Horace took upon himself the daunting challenge of encapsulating an epic narrative within a tightly woven, small-scale, and nonnarrative lyric mode. He met this challenge by alluding to epic narratives without narrating and by evoking characters, situations, or images in passages that occupy critical positions in the larger-scale structures of the Aeneid and Iliad. With the ode's transition to a scene set on the walls of a city (illum ex moenibus hosticis, 6), the outline of a landscape populated by characters comes into view, setting the stage for a narrative of a traditional type-scene in epic and tragedy in which the inhabitants of a besieged city look out from the ramparts over a battlefield. This cast of nameless but individually delineated characters springs into life when they are recognised as actors in the world of the Aeneid. Rather than narrating a background history, however, or, alternatively, a sequence of events, as these might be presented in Vergil's epic, Horace offers a static tableau in which the actions of the characters are only impressionistically represented. The imagery of Horace's scene summarises Vergil's ‘Iliadic’ Aeneid (Books 7–12) with bookending clues that point both to its highly marked proem, in which these characters enter the narrative, and to its dramatic conclusion with Turnus’ death.
After Aeneas’ departure from the underworld at the beginning of Book 7, Neptune's favouring winds blow his ships to the shores of Italy and they enter the waters of the Tiber. Here Vergil pauses his narrative and, invoking the Muse, makes a programmatic declaration that is positioned at the very turning point from the Odyssean to the Iliadic half of his narrative: ‘A grander sequence of events is emerging for me; I embark upon a grander work’ (maior rerum mihi nascitur ordo; | maius opus moveo, Aen. 7.44–5). Propertius’ nescio quid maius nascitur Iliade seems to echo Vergil's wording, taken by many to be a clear signal that Propertius – even if not Augustus – had at least some privileged knowledge of Vergil's text and a positive sign that Vergil was working on the composition of Aeneid 7 in the years when Propertius’ second book of elegies and Horace's Odes were being written.Footnote 17 Horace's own allusion to Vergil's ‘Iliadic’ proem looks not to the claim itself but to the narrative that follows, in which Vergil introduces the characters and motivating forces that will drive the events of the war soon to engulf Italy. Latinus, king of Latium, had one living child:
Only a daughter dwelt at home holding that palatial seat, a daughter now ripe for a husband, now marriageable in years. Many had been wooing her from wide Latium and from all over Ausonia; the most beautiful, beyond all others, who woos her is Turnus, powerful in his ancestral lineage, whom the royal wife had been hastening – with a passion to inspire wonder – to have united with her as son-in-law; but terrifying omens from the gods are standing variously in her way.Footnote 18
Horace's characters – the king at war; the watchful queen; their daughter, ripe for marriage (adulta virgo, 3.2.8; cf. iam matura viro, Aen. 7.53); and the betrothed prince on the battlefield – comprise the same cast as the characters introduced by Vergil in his ‘Iliadic’ proem, but their attributes, situations, and attitudes point to the very end of the Aeneid rather than to the beginning of Vergil's ‘Iliad’.Footnote 19 The war foretold in Vergil's proem is already raging in Horace's ode, and Horace's briefly sketched image of women watching fearfully from the battlements conveys a situation analogous to, although also strongly contrasting with, Vergil's action-packed narrative of the panic and confusion that break out within and outside the city as Aeneas approaches, as Latinus withdraws in despair and self-reproach, as women and children line the walls, and as Turnus arms himself eagerly for the final fatal encounter with Aeneas (Aen. 11.468–97). The transformation of Horace's ‘Aeneas’ into a figure who has lost his humanity (leonem) and is carried away by rage (quem cruenta | per medias rapit Ira caedes) parallels Aeneas’ shocking devolution away from pietas – forgetful of Anchises’ admonition to spare the conquered – as he himself is ‘inflamed by fury, terrible in his rage’ (furiis accensus et ira | terribilis, Aen. 12.946–7) when he plunges his sword into Turnus’ heart in his final act in the Aeneid. Although Vergil's Turnus is conspicuous for his violentia, Horace makes the transformation of his Horatian ‘Aeneas’ more pointed through contrast by simplifying and reversing the Vergilian characterisation of Turnus. He presents his Horatian ‘Turnus’ as the betrothed rather than as the rejected suitor and portrays him as unwarlike and vulnerable in the face of the lion's savagery. Horace's characters, then, do not in all respects duplicate Vergil's. The adjustments facilitate lyric compression – Horace resists showing character development as it would be presented in stages in an epic plot – and they illustrate one of the ways in which Horace takes over Vergil's developing story to tell it in his own way.
Horace employs a number of stylistic strategies to avoid narration. Of the three nameless characters introduced in two dense lines – the wife (matrona, 7) of a king (tyranni, 7) and a young woman of an age suitable for marriage (adulta virgo, 8) – the queen and king barely act. Their involvement in the scene is conveyed through present participles. The matrona is stationed on the battlements, looking out (prospiciens, 8), while the king's location and movements are not specified. We are not told whether he is on the battlefield, on the ramparts, or somewhere inside the city; we know only that he is a king at war (bellantis tyranni, 7), an attribute that conveys his ongoing state or condition rather than his actions and that gestures towards Latinus’ inability to take effective action in the Aeneid. The matrona lacks her own finite verb: in the zeugma matrona … et adulta virgo | suspiret (7–9), the singular verb suspiret must do grammatical duty for both subjects, although in sense – according to the usual Latin construction in prose – the verb belongs only to one of these subjects (virgo).Footnote 20 Moreover, the landscape in which these characters are fixed remains embedded in the grammatical framework of the injunction. ‘Let her sigh’ oddly culminates the string of earlier injunctions (‘let him learn’, 3; ‘let him harass’, 4; ‘let him lead [his life]’, 5), revealing the emotions of a character without becoming the narrative of an actual event: Horace does not say ‘she sighed’ or ‘she will sigh’ but rather ‘let her sigh’. While eheu verges on narration by delivering the sound of the sigh, the words that follow cannot be taken as a direct quotation of speech in any ‘realistic’, or even in any epic, world. Nor do these words, which describe the betrothed and the lion, give a literal rendition of the character's fears, as Heinze recognised: ‘What follows is of course not direct speech, but neither does it directly reproduce the women's fearful thoughts; rather, it converts them into the poet's own conception: only he can contrast the two warriors in this way’.Footnote 21 If we rephrase Heinze's remarks in narratological terms, we can recognise a distinction between the poet who speaks and the character who sees,Footnote 22 that is, between the voice of the poet, who structures his discourse with the use of zeugma (matrona … et … virgo suspiret, 7–9), transferred epithet (cruenta … Ira, 11–2), metaphor (leonem, 11), and other poetic figures and the feelings, judgments, and perceptions of a character or characters within the story. The discourse is the poet's, but the words describing the betrothed and the lion focalise the emotions of the characters and contribute to their characterisation.Footnote 23 Horace's deployment of this stylistic device constitutes something of a paradox: the focalisation allows him to avoid narrating even while he adopts, from Homer and Vergil, a distinctive narrative technique. The zeugma matrona … et … virgo suspiret works along with the focalisation to provide further characterisation. Although logic and the rules of prose insist that suspiret properly applies only to virgo, the verb stands in as a substitute for whatever else the matrona … prospiciens is doing. Some of its emotional colouring inevitably infects her.Footnote 24 The figures and techniques Horace uses to structure his discourse are therefore not merely ornamental. They say more than they seem to. Each encodes a narrative or exemplifies a narrative technique that alludes to Vergil's unfinished epic.
Three of these figures – a metaphor, a transferred epithet, and a striking personification – are particularly ingenious in that they achieve their Vergilian resonance not through adaptation of the characters, images, or plot structures that would later be encountered by readers of the Aeneid, but through reference to narrative contexts in Homer that have far-reaching significance for the Aeneid. To encapsulate this broader Vergilian narrative, Horace colours his allusions to Homer with Vergilian overtones and interpretive implications. Horace's lion metaphor draws on the lion similes that describe enemies of Aeneas in two extended and roughly parallel episodes in the Iliad in which Aeneas narrowly escapes death; while his personification of bloody Rage refigures a stylised personification depicted on Achilles’ shield. Horace's allusion to a work of art in Homer must also, however, be understood as a gesture towards the Aeneid, where the descriptions of works of art – such as the images of the Trojan War on display in Juno's temple in Carthage, the reliefs carved by Daedalus on the doors of the temple of Apollo in Cumae, and the images on the shield of Aeneas wrought by Vulcan – raise questions of interpretation and interpretability both for the characters within the narrative and for the reader.
From a Vergilian perspective, Aeneas’ encounter with Achilles at Il. 20.75–352 bears a special significance. In this long episode, one of several that interrupt the narrative of Achilles’ aristeia (which began with his arming in Book 19 and will end with the death of Hector in Book 22), Achilles taunts Aeneas, who responds with an account of his divine ancestry. The genealogy ‘recapitulates the history of Troy and its close associations with the gods at a time when the city's doom is rapidly approaching’.Footnote 25 This confrontation between the heroes thus provides an excuse for the narrative to dwell on the mythological foundation that serves as the background against which the epic events unfold. For Vergil, one passage in particular within the Homeric narrative of the clash could have been read tendentiously as a ‘prediction’ of his own epic narrative. At Il. 20.300–8 Poseidon, seeing that Aeneas is about to be killed by Achilles, reminds the other Olympians that Aeneas is not fated to die; rather, ‘it is his destiny to escape, so that the race of Dardanus not perish without seed, blotted out. … But as it is, mighty Aeneas will rule among the Trojans, as will the children of his children who are born afterwards’.
The centrality of Aeneas and the forward-looking emphasis on his fated destiny in this Homeric passage makes Horace's allusion to one of its finer details – a simile (Il. 20.164–73) – also, indirectly, an allusion to the Aeneid as a whole. Warriors are frequently compared to lions in the similes of the Iliad, and in several of the more developed of these similes a predacious lion is described defending itself against shepherds, bands of men, or dogs. In only two of these similes (Il. 5.136–42 and 20.164–73), however, is a lion roused to greater ferocity after having been wounded – as he threatens to be in Horace's metaphor, where the warrior-lion is provoked, through contact (tactu, 11) with a harassing weapon, into a violent rampage.Footnote 26 A reader's suspicion that this detail of imagery in Homer may point, for Horace, specifically to the Aeneid increases with the recognition that the earlier Homeric simile of the wounded, enraged lion also occurs in a narrative context (Il. 5.111–362, 431–53, 512–8, 561–75) in which Aeneas plays a critical role. Although neither of these unique similes duplicates, or even echoes, the other in its vocabulary or formulae, in many ways the broader context of the earlier narrative prefigures and prepares for the later one, which may perhaps be regarded as a kind of doublet: in both narratives, the simile characterises an enemy (Diomedes in Book 5; Achilles in Book 20) whom Aeneas confronts; in both, the gods intervene to save Aeneas by spiriting him away from the battlefield (Aphrodite, followed by Apollo, in Book 5; Poseidon, in consultation with Hera and other gods, in Book 20). Horace's allusion, through his imitation of Homer's similes, to Vergil's Aeneid may therefore be said to be extratextual. The reader's awareness of this double extratextual reference directs our attention away from the differences between the verbal formulae and particulars of each Homeric encounter (Aeneas against Diomedes; Aeneas against Achilles) and toward the overarching structure of the plot of Vergil's ‘Iliadic’ Aeneid (Aeneas against Turnus) and to the details that are relevant to that plot or that are repeated or shared in the Homeric similes: the motif of the wounded lion roused to greater fury and the application of each simile to an enemy of Aeneas. While Vergil drew on these very similes from Homer to characterise Turnus’ violentia at Aen. 12.4–8 and to foreshadow his death, Horace's application of the wounded lion similes to his ‘Aeneas’ rather than to his rival, the sponsus, accomplishes for the ode the same unsettling reversal that Vergil achieves through plot development.Footnote 27 This lyric inversion foreshadows, so to speak, or more precisely substitutes for the epic narrative of the increasing furor that finally overpowers and consumes Vergil's Aeneas.
Horace's artistry in transforming Vergil's epic narrative into a lyric discourse that is not narrated is further demonstrated in the Homeric allusion encapsulated in Horace's cruenta … Ira (11–2). The wounded lion in Homer's simile of Achilles is said to be ‘borne straight ahead by rage’ (ἰθὺς φέρɛται μένɛι, Il. 20.172). Horace picks up on this detail but takes it one step further in his personification of Ira, whom he characterises as ‘bloody’ with the addition of a transferred epithet that more properly applies to the carnage (caedes, 12) than it does, in a literal sense, to rage. In the Homeric simile, by contrast, the rage is not modified by an adjective, and it is figurative rather than personified: in the Iliad, personified abstractions interact with humans only in exceptional circumstances, only when undetected by humans, and only when sent by Zeus.Footnote 28 Yet the image of a bloody personification seizing and hurtling men through carnage is, in fact, Homeric, even though it does not occur in the narrative of an actual battle. This ghastly abstraction is represented in a figurative and, in Homer's terms, nonrealistic battle scene wrought by Hephaistos on the shield of Achilles (Il. 18.535–40). Here, Ἔρις (Strife), Κυδοιμός (Tumult), and ὀλοὴ Κήρ (deadly Fate) were shown fighting among men as if they themselves were mortal (ὡμίλɛυν δ’ ὥς τɛ ζωοὶ βροτοὶ ἠδ’ ἐμάχοντο, ‘and they joined in battle and fought like living men’, 539), one holding a man who had just been wounded, another an unwounded man, while the third, deadly Fate, dragged a corpse ‘by his feet through the press of battle’ (κατὰ μόθον, 537; cf. per medias … caedes, Carm. 3.2.12), and she ‘wore around her shoulders a garment crimsoned with men's blood’ (ɛἷμα δ’ ἔχ’ ἀμφ’ ὤμοισι δαφοινɛὸν αἵματι φωτῶν, 538; cf. cruenta … Ira, Carm. 3.2.11–2).
Horace's allusion to Homer's ekphrasis of Achilles’ shield does not necessarily constitute an allusion to Vergil's ekphrasis of Aeneas’ shield at Aen. 8.608–731, a passage that, after all, may not yet have been composed. The relevance to the Aeneid of Horace's Homeric allusion is simultaneously more diffuse and more pointed. Unlike the moments in the Aeneid's grander narrative that are relevant to Horace's scene on the walls – the introduction of Vergil's Italian characters in Aen. 7 and Aeneas’ final act of furor in Aen. 12 – Horace's allusion to Achilles’ shield points not so much to the plot or structure of the Aeneid as it does to Vergilian questions concerning interpretation and interpretability. Within a narrative context, the ekphrasis of a work of art is, almost inevitably, bound up with the reception of that art by an interpreting character who views it. A foundation for this stylistic technique of characterisation was already laid by Homer, who differentiates the reactions of the Myrmidons and of Achilles to Achilles’ armour depending on their respective abilities even to look at and to understand what they see. The Myrmidons – ordinary mortals – are overcome by trembling (τρόμος, 14) and shrink back in fear (ἔτρɛσαν) at the very sound of the clashing armour as Thetis places it before Achilles; not one of them dares to look directly (ἄντην ɛἰσιδέɛιν, Il. 19.15) at the divinely wrought handiwork, which Homer presents as beyond their ability to comprehend. Achilles’ eyes, by contrast, respond in kind by flashing with their own inner fire, which is compared to a sign, signal, or omen (like lightning, a fiery beacon, or a meteor: ὡς ɛἰ σέλας ἐξɛφάανθɛν, 19.17) when he sees the armour. He, son of an immortal, has the ability not only to satisfy his heart by gazing at the armour at length (ἐπɛὶ φρɛσὶν ᾗσι τɛτάρπɛτο δαίδαλα λɛύσσων, 19) but also to recognise and acknowledge the artefacts as the work of a god.
Vergil's Aeneas is both like and unlike Achilles in his reaction to Vulcan's armour in Vergil's description of the shield. He is like Achilles in that he is able to gaze at length upon what he sees and to rejoice in it (Aen. 8.617–25). But Vergil problematises Aeneas’ interpretation in that the images themselves bear a meaning that he cannot comprehend, as they refer to Rome's (to Aeneas, unknown and unknowable) future. This problematisation introduces an additional layer of complexity by drawing the reader into the text to complete, or complicate, the act of interpretation. The reader knows the Roman history that Aeneas does not know and can bring to bear on the artefact described in the text knowledge or a perspective that the interpreting character within the text lacks. A particularly eloquent development in the Aeneid of an ekphrasis that signals this analogous relationship between the interpreting character and the reader occurs in the first book.Footnote 29 Aeneas’ tearful declaration to Achates that the images of the Trojan War displayed in the temple of Juno in Carthage are motivated by a noble sense of humanity is at odds with the narrator's stark description of events whose representation in the temple offers an account that is, if anything, bleaker even than the reality Aeneas himself remembers when he later describes the fall of Troy to Dido.Footnote 30
Horace's intertextual allusion to a Homeric work of art may therefore be understood as yet another extratextual allusion to Vergil's Aeneid and, further, to the problems of interpretation and interpretability embodied in that text. The ode's lion metaphor reverses the larger Homeric narrative by applying the imagery to Horace's ‘Aeneas’ figure rather than to his enemy. The problem raised here is one of interpretation: what does it mean for the enraged lion to be Aeneas rather than Aeneas’ enemy? Horace's allusion to a Homeric work of art, in turn, raises the Vergilian question of interpretability. But the ode, unlike the epic, offers the reader neither a narrative nor an interpreting character. Its allusion to an epic ekphrasis serves as a signal that the work of interpretation must be done by the reader. How is the reader to understand Horace's own problematisation of dulce et decorum in this ode?
The Sixth Roman Ode
In its intertextual echoing, the contrasting story of the virgo in the sixth Roman Ode takes us from the sublime (Homer and Vergil) to the seamy (Automedon) in the form of an allusion to an obscene epigram, Anth. Pal. 5.129, whose text was probably already corrupt in Horace's own time.Footnote 31 In fact, as it has come down to us this short epigram arguably contains two textual corruptions. A literal translation exposes the difficulties of the received text.
The dancing girl from Asia, who moves with lascivious poses from her earliest infancy (?), I praise not because she shows passion in her every gesture or because she tenderly positions her tender hands this way and that, but because she knows how to dance around my worn-out peg and doesn't run away from an old man's wrinkles. She tongues, she tickles, she hugs; when she throws her leg upon (…?), she can bring my club back from the dead.
The meaning of ἐπιρίψῃ (7), ‘cast at, throw (something) upon’,Footnote 32 is problematic both grammatically – one would expect an indirect object with this compound verb – and as a description of what the dancer is doing with her (?) leg. Her previous actions in this verse (tonguing, tickling or teasing, and hugging) are all exercised on the old man's body. After these intimate details of foreplay, it is not easy to see how the mere kicking up of her leg could serve as the climactic gesture that brings the old man's club back from the dead. An effective remedy for the text's, and the old man's, problem is found if we emend ἐπιρίψῃ to ἐπιτρίψῃ and translate ‘when she rubs my thigh …’.Footnote 33 Another epigram by Automedon provides details that clarify the sexual significance of this act. In Anth. Pal. 11.29.3–4 Automedon worries about impotence: αὕτη γὰρ λαχάνου σαθρωτέρη ἡ πρὶν ἀκαμπὴς ζῶσα νɛκρὰ μηρῶν πᾶσα δέδυκɛν ἔσω (‘this thing of mine – more flaccid than a [wilted] vegetable – which used to be rigid when it was alive, has sunk entirely into my thighs, a corpse’). In 5.126 he faces a similar problem. Although his worn-out peg has presumably retreated into his thighs, when the dancer rubs his leg (σκέλος) she can bring it back – not as a peg, but as a club, no less! – from the realm of the dead. The physical stimulation of his σκέλος, rather than the performance of a sexy dance move in which she kicks up her σκέλος, will overcome his impotence.Footnote 34
With this emendation Automedon's epigram falls neatly into two thematic halves. The last four lines describe the dancer's sexual manipulations of the old man's body, while the first four lines concern her dancing moves. Within this context the phrase ἐξ ἁπαλῶν … ὀνύχων sits as awkwardly as Horace's de tenero … ungui (Carm. 3.6.24) sits within the ode, and for a similar reason: in both Greek and Latin the idiom ‘from (the time of her) tender fingernails’ has the troublesome meaning ‘from (her) earliest infancy.’Footnote 35 While Horace exploited this peculiar idiom for poetic effect by compressing the various phases of his virgo's life into a tight jumble of scenes in order to create the impression of her very rapid development and downfall, the thematic structure of Automedon's epigram admits no such poetic justification. The phrase ἐξ ἁπαλῶν κινυμένην ὀνύχων is as nonsensical within the epigram (‘moving … from her earliest infancy’) as it is gratuitous, since the old man is otherwise engrossed wholly in his own pleasure and betrays no actual curiosity about the dancer herself except as a talented sexual entertainer. For these reasons, Nisbet's tentative emendation to ἐξ ἄκρων κινυμένην ὀνύχων (‘quivering from the tips of her fingernails’) is persuasive: ‘perhaps at an early stage ἄκρων was corrupted to ἁπαλῶν under the influence of ἁπαλὰς ἁπαλῶς below’.Footnote 36
Why did Horace choose to imitate the very phrase that, both in the pointlessness of its content and in its difficulty to construe, contributes the least to the epigram's overall argument and thematic structure? The most economical answer would be that the conspicuous interpretive difficulty introduced with the corruption is a memorable textual feature in an otherwise rather unremarkable, if clever, epigram. Horace's echoing of the difficult text unmistakably signals his allusion to it. This explanation may well tell the whole story without the need for further elaboration, but the possibility that Horace himself might have suspected textual corruption is worth exploring. We have seen that two probable corruptions – both easily detectable and simply fixed – have crept into the text, one of which at least is ancient. The question asked above might now be reformulated to consider why Horace would choose to allude to a transparently corrupt text. I am tempted to propose that in making the allusion to a flawed or inferior reading, Horace is commenting on the poor quality of the physical text itself and, by implication, on the common circulation of that text, with its lewd content, to an uncritical and coarse public. In the form in which Automedon's closely contemporary epigram was circulating in Horace's time, the physical artefact presented to readers was anything but a lepidum novum libellum (Catull. 1.1) – a curated, polished edition; rather, the epigram seems to have been preserved as a carelessly copied product that might be compared, in modern terms, to a badly edited, throwaway paperback published for consumption by an idle, undiscriminating audience seeking raunchy entertainment.
Horace's implicit condemnation of the epigram's content is signalled both by the parallels that link the ode closely to its epigrammatic source and by the reversals that distinguish it from that source. Both Automedon and Horace refer either to the dancer's, or to the dance's, Eastern origin (τὴν ἀπὸ τῆς Ἀσίης ὀρχηστρίδα, 1; motus … Ionicos, 21); to the arts the dancer employs, or which shape her (κακοτέχνοις σχήμασιν, 1–2; fingitur artibus, 22); and to the age of her lover(s) (γηραλέας ῥυτίδας, 6; iuniores adulteros, 25); in all these cases the parallelism also serves to some degree as a contrast. But while Automedon makes abundant use of words relating to dancing in the first part of the epigram and, in its second part, of words relating to sex and parts of the body, Horace refers to dancing only with the words motus … Ionicos; to the act of sex only with gaudia donare; and to the body only indirectly, with the idiom of the fingernail. These parallels and contrasts were noted by Colaclidès and McDonald, who also observed that, ‘as if to counterbalance his reserve in the use of such terms’, Horace ‘has put Automedon at a distance in another sense, in relying largely on an intellectual, moral, and social vocabulary’, with reference also to cognitive processes and choice (doceri, 21; meditatur, 24; eligit, 26; conscio, 29); while Automedon refers only to the dancer's knowledge (οἶδɛ, 6) of how to rouse the old man.Footnote 37
3. Conclusion
My analysis has attempted to illuminate two aspects of Horace's lyric discourse in the Roman Odes through a consideration of two parallel, yet contrasting, figures, a puer and a virgo. The aim has been, first, to unpack the ways in which the stories of each figure's progression from youth to maturity are presented in a compressed – some might say almost tortured – lyric discourse that eschews narration; the second aim has been to expose their poetic genealogies. In the second Roman Ode, the exhortations to inure a puer to military discipline break down unexpectedly with the injunction suspiret in the culminating image of a landscape peopled by characters from Vergil's Aeneid. In miniaturising Vergil's epic narrative, Horace also adapted or encoded stylistic techniques that are well documented in Vergil, including characterisation through allusion to Homer. The subtlety of Horace's own Iliadic allusions, in which he displays his nuanced understanding both of the minute details of imagery in the Iliad and of the ways in which any given simile or image fits within the broad strokes of Homer's Aeneas plot(s), shows Horace to have been as careful a reader of Homer as was Vergil.
In the sixth Roman Ode, Horace constructs a second unnarrated story that plays off his first. The adulta virgo of Carm. 3.2, a character inspired by Vergil's Lavinia, gave rise to the contrasting figure of the matura virgo in 3.6, but here Horace's allusion is not to Vergil's Lavinia but to his own lyric virgo. The girls’ situations in life are starkly opposed: the adulta virgo is under the protection of a royal mother and father (matrona bellantis tyranni, 3.2.8) and anticipates a suitable marriage (sponsus … regius, 10); while the matura virgo is under the tutelage, not of parents, but of unspecified people who school her in the arts of sexual enticement while her character is still being formed (doceri … gaudet, 3.6.21; fingitur, 22). Her intentions are directed not towards honourable marriage but towards plotting adulterous love affairs (incestos amores … meditatur, 23–4). Horace's matura virgo is the antitype of his adulta virgo, but her passage through the various stages of life, which are piled on top of and interlaced with each other at the very level of syntax, is analogous to that of the puer. Inserted into Horace's sketch of the virgo's downfall are a number of details derived from a frivolously obscene epigram by Automedon, including – puzzlingly – Horace's duplication of a probable textual error. Horace's reproduction of this senseless but easily corrected corruption, which may very well be deliberate, arguably has a metatextual function. The corruption reveals the low quality of an uncurated and (by implication) popularly circulating source text by an undistinguished author, whose sloppy production implicitly reflects its low moral content. This textual carelessness and the epigram's disreputable morality stand in sharp contrast to Vergil's polished and curated text of the Aeneid, with its serious ethical concerns.
The metatextual relationship between the corrupted text and corrupted morals touches on a Callimachean topos given rather delicate expression in Epigr. 30: the implied equivalence between a commonly circulating poem and a sexually desirable but promiscuous boy.Footnote 38 An amusing variation on this topos would be developed more extensively by Horace in the closing epistle of his collection of hexameter letters (Epist. 1.20), in which he anxiously addresses his personified book as a slave who is (overly) eager to be put up for sale and go out in the world, but whose release to the public means that he will be well-thumbed by vulgar hands and opens him to the danger of being debased and soon forgotten once his novelty and youthful sexual allure have worn off. Within the context of the Roman Odes, the underlying contrast between vulgar circulation and avoidance of the common crowd picks up on the programmatic declaration – itself a probable echo of Callim. Epigr. 30 – that opens the cycle at the very beginning of the third book: odi profanum vulgus (‘I shun the profane crowd’, Carm. 3.1.1). Horace's insistence on banishment of the common crowd before he utters his carmina non prius audita (‘songs not heard before’, 3.1.2–3) and his anxiety concerning the epistles’ cheapening, corruption, and even, perhaps, the jadedness or eventual oblivion that may come with publication are two sides of the same coin. Stewardship and protection of the text and a resolve to share it only when it is ready to be presented to the world are evident also in Vergil's refusal to accede prematurely to Augustus’ demand to see part of the Aeneid and in Vergil's deathbed wish that any writings he would not have published himself be burnt.
These motives may also account for Horace's underplaying – one might almost say his disguising – of his allusions to the Aeneid. Would Horace's readers in 23 BC have recognised his inexplicit gestures towards the Aeneid? I am tempted to propose that at the time of the Roman Odes’ publication Horace had in mind two contemporary classes of readers. One was the larger public, who might or might not grasp that the epic colouring of the second Roman Ode pointed to the much-anticipated, but as yet unpublished and still unknown, Aeneid.Footnote 39 The other was a carefully selected and exclusive audience, including Vergil (and quite possibly Propertius), consisting of an inner circle of those in the know. To these contemporary audiences we must, however, add Horace's proudly anticipated and ever-growing audience of future readers. These readers will have read the Aeneid and, in Horace's own words, would be reading his Odes ‘as long as the Pontifex climbs the Capitol with the silent [Vestal] Virgin’ (dum Capitolium scandet cum tacita virgine pontifex, Carm. 3.30.8–9).
An anecdote preserved in Donatus’ abridgement of Suetonius’ Life of Vergil (23–5) reports that Vergil first composed a draft of the Aeneid in prose and then proceeded to turn it into verse in no particular order, propping up the unfinished sections with very slight wording, ‘like scaffolding, Vergil would jokingly say, to hold up the structure until the solid columns arrived’ (quos per iocum pro tibicinibus interponi aiebat ad sustinendum opus, donec solidae columnae advenirent). This anecdote is consistent with the conclusions that can be drawn from an analysis of the ways in which Horace constructs his tribute to the unfinished Aeneid, an engagement that I have described as metatextual, intertextual, and extratextual. The anecdote would account both for the slight traces of Vergil's wording in Horace's text but also for why Horace does not show – or, perhaps, was not yet permitted by Vergil to reveal – a more detailed familiarity with the text of the ‘Iliadic’ Aeneid as we know it, even while he demonstrates his awareness of its larger structure. It can plausibly be argued that Horace alludes directly to the ‘Iliadic’ half of the Aeneid with two bookending intertextual references to Aen. 7.53 and 12.946–7. But he shows his deeper acquaintance with Vergil's unique narrative style, techniques of characterisation, and problematisation of interpretation and interpretability – features that can be observed already in the first book of the Aeneid – not through further direct imitation or adaptation of the text of the later books of the Aeneid, but through reference to narrative contexts in Homer that had, or would come to have, significance for the Aeneid. I have proposed the term ‘extratextual’ for this type of allusion, through Homer, to Vergil's intended design for his ‘Iliadic’ Aeneid.
Acknowledgements
I am grateful to my anonymous readers for their incisive criticism and to Bob Cowan for his further valuable observations, all of which served to strengthen my argument. Its weaknesses are my responsibility.
Open access
