Ennius' Scipio is represented for us by three fragments explicitly attributed to the poem by our ancient sources,Footnote ¹ frr. 31, 32 and 33 Courtney (= var. 9–12, 13 and 14 Vahlen), along with a detail from the Suda s.v. Ἔννιος (Ε 1348, p. 2.285 Adler = fr. 29 Courtney = var. I Vahlen) asserting Homer's pre-eminence as a panegyrist (and Scipio's as a recipient of panegyric), which is echoed (and perhaps amplified) by Valerius Maximus (8.14.1).

Editors have supplemented this small rump of undisputed fragments and testimonia by a variety of criteria, and the result is a Scipio comprising anything between three fragments of six lines and seventeen fragments of about twenty-six, the diversity of opinion graphically represented by Russo in his recent commentary.Footnote ² It is fair to say that there is little about this poem (or poems) that is free of controversy: representatively, Russo, after careful weighing of the complex and intensely debated evidence, is obliged to admit that most of the questions we would like to be able to answer about the Scipio, whether its date of composition,Footnote ³ its metreFootnote ⁴ or its genre,Footnote ⁵ remain stubbornly moot.

If the evidence available to us, particularly fragments open to a range of metrical reconstructions, leads us to an impasse, it might be worth trying a different kind of approach to understanding the Scipio. This article will thus start from two interrelated assumptions, neither of which commands any kind of consensus from scholars or unequivocal support from the evidence, but which I would still like to insist are the most likely to be true. I shall then show how, if these assumptions were indeed correct, Ennius' Scipio could make sense as a metrical artefact, as a composition alert to the ‘rules’ of genre, and even as the product of a particular juncture in Scipio's career. These guiding assumptions are that the Scipio was one poem, not a miscellany; and that its metre, from start to finish, was the catalectic trochaic tetrameter, better known in Rome as the trochaic septenarius or the uersus quadratus.Footnote ⁶ Once these assumptions are made (and I am far from the first to make them),Footnote ⁷ the fundamental peculiarity of the Scipio snaps into focus, namely the arresting contradiction between the contribution of the Suda (that only Homer could praise Scipio adequately) and the trochaic form of Ennius' poem of praise, since Homer and the uersus quadratus have very little in common. But as I have commented elsewhere,Footnote ⁸ ‘it is when a metre is used anomalously … that it becomes most interesting’. We may not be able to say very much at all about the Scipio, but it may at least reveal itself to embody a bold (and, I suspect, quietly influential) exploitation of metrical form.

The assertion of the Suda that Ἔννιος, ῾Ρωμαίων ποιητής … Σκιπίωνα … ᾄδων καὶ ἐπὶ μέγα τὸν ἄνδρα ἐξᾶραι βουλόμενός φησι μόνον ἂν Ὅμηρον ἐπαξίους ἐπαίνους εἰπεῖν Σκιπίωνος is supplemented by Valerius Maximus, who assesses the elder Africanus, subject of Ennius' poem, as uir Homerico quam rudi atque impolito praeconio dignior: it is clear that by the rude atque impolitum praeconium is meant Ennius' Scipio.Footnote ⁹ The claim that only Homer was equal to the task of praising Scipio is thus certainly to be traced to Ennius, but it is entirely likely that Valerius' dismissive evaluation of Ennius' own efforts at panegyric also derives from Ennius' poem, since it sits comfortably with the modesty of the implicit claim of inadequacy conveyed by the Roman poet's assertion of Homer's incomparable ability.

The beauty of assuming that Ennius composed the Scipio in trochaic septenarii, furthermore, is that the uersus quadratus would give Ennius' stance of poetic inadequacy metrical shape, since what makes trochaic septenarii so unlikely for a poem of epic aspirations (however qualified) – a ‘riddle’ is Friedrich Leo's apt descriptionFootnote ¹⁰ – is that a poem that establishes as an ideal the praise forthcoming from Homer is adopting a metrical form as far removed from the Homeric paradigm as it is possible to imagine. Res gestae regumque ducumque et tristia bella | quo scribi possent numero monstrauit Homerus (Ars P. 73–4), Horace famously insists: the dactylic hexameter, in other words, which under its most familiar ancient appellations (τὸ ἡρωικόν or herous) actively advertised its proper application, the Great Man. Even if the citation of Homer on its own did not raise the issue of the proper metrical form for praise poetry (although the Suda interprets Ennius' remark as an evaluation of τοῦ ποιητοῦ τὴν μεγαλόνοιαν καὶ τῶν μέτρων τὸ μεγαλεῖον καὶ ἀξιάγαστον),Footnote ¹¹ Ennius' adoption of septenarii surely would have done. Far from heroic, Ennius' chosen vehicle is one characterized by Fraenkel as a ‘simple, almost rustic’Footnote ¹² form long used in Rome ‘for scoptic poetry and verse accompanying play, in short for any kind of easily composed and barely poetic impromptu expression’.Footnote ¹³

The possibility that Ennius might have employed trochaic septenarii in a poem of praise to Scipio Africanus the Elder has contributed significantly to doubts about the poem's genre and unity. Even among scholars who have shared with me the assumption that Scipio was entirely composed in septenarii, the choice of metre has seemed prima facie peculiar and in need of explanation; and attempts to explain it have followed. But where all these efforts to elucidate Ennius' choice of metre seem to me to fall down is in their impulse to soften and rationalize the oddness of Ennius' choice. What if, on the contrary, the trochaic septenarii of the Scipio were meant to be anomalous and eye-catching, a bold and meaningful ‘abuse’ of metrical propriety that forces the reader to adopt an unanticipated view of the subject – further evidence of Ennius' readiness to court controversy in his metrical practice (and, I would be inclined to add, of the metrical self-consciousness that Roman poetry seemed to carry with it from its earliest days)?Footnote ¹⁴ Leo's approach was to posit lost Hellenistic models, whilst Richter explains the metre as ‘metrically a kind of intermediate stage between the saturnian that Ennius had rejected and the hexameter he had evidently not yet mastered’: ‘a dialogue metre from drama’ which could make shift in the interim.Footnote ¹⁵ The septenarius was indeed an important metre in Republican drama,Footnote ¹⁶ although that role seems not to have shaped its associations nearly as much as its other applications. But the weakness of Richter's explanation is that it leaves Ennius needlessly exposing his formal shortcomings, advertising his inability, as yet, to write proper panegyric in Homeric hexameters without any obvious pay-off for the poem he is writing.

Scholz recognizes the weaknesses of these arguments, and rightly concludes that the choice of trochaics for the Scipio should be assessed as a deliberate artistic decision on Ennius' part rather than some kind of makeshift solution,Footnote ¹⁷ and if I find his own explanation in turn ultimately less than satisfactory, it will doubtless seem a matter of emphasis rather than outright disagreement. Scholz draws on a scholarly tradition stretching back to RibbeckFootnote ¹⁸ which has identified the rationale of the trochaics of the Scipio, a nationalistically tinged praise of ‘the great Roman commander and victor over Hannibal’, in ‘the Roman tradition of the triumphal song’, for which the uersus quadratus was undoubtedly the established vehicle, although Ennius' septenarii in the Scipio seem to conform to a more refined register of versification than such extemporized examples, or indeed than was the norm in Roman drama.Footnote ¹⁹ A small problem here, acknowledged by Scholz, is the failure of this innovation in Roman panegyrical form (on the part of a poet whose influence on Roman metrical practice was generally immense) to establish a precedent for later literary practice,Footnote ²⁰ but my larger concern (as anticipated) is that such a reading of the metre of the Scipio as a (comparably elevated) alternative to the Homeric hexameter neutralizes the transgressive force of Ennius' metre – robs the metre of its power to compel on its own account.

That notion I shall develop later, but a more basic objection would be Scholz's (and his predecessors') isolation of a distinct genre of trochaic ‘triumphal songs’, when it seems easier to regard the chants of common soldiers at triumphs as a subset (and not a very clearly defined one) of a broader tradition of popular song. Long before the advent of Roman formal drama, the trochaic septenarius had found a use, as Scholz says, in ‘volkstümliche Spiel- und Spottverse, aber auch die Tradition der Triumphallieder der Soldaten’,Footnote ²¹ but the distinction implied here between light popular verse and the songs of soldiers at triumphs lacks cogency in the face of the surviving examples.

The soldiers' chants are represented for us by examples such as the refrains from Caesar's Gallic triumph in 46 b.c. (Gallias Caesar subegit, Nicomedes Caesarem. | ecce Caesar nunc triumphat, qui subegit Gallias: | Nicomedes non triumphat, qui subegit Caesarem, and urbani, seruate uxores; moechum caluom adducimus. | aurum in Gallia effutuisti, hic sumpsisti mutuum, Suet. Iul. 49.4, 51) and the triumphs of the alleged fratricides M. Aemilius Lepidus the triumvir and L. Munatius Plancus in December 43 (de germanis non de Gallis duo triumphant consules, Vell. Pat. 2.67.4), but in style and sentiment these military quadrati are very much in line with the civilian Spottvers coined in response to the death of the orator L. Licinius Crassus in 91 b.c. and referring to the enmity between Crassus and the corpulent C. Papirius Carbo ‘Arvina’ (aruina is ‘lard’), postquam Crassus carbo factus, Carbo crassus factus est, or the attack on Pompey preserved in the same context (an account of the trope of astismos, ‘urbanity’) of Sacerdos (Gramm. Lat. 6.461.26–462.3 Keil),Footnote ²² or for that matter the popular (i.e. non-triumphal) responses to Caesarian policy (Suet. Iul. 80.2, 3): Gallos Caesar in triumphum ducit, idem in curiam; | Galli bracas deposuerunt, latum clauum sumpserunt and (a graffito) Brutus, quia reges eiecit, consul primus factus est. | hic, quia consules eiecit, rex postremo factus est. At Dio Cass. 43.20 we are told that Caesar's soldiers made jokes similar to the first of these from the general populace (about new entries to the senate), and the close kinship of triumphal song and popular song in septenarii is also illustrated by yet another chant by Caesar's soldiery recorded by Dio, ἂν μὲν καλῶς ποιήσῃς, κολασθήσῃ, ἂν δὲ κακῶς, βασιλεύσεις, a parody of the trochaic nursery rhyme rex erit qui recte faciet, qui non faciet non erit (Porphyrio on Hor. Epist. 1.1.62).Footnote ²³ Horace, as we shall see, implies in a similar way that even during a triumph a distinction between soldiers' songs and those of the wider public might be hard to draw.

Beard insists, in reaction to the long-established interpretation of ribald triumphal song as apotropaic, that in fact soldiers' songs (according to the sources) could be eulogistic as well as insubordinate.Footnote ²⁴ An example of a eulogistic uersus quadratus from the popular domain is the response of the Roman ciuitas to (false) news of Germanicus' recovery from illness (Suet. Calig. 6.1), salua Roma, salua patria, saluus est Germanicus, loud enough to wake up Tiberius, we are told. But such positive chants differ only in tone from the ribald kind, not in register. Safer, in general, than an arbitrary distinction between extemporized song voiced by soldiers and by the wider public is a recognition that what unites the rather meagre remains of non-dramatic septenariiFootnote ²⁵ is rather more than what distinguishes them, and that is their consistently demotic character: popular effusions, as opposed to those emanating from the elite, whether from the processing soldiery or any other kind of popular collective, found their appropriate medium in the ‘simple, clear structure’ of the uersus quadratus.Footnote ²⁶

Consideration of a later exercise in praise of a great man, Horace, Odes 4.2, will support this subtle, but significant, recalibration of our reading of Ennius' trochaics – as a popular rather than specifically military form, and as a low-status form that Ennius wanted to be apprehended as such, not somehow to pass muster as a heroic medium. The latter clarification matters, I should say, because the contrast in register between Homeric and popular poetic forms will have been a dynamic one, a dramatic contradiction of readerly expectations surely central to the effect of the Scipio; and further evidence of Ennius' status as a revolutionary metrical innovator.Footnote ²⁷

Certainly the equivalent movement in Odes 4.2 is critical to the meaning of that poem, whether we are simply dealing with a close parallel to Ennius' procedure in the Scipio, or rather, in a book acknowledged to owe much to Ennius,Footnote ²⁸ a rather more direct relationship. Horace's poem is a subtly oblique panegyric of Augustus which is also highly self-conscious about its metrical form, setting the restrictive structures of its sapphic stanzas against the exuberant Pindaric forms to which Horace claims to be unequal.Footnote ²⁹ To this extent already 4.2 resembles the Scipio in its identification of a figure better able to provide the laudandus with appropriate praise, in Horace's case Iullus Antonius – though of course that conceit does not lack parallels elsewhere.Footnote ³⁰ But 4.2 takes a further step towards self-effacement, and achieves it as much as anywhere on the metrical plane. After describing Pindar's superlative achievements at 1–24, and his own more limited ambitions at 25–32, Horace sketches the panegyric that Iullus, maiore poeta plectro, will deliver (33–44). Then, however, the poem reverts to Horace, and Fraenkel draws out the implications of an exquisite metrical play first noted by Heinze:Footnote ³¹

Describing the shift back to Horace's perspective as ‘extraordinary’, Fraenkel sets out the later development of the poem: Iullus Antonius had suggested a Pindarizing treatment of the homecoming of the victorious ruler, but ‘Horace declines and steps aside, only to reappear after a moment, not in the role of a Pindarizing bard or any bard at all, but indistinguishable from any ordinary citizen’.Footnote ³² Plural verbs take over from the first person singular as the poet imagines ‘the whole ciuitas’ in celebration, himself lost in the throng: ‘Thousands and thousands of excited people will be lining the road on both sides and thronging in the wake of the procession, and one amidst the shouting crowd, no more conspicuous than all the others, will be Horace.’ And Horace gives his anonymity metrical expression, since o sol | pulcher, o laudande is a fragment, brilliantly intruded into the sapphic stanza, of a trochaic septenarius, in style as well as metre typical of public expressions in the uersus quadratus at such occasions: ‘No Pindaric ode now’, continues Fraenkel, ‘no elaborate poetry at all; the medium through which he voices his enthusiasm is the homely versus quadratus’; in metrical terms, not least, Horace is constructing himself as ‘quiuis unus ex populo’. We may further note that the Roman ciuitas chants a refrain, io triumphe, typically associated with the processing soldiery,Footnote ³³ and at a military celebration: Horace at least recognizes no distinction between the soldier's and citizen's voice.

Horace's approach to panegyrizing Augustus in 4.2 is thus to adopt the perspective, voice and metre of the ciuis Romanus. ‘Just as devotion to Augustus was not the privilege of any individual, but was felt by thousands of ordinary citizens, so the language in which that devotion was voiced should be one that seemed to come from the heart of the common man. “O sol pulcher, o laudande”, that is how the poet echoes the grateful rejoicing of the mass of the people.’Footnote ³⁴ What we discover, or rediscover, about this trochaic length is that it shares the mysterious power of other metrical forms to embody a particular personality. If, as Varro succinctly put it, ᾿Αχιλλέως ἡρωικός, ἰωνικὸς κιναίδου (Men. Sat. 360 Cèbe), the heroic hexameter is the metre of Achilles and the sotadean the metre of the cinaedus (cf. Ov. Rem. am. 381–2; Demetr. Eloc. 189),Footnote ³⁵ the uersus quadratus evoked equally vividly the common man. This is surely also what Ennius was aiming to achieve with the trochaic form of the Scipio, conveying the greatness of his subject and the universal gratitude he could command (as well as his own inadequacy to a properly Homeric task) by adopting the metrical standpoint of the ordinary citizen – as radical and arresting a strategy as his self-construction as alter Homerus in the Annals, though metrically and otherwise very different.

Scipio fr. 30 Courtney (var. 1 Vahlen) is about Scipio, our source informs us, but we cannot be certain that it originated in the Scipio. The passage appears to contemplate the physical memorials, a statue or a column, with which the populus Romanus will express their respect for the achievements of the great general, and it has been related, very plausibly, to the (textually controversial) section of Horace, Odes 4.8 which compares the poetic celebration of great achievements to the notae publicae engraved on marble, exemplifying the power of poetic commemoration with the Calabrae Pierides of Ennius, seemingly with specific reference to his praise of Scipio in the Scipio (13–20).Footnote ³⁶ If Odes 4.8 strongly supports the attribution of fr. 30 to the Scipio, that fragment in turn parallels rather nicely my characterization of the perspective adopted by Ennius in this poem, an expression of gratitude and respect to a great man emanating from the general populace. Add the parallel in Odes 4.2, where Horace, as one of the crowd, will deliver his trochaic praise when Augustus returns in triumph, and one might be forgiven for suggesting that a panegyric of Scipio couched in so overtly popular a metre looks very much like something written, as Ribbeck long ago suspected,Footnote ³⁷ in the immediate context of Scipio's return to Rome after Zama in 201 b.c., or at least is a work that harks back very vividly to those circumstances. The sources agree on the effusion of popular enthusiasm that greeted his return (Polyb. 16.23; Livy 30.45).

In conclusion, if indeed the metrical form of the Scipio embodied the modest status (the voice of the people) assumed by Ennius in the poem, it may be worth returning to Valerius Maximus' formulation, uir Homerico quam rudi atque impolito praeconio dignior. ‘Rough and unpolished’ is an extremely apt summary of the character of the uersus quadratus, subliterary, improvised and popular. But praeconium is a suggestive term, too. Praeco might denote an official herald or an auctioneer, and praeconium might also of course just convey an unspecific sense of proclamation. Thus at Cic. Fam. 7.24.1 C. Licinius Calvus' choliambic attack on Tigellius, Sardi Tigelli putidum caput uenit, is an ‘auction cry’, Calui Licini Hipponacteo praeconio, whilst at Arch. 24 Alexander at Achilles' tomb congratulates Achilles for having Homer as his praeco: ‘o fortunate’ inquit ‘adulescens, qui tuae uirtutis Homerum praeconem inueneris!’ That is almost certainly where we should leave it in Valerius Maximus, praeconio coordinated with both Homerico and rudi atque impolito, and thus communicating an unmarked sense of ‘the action of blazoning abroad, publicity (usu. w. the sense of commendation)’.Footnote ³⁸ But it is still tempting to wonder whether Valerius has in fact compressed his Ennian source, and praeconium, in the sense of the words of the praeco publicus, was originally an exclusive description of Ennius' un-Homeric undertaking. It wouldn't have been a bad one. Praeconium could simultaneously convey a humility of register (the praeco was not a distinguished position, at least from an elite perspective)Footnote ³⁹ and a capacity symbolically to represent the populus Romanus. We must assume that the praeco was a ubiquitous sight in the day-to-day functioning of the Roman res publica, proclaiming bills, summoning assemblies, advertising upcoming festivals, announcing election results (and much more), potentially a physical embodiment of the will of the Roman people:Footnote ⁴⁰ Cicero perhaps implies how far the praeco publicus could pass for the voice of the people when he claims, me non extrema diribitio suffragiorum, sed primi illi uestri concursus, neque singulae uoces praeconum, sed una uox uniuersi populi Romani consulem declarauit (Leg. agr. 2.2.4). Certainly, if one were minded to represent the proclamation of a praeco in verse, the uersus quadratus would be an appropriate choice of metre.