Sideshadowing Modern and Ancient

Counterfactual, or alternative, history has become very fashionable in the last thirty or so years, in historical studies at least; in certain areas of literature, especially science fiction, it has always been prominent. In addition to fully-developed depictions of what might have happened if a particular event had turned out differently (such as if the Axis Powers had won the Second World War in Philip K. Dick’s The Man in the High Castle or Robert Harris’ Fatherland), many works in most genres include more incidental speculations on how their own plot might have moved along a different path from that which it actually proceeds to take.Footnote ⁸ Already in the Iliad, the listener is told that Patroclus would have captured Troy but for Apollo’s direct intervention in its defence.Footnote ⁹ The effect on the reader (or listener) of such near-misses, what Nesselrath calls Beinahe-Episoden, is complex. In the majority of ancient examples, the signalling of a path not taken almost paradoxically reinforces the inevitability of the actualised course of events. As I have written elsewhere,

Virgil may tell his reader that, if Turnus had opened the gates of the Trojan camp to his comrades, that ‘would have been the last day for the war and for the [Trojan] people’ (ultimus ille dies bello gentique fuisset), but her awareness of Turnus’ solipsistic lust for glory assures her that this scenario must always remain in the unfulfilled subjunctive mood.Footnote ¹¹ More importantly, the sense that other paths of fate were possible reinforces her feeling that Jupiter’s plan is unfolding, not pis aller, and hence with the suspicion that it would have been trumped by a genuinely viable alternative, but as the teleologically dominant march of history, capable of seeing off all-comers. However, whereas these epic counterfactuals are emphatically unrealised even within their own narratives, Horace in Epistles 1.18 leaves the picture open. He does not write that Antony would have won if something else had or had not happened (something which, of course, respectively did not or did happen). Rather he leaves the question entirely in the balance as to ‘which of the two’ will win, and though coronet is indeed subjunctive, it is not the pluperfect of a safely unfulfilled condition, but the dangerously potent present of a final clause in primary sequence.

Even without this radical difference in the case of Horace’s ‘what if?’, the emphasis on the counterfactuality of counterfactual scenarios cannot totally destroy their power to suggest that alternative outcomes are possible, that the path of history is not entirely linear and teleologically predetermined. A particularly subtle articulation of this is to be found in the work of Gary Morson and Michael André Bernstein on 19^th and 20^th century novels, particularly Tolstoy, Dostoevsky, and Holocaust literature.Footnote ¹² They coined, by analogy – and contrast – with foreshadowing, the term ‘sideshadowing’, whose connotations are most clearly summarised by Bernstein:

The relevance of sideshadowing to speculation that Antony might have won Actium is surely clear, since it is just such a ‘triumphalist, unidirectional view of history’ which Augustan ideology adopted, and to be ‘found wanting by some irresistible historico-logical dynamic’ is precisely the manner in which Antony and Cleopatra were depicted, even – perhaps especially – when most sympathetically painted as doomed lovers.Footnote ¹⁴

No less suggestive for our purposes is Morson’s simile for the relationship between sideshadows and the actualised narrative:

This is, of course, a simile, but, as with all similes, the vehicle has the potential to bleed into the realm of the tenor. Kings and other monarchs, even when not ruling by ‘divine right’, rely in part on narratives which reinforce their legitimacy by constructing a fiction of inevitability around the events which brought them to power. Not only the narrative of Octavian’s victory at Actium but his position as princeps is challenged by the sideshadowing ‘pretenders’ of an Antonian victory and an Antonian principate.Footnote ¹⁶ It is the official inevitability of Octavian’s victory at Actium and hence of Augustus’ principate which Lollius’ counterfactual speculation subverts.Footnote ¹⁷ In order to appreciate the full force of this subversion, it must be situated in its historical and historiographical context.

Counterfactual speculations are not alien to classical historiography. The most famous example is probably Livy’s extended discussion of what would have happened if Alexander had fought the Romans.Footnote ¹⁸ However, our focus here must be, not on the ramifications of a large-scale difference of strategy, such as if Alexander had gone West instead of East, or for that matter if Hitler had concentrated his forces on the Middle East rather than the Soviet Union.Footnote ¹⁹ Rather it is those critical turning-points of history, battles, which are more pertinent to Lollius. The notion that battles could go either way, with the implicit or explicit suggestion that this would have major ramifications for subsequent events, is hardly unusual. The idea probably has a fairly strong basis in reality, since battles can turn on relatively small factors (‘for the want of a nail…’), and can in turn have a disproportionate impact on the outcomes of wars, and hence of regional or even global power. However, for our purposes, the important thing is the strong hold which the contingency of battles has and always has had in the popular and historiographical imagination. Indeed, battles are a particularly common focus for counterfactual speculation in collections of ‘alternative histories’ (‘What if the Japanese had won at Midway? The Confederates at Gettysburg? Xerxes at Salamis?’)Footnote ²⁰

In classical historiography, the topos is particularly at home in pre-battle hortationes urging troops and embassies begging allies to avert the world-changing consequences of defeat. Yet it also features in authorial commentary before and after the description of such turning-point battles. The former, of course, are aimed more at rhetorical persuasiveness than at historical analysis, but that very persuasiveness depends on its audience’s conceptualising the alternative outcomes which turn on the battle. Thus Herodotus has Pausanias send a messenger begging the Athenians to help at Plataea, ‘since the greatest struggle lies before [them] as to whether Greece will be free or enslaved’.Footnote ²¹ Livy has both Scipio and Hannibal, in oratio obliqua, stress Zama’s significance as deciding not merely the Second Punic War but world domination, a notion which Polybius appends to their hortationes in the authorial voice.Footnote ²² Turning to the civil wars, and to more reflective authorial comment, both Appian and Cassius Dio figure Philippi as a great contest which determined whether Rome would thereafter be a republic or a monarchy, with Dio in particular magnifying its importance in contrast to later civil-war battles which merely decided between monarchs.Footnote ²³ Perhaps the most intriguing counterfactual speculation about the battle is that by Velleius Paterculus, who sets up two pairs of counterfactual alternatives within a third, as the two possible outcomes of Philippi (victory for either Liberators or Caesareans) themselves bifurcate into which half of each possible victorious duo would then go on to individual dominance, Brutus or Cassius, Antony or Octavian.Footnote ²⁴

The Inevitability of Actium

Depictions of Actium seem to have differed somewhat from this pattern. Such was the centrality of the battle in Augustan ideology and so extensive the manipulation of its image that it is extremely difficult to excavate either the precise nature of the battle or even immediately contemporary attitudes from underneath the layers of significance given to the ‘secular miracle’ which Actium became.Footnote ²⁵ Osgood has shown how contemporary evidence already demonstrates the magnification of the significance of Actium and the increase of its status as a turning point in the months following the battle.Footnote ²⁶ Alongside this was an increasing emphasis on the inevitability of Octavian’s victory. As Welch writes of the whole period from 35-30 bce, covering the lead-up to Actium, the battle itself and its sequel in Alexandria, ‘[l]ater descriptions of the period are invested with a certain inevitability, but the future cannot have seemed secure at the time.’Footnote ²⁷ Certainly, there is evidence that the battle’s outcome was far from a foregone conclusion and that this was widely recognised at the time and in the immediate aftermath. We need only think of Macrobius’ famous anecdote (2.4.29) about the man who had trained two crows, one to say Aue, Caesar uictor imperator and the other to say Aue, uictor imperator Antoni.

Ultimately, it is not essential for our current discussion of Lollius’ naumachia to establish what ‘really happened’ on 2^nd September 31 bce. Of course, this question is one of immense intrinsic historical interest and it even has relevance to the later depictions and constructions of Actium, inasmuch as they may have occluded, reacted against and inverted aspects of the battle as it was fought which were ideologically undesirable. However, any ‘official’ version or versions – or at least those dominant features of the incipient tradition which can plausibly be taken as congenial to Augustus and his allies – would need to react against and counter rival versions which might have no greater basis in reality than themselves, quite as much as ‘the truth’. Indeed, even if we stop short of a hyper-sceptical position in which there are no historical facts, only perceptions and interpretations of them, Actium is a prime candidate for the title of a historical event which cannot be separated from the deeply partisan perceptions and highly tendentious interpretations which must have been felt and produced even while the battle was taking place, let alone in the immediate and longer-term aftermath.

In evaluating how the counterfactual speculation implied by the word alterutrum in Epistles 1.18 would operate within and against Augustan ideology about Actium, our main interest must be in that ideology, almost independent of the ‘reality’ behind it. Indeed, since the principal aspect of that ideology will be the teleologically predetermined and divinely guaranteed inevitability of Octavian’s victory, it need hardly be established for most modern readers that, in ‘reality’, that victory was not so preordained as part of a manifest destiny. Although there is a range of contemporary sources, including numismatic, architectural and other material evidence, it is on the evidence of poetry that we shall focus for two reasons. In the first place, these poetic Actiums are the most directly comparable form of evidence for Lollius’ naumachia, which is itself doubly poetic, being both embedded within Horace’s poetic epistle and metapoetically representing Lollius’ (or Horace’s) own poetic mimesis of the battle.Footnote ²⁸ Secondly, the poetry of Virgil, Propertius, and Horace himself combines the virtue of being more-or-less contemporary with Epistle 1.18, which is lacked by the far later historiographical evidence of Plutarch, Appian, Dio, and others, with that of having language’s ability to construct the notion of inevitability, which is much harder (though not impossible) to convey in visual media.

The key aspects of the way in which Actium was depicted in poetry of the 20s and teens bce are threefold and all three may be seen to be subverted to a greater or lesser extent by Lollius’ naumachia. The first is the figuring of the battle as the climax of a bellum externum against Cleopatra’s Egypt. This, as we have already noted, following Ahl and others, is succinctly debunked by the civil-war associations of the phrase aduersarius est frater.Footnote ²⁹ The second is closely related to the first, but stands in some tension with the third. The battle was a climactic, almost cosmic struggle between Western civilization and Oriental barbarism, but at the same time a struggle whose outcome was never in doubt, indeed was so teleologically predetermined as to exclude even the contemplation of any alternative outcome. The scale and significance of the conflict could conceivably be belittled by Lollius’ small-scale re-enactment, as Gurval suggests.Footnote ³⁰ Against this interpretation must be set two considerations: first, that the naumachia is a mimesis rather than a replica, and hence a reduction in scale is appropriate, and, second, that Horace himself recommends restraint and discretion when re-enacting Actium. Indeed, a mini-Actium could even be considered in keeping with the implicit Callimacheanism which pervades every aspect of Horace’s poetics, ethics and politics from Satires 1 onwards.Footnote ³¹ If anything, the traumatic impulse to replay the battle is rather testimony to its magnitude and significance for Lollius. That the uncertain outcome of the virtual Actium subverted the inevitability of the future Augustus’ victory in the official Actium is the key point I wish to make. Much has been written on the four poetic Actiums by Horace, Virgil, and Propertius, both individually and in comparison with each other, so I shall only briefly (re-)establish the sense of an inevitable Augustan victory which they produce.Footnote ³²

Something of the shift between initial constructions of Actium and those which developed in the course of the 20s bce can be traced by looking at Horace’s own two depictions of the battle, in Epode 9, probably written and published in the year or two following the battle, and Ode 1.37, published in 23. The scholarly controversies over Epode 9 and the bibliography which they have generated are both too large and too tangential to the current discussion to justify more than a glance here. For our purposes, it is noticeable that many but not all of the features of what would become the dominant version of Actium are already present, though with different emphases and even some differences of detail.Footnote ³³ Cleopatra is emphasised, and with her all the associations of Eastern luxury, decadence and tyranny (11-16). Yet, for all that the comparisons with Marius’ triumph over Jugurtha or Scipio’s over Carthage (23-6; which Scipio need not detain us here) align Octavian with other victors over foreign, and specifically African, foes, the parallel with Naulochus and between Antony and Sextus Pompeius (7-10) draws attention to the battle’s civil nature, albeit while clearly constructing the Roman enemy as morally depraved.

However, the main divergence from later depictions is the extent of the fear and anxiety which the pro-Octavianic party (including Horace) are shown as feeling (37: curam metumque Caesaris rerum). This can in part be attributed to the dramatic setting of the poem, most probably spanning the course and the immediate aftermath of the battle itself, but that answer only shifts the question as to why Horace chooses such a dramatic moment. We shall see a similar focalization through anxious pre-Actium Romans in Odes 1.37, but their anxieties are then shown to have been groundless. In Epode 9, there is no such reassurance, and indeed the emphasis on wine as a means of dispelling anxiety and fear, as opposed to celebrating victory, adds to the sense of both their justification and even their continued validity. Epode 9 is by no means an innocent response to Actium, but rather one which reflects (and, like all Triumviral and Augustan poetry, helps to shape) the incipient ideological construction of the battle. At this stage, in the early 20s, the inevitability of victory does not seem to have been part of that construction.Footnote ³⁴

The depiction of Actium in Odes 1.37 focuses on Cleopatra to the exclusion of Antony and emphasises the ‘Oriental’ frenzy and drunkenness of her entourage, setting up a battle ‘between West and East, man and woman, virtus and impotentia, reason and passion, Republican liberty and monarchical enslavement.’Footnote ³⁵ Almost all studies of this ode are – quite rightly – principally concerned with the depiction of Cleopatra, in particular the shift from the crazed tyrant of the first four stanzas to the more complex, admirable and even sympathetic figure of the last three.Footnote ³⁶ This in itself is relevant to the magnitude with which the battle is invested, since one of the standard ways in which a panegyrical reading of the poem accommodates the sympathetic portrayal of Cleopatra is to see the greatness of the adversary redounding to the glory of the victor.Footnote ³⁷ The ode is also interesting in terms of the way in which it sets up the battle as the climax of a bellum externum between Rome and Egypt (as well as between various corresponding gender, ethnic, social and political polarities), but also subtly signals its status as a civil war battle, notably through the use of the motto from Alcaeus celebrating the death of Myrsilus, an opponent in domestic Mytilenean stasis.Footnote ³⁸

In terms of the inevitability of Octavian’s victory, this ode is unusual in that it does offer an explicit sideshadowing of what would have happened if Cleopatra had won. Indeed, the famous Alcaean motto nunc est bibendum is set in antithesis with the inadmissibility (nefas) of drinking ‘while the queen was devising crazed destruction for the Capitol and death to the empire’ (1.37.6-8: dum Capitolio | regina dementis ruinas | funus et imperio parabat). This (for a Roman) horrifying scenario bears a resemblance to the scare tactics employed in historiographical pre-battle hortationes and, according to Morson and Bernstein’s principles of sideshadowing, any such indication of an alternative outcome disrupts ‘the affirmations of a triumphalist, unidirectional view of history’.Footnote ³⁹ Yet Horace does as much as is poetically possible to sequester the Ptolemaic assault on the Capitol within the fevered delusions of Cleopatra herself, ‘incapable of controlling her hopes and drunk on the sweetness of fortune’ (10-12: impotens | sperare fortunaque dulci | ebria), and doomed to be driven back by Octavian into ‘fears which are grounded in reality’ (15: in ueros timores).Footnote ⁴⁰ A counterfactual Ptolemaic victory (for it can hardly be called Antonian in the discourse of this ode) can only be viewed as possible through Cleopatra, a focaliser who is explicitly framed as deluded and indeed self-deluding, or through the understandably but groundlessly anxious citizens of Rome. The objectively (within the ode’s own ideologically-determined construction of objectivity) ‘true’ or ‘justified by reality’ nature of her fears entails an inevitability about their fulfilment, the might of Caesar securing victory. The double simile comparing victor and vanquished to hawk and dove, hunter and hare reinforces the sense of a mismatch in the battle, that the outcome could never have truly been in doubt, since one side was so immensely stronger than the other.Footnote ⁴¹ If anything, Horace comes close to losing the sense of the battle’s momentousness in bringing out, through the twin focalisers of the increasingly despairing Cleopatra and the increasingly relieved citizens back in Rome, just how groundless their earlier respective hopes and fears had been. Nevertheless, in the mimetic journey through the emotions and realizations of the battle, he ultimately preserves precisely the sense that Actium was a turning-point of supreme significance, but also that it could only ever have turned in one direction.

Moving beyond the 20 bce publication of Epistles 1, but still well within the same historical moment, and, considering the Aeneid’s fabled long gestation and probable circulation within the circle of Maecenas, perhaps even pre-dating it in time of composition, we turn to Virgil.Footnote ⁴² A detailed analysis of the depiction of Actium on the shield of Aeneas in Aeneid 8 would be far beyond the scope of this article, and would inevitably duplicate much of the even greater quantity of scholarship which has accrued to this passage. Its orientalising of the forces of Antony and Cleopatra (though Antony is included), and its cosmic dimension as a clash of order and chaos are too familiar to be rehearsed here.Footnote ⁴³ Rather I shall briefly focus on the elements of the ecphrasis which contribute to a sense of the Octavianic victory’s teleological inevitability. The first is the very nature of the depiction as a divinely crafted item of plastic prophecy, a victory whose fulfilment was predetermined in the mind of Jupiter and the craftsmanship of Vulcan at least a millennium before it took place. Pagán, herself drawing on the work of Morson and Bernstein, eloquently expresses the effect of foreshadowing in the shield, and its concomitant exclusion of sideshadowing: ‘The universe is indeed closed in the compass of the shield. The choices have been made. There is no room for what might have been. The course of events is inevitable, and Aeneas has no choice but to accept it.’Footnote ⁴⁴

This providential divine ordinance of victory is reinforced by a second element, the role in the battle itself of the Olympian gods, whose alignment with the forces of Octavian not only defines that side as representative of order but guarantees its victory, ultimately an easy one, as the triumph of light over darkness conventionally is in Graeco-Roman culture.Footnote ⁴⁵ Finally, Quint has influentially shown how the formal elements of epic reinforce a sense of teleological inevitability for the triumph of the victors.Footnote ⁴⁶ Epic, at least the ‘epic of the winners’, does not merely depict that the victors were victorious, but excludes the very possibility that it could have been otherwise. All these elements, however much one might or might not wish to see them subverted and problematised by the Aeneid’s wider polyphony and ambivalence, nevertheless, on the surface at least, contribute to and reflect a construction of Actium as an inevitable and teleologically predetermined victory for Octavian.

Proceeding further into the teens bce, an even clearer sense of inevitability is presented by Propertius’ controversial depiction of the battle as an aetiology of the temple of Actian Apollo in 4.6. For our purposes, it is mercifully unnecessary to decide whether the tone of the elegy is panegyrical or parodic, politically committed or subversively reluctant, since all interpretations tend to agree that Octavian’s victory is depicted as easy and inevitable, however they interpret the further implications of that depiction. Whether the reason lies in the poem’s exploitation of Hellenistic and rhetorical narrative techniques, in elegiac poetics, in an intention to parody, or in a grudging reluctance to dwell on uncongenial subject-matter, Propertius devotes only one couplet to the fighting itself (57-8).Footnote ⁴⁷ As Hutchinson puts it, ‘the battle is dispatched with ‘mannerist’ rapidity. The ease of victory is also conveyed’.Footnote ⁴⁸ For our purposes, this is the key point. The parallelism of Apollo’s Olympian and Octavian’s human might is made even clearer here than in Virgil, as they are respectively described in the hexameter and pentameter of a single elegiac couplet (55-6). Apollo himself acknowledges the anxiety of Romans at the possibility of an Egyptian victory (41-4), but only to emphasise its utter groundlessness. Propertius effectively combines magnification of the victory with a sense of its inevitability by having Apollo stress the size and number of the enemy forces, and to place even greater weight on the idea that it is the justice of Octavian’s cause which makes his success inevitable (47-52). Whatever tone we, or indeed a necessarily diverse contemporary readership, might wish to detect in all this, it clearly reflects, even if it exaggerates, a discourse of inevitable victory.

It has emerged from this brief survey of Augustan poetic Actiums, not only that there was a schematization of the battle as one between West and East, order and chaos, and all the other polarities which map onto them, and that Octavian’s victory was depicted as teleologically inevitable, but that these two elements were inextricably linked. Especially in Virgil and Propertius, the inevitability of Octavian’s victory derives from the inevitability of victory for the forces of Olympian order, masculinity, the West, liberty, and, in the broadest terms, right. Inferentially, the syllogism can be read in the opposite direction and the victory has to be inevitable if it is to be considered a victory for these forces and values. If victory could conceivably have flitted lightly to the other side and Octavian’s victory was merely contingent, or the result of luck, then his claim to represent those forces and values is seriously undermined, if not invalidated. By rejecting the inevitability of Octavian’s victory, Horace and Lollius reject its status as the triumph of order over chaos. In many ways, this is not far different from the very basis of Morson and Bernstein’s notion of sideshadowing. Bernstein writes of how narrative, unchallenged by sideshadowing, implicitly constructs an ‘irresistible historico-logical dynamic’ by which the winners are meant to win and the losers to lose.Footnote ⁴⁹ The Augustan myth of the inevitable victory at Actium took that implicit, abstract, amoral dynamic and gave it ethical, ethnic, religious, cosmic and political dimensions which made explicit (or all but so) the notion, not just that the winners were meant to win and the losers to lose, but that it was right for them to do so. It is this complex of ideas, and not merely the inevitability of the victory itself, which Horace and Lollius’ sideshadowing challenges. It remains to explore, however, one aspect of the mechanics of how that challenge is made, and also to consider whether it may have been even more subversive.

Counterfactual naumachiae and Ships’ Rigging

The choice of a naumachia as the form of mimesis which Horace the poet and Lollius the battle re-enactor simultaneously perform is significant in itself. It might be objected that the naumachia was the principal form of battle (re-)enactment in Roman public ludi, and hence that it was the obvious choice for Lollius’ smaller-scale, private version. The naumachia certainly became the most common – or at least most frequently described – large-scale combat in the ludi of the Principate, but Horace was writing in the comparatively early years of such spectacles. It is worth remembering that, according to Velleius and Suetonius, in Caesar’s triumphal games of 46 bce, there was not only the naumachia between the Egyptians and Phoenicians but also the staging of a land-battle involving infantry, cavalry and even elephants, not to mention pitched camps, to make room for which the turning-posts had to be moved.Footnote ⁵⁰ Later Claudius re-enacted the storming of a British town (Suet. Claud. 21.6). It would therefore be by no means unthinkable for Horace to imagine Lollius staging a mocked-up land-battle in a field on his father’s estate.

Of course, only by a naumachia could Actium be re-enacted. Yet Actium itself is not the only or even perhaps the most obvious choice. Certainly, it was the foundational battle of the Augustan principate, but, apart from the more-or-less contemporary Epodes and the significant exception of Odes 1.37, Horace makes surprisingly few references to Actium, and it is Philippi which is the more prominent battle throughout his oeuvre.Footnote ⁵¹ It was Philippi at which Horace himself fought on ‘the wrong side’ and from which he fled, like Archilochus or Alcaeus, without his shield, Philippi which he repeatedly recalled, perhaps most memorably in the most pregnant adonius in all Horace’s Sapphics, consule Planco.Footnote ⁵² The earlier battle might also be thought more worthy of subversive counterfactual speculation, inasmuch as it was a more significant watershed between Republic and Principate, as opposed to one which merely decided which Caesarean would achieve individual supremacy. Indeed, we have already seen precisely this sort of sideshadowing (though its subversive quality is doubtful) in the later historiographical tradition about Philippi, as represented by Velleius, Appian, and Dio.Footnote ⁵³ While it would be going much too far to suggest that Actium was chosen purely because it could be re-enacted by a naumachia, there is a peculiarity about this particular form of spectacle which unquestionably contributes to the effectiveness of his exercise in counterfactual history, for the following reason.

Naumachiae were of two kinds and in neither was the outcome predictable.Footnote ⁵⁴ Some, like that mentioned above between ‘Egyptian’ and ‘Phoenician’ navies at Caesar’s triumphal games, seem to have been fantasy battles, not rooted in any specific historical context, between forces which the people’s imagination might wish to see matched for maximum spectacle, like modern children arguing in the playground about who would win if the Hulk fought Superman.Footnote ⁵⁵ In such spectacles, it is reasonable to assume that the outcome of the battle would be unpredictable, for the audience at least. The majority of attested naumachiae, however, were re-enactments of famous historical battles, such as Salamis or Syracuse, though – significantly for our passage – not battles involving Roman forces, let alone between Roman forces.

A striking feature of the ancient accounts of these spectacles is the way in which the result of the re-enactment did not always replicate that of the original battle. The clearest example of this is in Dio’s account of Titus’ re-enactment, at the opening of the Flavian Amphitheatre in 80 ce, of the Syracusan defeat of the Athenian navy during the second and final year of the Sicilian expedition, in 414 bce. Despite the famous (or notorious) historical victory of the Syracusans and their Peloponnesian allies, as immortalised by Thucydides, Titus’ naumachia and the land battle into which it segued depicted an Athenian victory.Footnote ⁵⁶ The other piece of evidence for such re-writing of history is more problematic. Dio’s reference to the re-enactment of Salamis during Augustus’ games to celebrate the dedication of the temple of Mars Ultor in 2 bce includes the remark ‘and the Athenians won this time too’ (55.10.7: καὶ ἐνίκων καὶ τότε οἱ Ἀθηναῖοι). Coleman tentatively suggested in 1990 that this comment ‘may imply that [Dio] thought the outcome was a coincidence’, though, three years later, she added the rider that ‘equally his remark could be a neutral comment merely recording the result.’Footnote ⁵⁷ Even if we rely only on the example of Titus, there remains a parallel for naumachiae re-writing history in the same way as Lollius and his brother are imagined as doing. As such, a naumachia is by its nature a perfect vehicle not only for the re-enactment of the sea-battle of Actium but for counterfactual speculation about its outcome.

The further question raised by the examples of Titus’ and perhaps Augustus’ naumachiae is whether or not their outcomes were pre-arranged. Against this is the presumable difficulty in both persuading and organising the gladiators, slaves or condemned criminals who would have taken part in the spectacle to let themselves be defeated and killed according to script, as opposed to enjoying the possibility of winning victory and survival by their own prowess. Less compelling is the argument that a pre-arranged outcome would detract from the audience’s satisfaction.Footnote ⁵⁸ It is quite conceivable that the result of the battle could be rigged without the audience’s being aware of it, and even if they did know who was going to win, the impressive spectacle may well have been sufficient entertainment without the added dimension of sporting unpredictability. In any case, it is striking that the one counterfactual outcome for which we have unambiguous testimony re-wrote history so that the privileged historical model of enlightened imperialism, Athens, should be victorious.Footnote ⁵⁹ Moreover, the details of Dio’s narrative suggest a carefully choreographed sequence of events climaxing in the set-piece storming of the Ortygian walls and it is hard to imagine, should the ‘Syracusans’ have fortuitously happened to win the naumachia, what use could have been made of the doubtless elaborate mock-up of those fortifications. It seems, therefore, on balance, more probable that the outcomes of naumachiae, both factual and counterfactual, were pre-arranged.

What implications does this have for Lollius? We have seen that it was profoundly subversive even to suggest the contingency of Octavian’s victory at Actium in the face of its official portrayal as a teleologically pre-determined inevitability. If Horace is suggesting that Lollius and his brother might deliberately choreograph the naumachia so that ‘Antony’s’ forces win, then the subversion would be somewhat greater.Footnote ⁶⁰ Of course, alterutrum does not in itself necessarily indicate an Antonian victory, but it could mean ‘whichever you have pre-arranged will win’ just as easily as ‘whichever happens to win’ and it would be hard to imagine Horace being quite so blatant as to be explicit about the outcome. The coded nature of the reference would be appropriate both to the historical reality of Horace’s publication of the Epistles in Augustan Rome under Maecenas’ patronage and to his mimesis and modelling of the sort of tact and discretion which Lollius should employ when dealing with dangerous subject-matter. Alternatively, Horace could be constructing himself as actually giving Lollius the choice as to how subversive he wants his virtual Actium to be, whether he chooses to make Octavian or Antony the victor.

Two Thumbs Up: The naumachia within Epistle 1.18

How does Lollius’ subversive sideshadowing fit within the wider context of Horace’s advice to his young friend?Footnote ⁶¹ In particular, how does the speculation about a possible Antonian victory fit with the final couplet, which closes not only the description of the naumachia but the whole section on balancing Lollius’ and his patron’s preferred pursuits?Footnote ⁶² If the counterfactual speculation is truly subversive, why would the patron ‘praise’ it? The lines were quoted with the rest of the passage at the beginning of this article, but it will be convenient to repeat them here (1.18.65-6):

The full import of these lines is, like so much else in this epistle, far from clear. In the broadest terms, it fits with the overarching issue of the balance between freedom and concession, boorishness and toadying, and the elusive Aristotelian mean between them, as set out explicitly in the poem’s opening section (1.18.1-20). It is worth stressing that the particular emphasis here is on speech, its freedoms and its constraints.Footnote ⁶³ As such, there seems to be a degree of give-and-take, albeit a highly asymmetrical one, as the patron offers approval for Lollius’ ludus in reciprocity for the latter’s (perceived) compliance with his own studia. However, it is not clear how this reciprocity works. Is it because Lollius complies with the patron’s desire to go hunting (a characteristic of the flatterer, as Hunter notes)Footnote ⁶⁴ rather than write poetry (39-52) and undertakes military exercises and service (54-7) that the patron tolerates the naumachia and similar nugatory games, even though he is himself indifferent or even unsympathetic to them? Or does the patron think of the naumachia as part of Lollius’ compliance with his studia, so that, as Macleod puts it, ‘even in his free time he will be wise to avoid verse and stick to something with a military and pro-Augustan flavour to it, like playing at the battle of Actium on his estate’?Footnote ⁶⁵ It is not even certain what is meant by the ludus: the poetry-writing which Lollius must subordinate, but not abandon, to his patron’s enthusiasms, or the naumachia, or both.Footnote ⁶⁶

Critics such as Bowditch and Porter, building on Ahl’s observation that Lollius and Horace allude to the taboo subject of Actium’s civil-war nature, have offered sophisticated accounts of the complex and ironic interplays at work in this passage, accounting for the presence of such a subversive subtext in an approved ludus. Porter couples the naumachia with the earlier exemplum of Amphion and Zethus (1.18.41-5) to argue that the two passages ‘suggest both the severe limitations and compromises that Lollius – or Horace – must accept as conditions of the client/patron relationship and also the freedom that remains possible through the ludus of representation.’Footnote ⁶⁷ This beautifully expresses the tension inherent in Lollius’ and Horace’s situation, but does not in itself account for how a ‘freedom’ which is tantamount to subversion and dissent can be accepted and even ‘praised’ by the patron. Bowditch emphasises the reciprocity operative in the scenario and the concession which Lollius’ partial acquiescence earns, as ‘the skit becomes not an activity of the patron’s to which Lollius defers, but rather an act of risky independence of a type which would win the patron’s approval in exchange for participation in his affairs’.Footnote ⁶⁸ Yet, unlike Porter, she goes further to address head-on the apparent contradiction of approved dissent. She shows how the paradox at the heart of the patron-client relationship and its tension between freedom and constraint makes the expression of liberty itself an act of submission, as ‘the independence of the act of representation itself depends on compliance with the patron … the compliance with the overall fiction or scheme of the Principate permits or allows a critical gesture that diverges from the party line.’Footnote ⁶⁹ Such an interpretation, formulated to account for the subversive assertion of Actium’s civil-war nature, works equally well for the further subversion of the counterfactual speculation about an Antonian victory, a trade-off but also perhaps an expression of power. The dominance of the ‘party line’ is demonstrated by its ability to accommodate (suitably circumscribed) dissent.

Bowditch’s elegant and subtle position is more than sufficient to situate Lollius’ sideshadowing within the wider argument of Epistle 18, but there is one further possibility. Bowditch herself gestures towards it when she argues that Horace uses Lollius and his virtual Actium to dramatise his own struggles with constraint and circumscribed freedom and that ‘this situates the liberty that Horace does take on a subtextual level.’Footnote ⁷⁰ I would be reluctant to separate Lollius and Horace too sharply, rather seeing them as parallel expressions of the client’s (and especially the poetic client’s) dilemma. However, Bowditch’s recurrent emphasis on subtext does fit with a reading of Epistle 18 in which both the civil-war nature of Actium and the contingency of its outcome are encoded so subtly as to elude the notice of both the internal audience that is Lollius’ patron and the external audience that is (a portion of) Horace’s readers. Such a reading would also account for Horace’s curious formulation not that approval will come if Lollius complies with his patron’s enthusiasms but if the patron believes (crediderit) that he does.

Of course, to write that someone ‘believes’ something does not necessarily suggest scepticism about the validity of that belief. Nevertheless, it is striking how often in Epistles 1 the verb credo is used in contexts of explicitly mistaken belief or at least a value judgment from which Horace dissents. Thus Maecenas (or the generalised ‘you’ into which the addressee has morphed) ‘sees with how much effort of head and mind he avoids what he believes to be the greatest evils, a small fortune and a disgraceful electoral defeat’ and what Horace’s uilicus ‘believes to be wilderness and inhospitable wasteland, the man who feels the same as [Horace] calls idyllic’.Footnote ⁷¹ There is no such explicitness in the present context, and it is certainly possible to take crediderit here as ‘rightly believes’, putting emphasis on the overwhelming importance of the powerful amicus’ perceptions.Footnote ⁷² However, it remains a jarring qualification and one which opens up a range of possibilities, especially following the careful framing of the naumachia as a potentially controversial activity. ‘Playing at the battle of Actium’ may well seem to have a ‘pro-Augustan flavour’, leading the patron to believe that Lollius not only complies with his pursuits (OLD s.v. studium 4) but agrees with his devotion to his political cause (OLD s.v. 5) and partisanship (OLD s.v. 6). However, this also makes it an ideal cover for coded subversion. Ahl, Bowditch and Porter have shown how subtly the civil-war nature of Actium is asserted. Just as subtly, and just as subversively, Lollius asserts the contingency of the Augustan victory. If he is lucky and careful, the patron will not notice, but will believe that he agrees with his views and give him two thumbs up.Footnote ⁷³

Coda: Lucan Reads Lollius

As a coda, let us turn to intertextual evidence that at least one ancient reader of the Epistles interpreted the description of Lollius’ naumachia in terms of a turning-point in history which could have gone either way. The word alteruter is, quite apart from its cacophonous jingle, very prosaic, so much so that it only occurs on four other occasions in extant Latin poetry before late antiquity.Footnote ⁷⁴ Three of these are in Lucretius, revelling in the uncompromisingly prosaic language of logical argumentation. Despite Horace’s frequent allusion to Lucretius, in the Epistles as elsewhere, it is hard to detect in line 64 any specific evocation either of the dilemma proving the infinity of the universe (1.974, 1012) or the ‘proof’ that the sun and stars are approximately the same size as they appear to be (5.589).Footnote ⁷⁵ However, in the context of undermining an assertion of predestined inevitability, a more general evocation of Epicurean anti-teleology by the use of a word with Lucretian flavour is distinctly possible.

The other occurrence is in Lucan, and is not only the same unpoetic word but in the same form (accusative singular)Footnote ⁷⁶ and in the same metrical sedes (start of the line) as in Horace (Luc. 6.6-8):Footnote ⁷⁷

The Bellum Civile frequently oscillates between ‘despairing anarchism’ and ‘the claustrophobic sense of trapped inevitability which is provided by the language of Fate’.Footnote ⁷⁸ Yet both of these are held in further tension with a counterfactual hope-against-hope that, from the perspective of Lucan’s pro-Republican implied reader, all may yet be well. This self-deluding hope is partly generated by the illusion that the poem’s events are playing out in an eternal present and that its outcome is as yet undecided, an illusion in turn produced by Lucan’s manipulation of enargeia, apostrophe, present tenses, and, of course, the explicit prescription that the poem’s future readers will treat its events as future rather than past (7.212-13: attonitique omnes ueluti uenientia fata, | non transmissa, legent).Footnote ⁷⁹ Yet for this illusion to instil hope in the Republican cause, it must be combined with a sense that a Pompeian victory is a possibility, a counterfactual scenario to anyone living outside that Lucanian bubble, or in the conjured eternal present of the narrative. The possibility and even the probability of victory for Pompey is virtually omnipresent from books one to seven as he is (unjustly) depicted as snatching defeat from the jaws of victory, from his decision to abandon Rome to his yielding to pressure from his troops to offer battle at Pharsalus.

Virtually omnipresent as this counterfactual ‘possibility’ is up until Pharsalus, there is a particular concentration in the Dyrrachium episode at the start of book six. Lucan depicts Pompey’s victory over Caesar here as all-but-certain so that the counterfactual scenario is particularly vivid and its frustration particularly demoralising. First, Lucan upbraids the Caesarian hero Scaeva who, in an inversion characteristic of the poem and its worldview, performs a hyperbolically heroic aristeia whose result is not glory and salvation but enslavement to an emperor for Scaeva, for Rome and for the world.Footnote ⁸⁰ Shortly afterwards, the narrator criticises Pompey for his lack of Sullan ruthlessness, a display of pietas which paradoxically leads to disaster for the Republican cause and Rome as a whole.Footnote ⁸¹ He goes on to list the subsequent disasters which would have been avoided by Pompey’s counterfactual victory here, climaxing in the death of Cato. It is in this context of counterfactual speculation that Caesar makes his characteristic prayer for a decisive battle which will mean destruction for either him or Pompey. With typical Lucanian pessimism, the emphasis is not, as in Horace, that ‘one or other’ will win but that chance will drown ‘one or other’. These are, of course, two sides of the same coin, but it is characteristic of the Bellum Civile to describe the side which falls facing down. Lucan has identified Horace’s subtle use of the very rare (in verse) word alterutrum to signal the contingent outcome of a civil-war battle and to evoke counterfactual scenarios where the other side might win. Then he has drawn attention to this reading by imitating it in an act of creative commentary, applying it to his own depiction of a civil-war battle with a contingent outcome.