I

In the second half of the third century bce Roman historical epic (notably that written by Naevius and Ennius) and Roman historiography (notably that of Fabius Pictor) came into being at roughly the same time. Whether and in what ways these two literary forms may have mutually influenced each other in their early development is a matter of debate,Footnote ¹ but it is obvious that there are both similarities and a generic difference, demonstrated by the use of prose or verse respectively and the accompanying style. Such characteristics enable a distinction between different types of narrative, even if the same events in Roman history are covered.

From a modern perspective, the presentation of historical events in epic form may seem strange, since nowadays recipients expect historical accounts to be objective, factual, and neutral, possibly written in a rather dry style with little poetic or rhetorical embellishment. At the same time historical periods and individuals may be portrayed in historical novels (such as those by Tom Holland or Robert Harris) or historical films. Because of the nature of such genres, poetic embellishment and ‘non-factual’ elements are included and accepted by audiences, though it may be asked whether recipients are always aware of what is ‘factual’ or ‘fictional’ or even make an effort to distinguish between the two, rather than preferring to accept the narrated story as a version of events.

The status of historiography and poetry, as well as their respective truthfulness as a mirror of reality, was discussed in the ancient world.Footnote ² Among the Greeks the most famous view is probably Aristotle's statement in the Poetics (9.1451a36–b7) in the fourth century bce:

When Aristotle makes a distinction between a literary genre describing what ‘has happened’ and another literary genre telling what ‘might happen’, this is basically a distinction between ‘fact’ and ‘fiction’.

In ancient Rome, too, literary people were conscious of differences between historiography and poetry. In Late Republican times Cicero has his friend Atticus, his brother Quintus, and himself discuss historiography in the treatise De legibus (c. 52 bce): Cicero defends ‘non-factual’ elements in his historical epic on his fellow-townsman Marius by claiming that for the epic genre the kind of truthfulness expected of a witness in court is not appropriate. This leads to the conclusion that different laws apply to historiography and to poetry, because historiography is concerned with truth and poetry is concerned with pleasure, although the interlocutors admit that in some early Greek historiographers there are a lot of ‘tales’, which blurs the boundaries (Leg. 1.4–5).Footnote ⁴ Elsewhere, when Cicero asks a friend to write a historiographical work about Cicero's consulship in 63 bce and to present his actions favourably (55 bce), he encourages the addressee to ignore the laws of historiography (Fam. 5.12.3). This implies the underlying view that in historiography a truthful and objective narrative is required; yet Cicero is prepared to allow the rules to be flouted on this occasion.Footnote ⁵

In the first century ce Petronius, in the satirical novel Satyrica, has one of the characters insist that one does not create a poem by describing historical events in verse, but that the presentation must be informed by an inspired and lofty style (Sat. 118). In principle, it seems, ancient writers believed that poetry and historiography differ and are each characterized by distinctive features, and that this has to do with what is narrated and how it is narrated. Nevertheless, in the ancient world, too, historiography was expected to be well-written prose, while epic poetry could deal with historical events in living memory.

Even though the passages looked at so far suggest, unsurprisingly, that historiography was seen as the more factual of the two literary genres, at the end of the first century ce the rhetorician Quintilian, assessing them from an orator's point of view, stated that ‘history is very close to the poets’, since historiography, like poetry, tells a story, while oratory aims to prove a point (Quint. Inst. 10.1.31–3). With this assumption of the similarity of historiography and poetry, Quintilian anticipates the modern discussion of the narrative character of historiography, associated with the American historian and literary critic Hayden White.Footnote ⁶ In research on classical literature, scholars tend to assume that there is a difference between historical poetry and historiographical descriptions, even though recent research has emphasized that there is not a simple opposition of ‘epic = fiction’ and ‘history = fact’,Footnote ⁷ for both literary genres are ultimately based on historical facts and elaborate them, yet they follow different conventions and pursue different aims. In order to define more precisely what is distinctive in classical historical epic, it seems promising to take advantage of current developments in research on fiction. An approach based on the categories established by this kind of research will aim to determine the respective proportions of ‘factual’ and ‘fictional’ elements,Footnote ⁸ as well as the relationship between them, and to analyse how ‘facts’ are employed in fictional texts.Footnote ⁹

Obviously, the terms ‘fact’ and ‘fiction’ are themselves in need of definition, both as descriptors in their own right and also with respect to whether they refer to elements of texts or to entire texts. According to the literary scholar Frank Zipfel, ‘fiction’ describes a text as a whole, while ‘fictional’ applies to individual elements.Footnote ¹⁰ For ancient texts, ‘fictional’ elements can be difficult to define both because of the historical distance and also because of the lack of available evidence on matters narrated in Latin historical epics. ‘Historical facts’ from the ancient Roman world are often based on literary sources (that is, historiographical narratives), such as those of the Latin historians Sallust, Livy, and Tacitus; yet these writers, too, pursue their own agendas, shape the subject matter accordingly, and present the material in line with literary conventions rather than as an objective record.Footnote ¹¹ For the purposes of this exercise, one can accept as ‘historical fact’ what emerges as a factual basis from such sources and is ideally supported by non-literary evidence.

A particular problem in this investigation is the status of gods, because assessing their role includes reconstructing and presupposing attitudes.Footnote ¹² Since people in the ancient world believed in gods and worshipped them, as non-literary evidence shows, there is a ‘factual’ basis, although, of course, the existence of gods can never be proved in the same way as, for instance, the details of a battle. One might therefore say that there are different shades of ‘factuality’, ranging from what is historically attested to views and beliefs generally accepted in a society. When gods are presented as making speeches or taking actions, this is undoubtedly ‘fictional’, in the same way as otherwise unattested speeches of generals or politicians in historical accounts cannot be regarded as ‘factual’. Such a description of the gods' status is not at odds with views of ancient authors: the Late Republican scholar Varro notes that gods in literature are different from gods in the ‘real’ world and have a ‘fictional’ dimension.Footnote ¹³

A similar difficulty applies to ‘mythical’ epics, such as Valerius Flaccus' Argonautica or Statius' Thebaid (and also Virgil's Aeneid, with an ambiguous status in this respect), since for the analysis of the mixture of ‘factual’ and ‘fictional’ elements in those texts one would first have to establish the ‘historical’ status of the mythical stories in the view of contemporary audiences; these epics will therefore not be discussed in the present investigation. Here, events and details in historical epics that are not attested in the historical record in the widest sense or are implausible in ‘real’ life, but that contribute to constituting another world with its own rules, will be regarded as ‘non-factual’ and thus likely to be ‘fictional’, especially when such features are characteristic of literary narratives or develop standard literary motifs.

This last point is important since the mixture of ‘factual’ and ‘fictional’ elements in a text does not in itself constitute a sufficient criterion to decide whether it is to be defined as ‘fact’ or ‘fiction’ in its entirety. Literary scholars such as Alexander Bareis stress that, ultimately, the status of texts can only be determined pragmatically, depending on how readers react to them.Footnote ¹⁴ This is connected not only with the subject matter presented but also with circumstantial factors such as context or style; an example from the modern world could be the different presentation and consequent reactions to a story about a crime in a newspaper article or in a detective novel.

On the basis of these points concerning terminology and definitions, the remainder of this article will be devoted to a look at paradigmatic passages from Latin historical epics. An attempt will be made to determine how ‘factual’ and ‘fictional’ elements have been combined and how they are presented. Discussion of some remarks by contemporaries is included so as to investigate historical views on ‘fact’ and ‘fiction’. Against this background, the status of Latin historical epics between ‘fact’ and ‘fiction’ can be considered. Since this study is concerned with general narrative forms rather than with the detailed interpretation of particular texts, a broad overview of key examples seems most useful, even if that means that individual instances cannot be discussed in depth.

II

The epic genre was introduced to Rome by a Latin version of Homer's Odyssey, the poem Odusia (only surviving in fragments) by Rome's first poet Livius Andronicus (c. 280/270–200 bce), probably written just before 240 bce.Footnote ¹⁵ This provided the basis for Naevius, Rome's second poet (c. 280/260–200 bce), to make the fledgling literary genres more Roman: his epic Bellum Poenicum deals with the First Punic War (264–241 bce); thus it is an epic on an event in Roman history and was probably written shortly afterwards. Testimonia on this fragmentary epic show that Naevius said in the poem that he had taken part in the war described (Gell. NA 17.21.45).

That authors make efforts to increase the probability of their narrative by claiming that they are eyewitnesses or have other trusted sources is a common topos in historiography, in the tradition of the Greek historian Thucydides (e.g. 4.104).Footnote ¹⁶ For instance, Naevius' contemporary Fabius Pictor (c. 270–210 bce), whose historiographical work also survives in fragments,Footnote ¹⁷ is said to have talked about a war in which he himself participated, and he is likely to have referred to this detail in his historiographical work. The reference to eyewitness accounts must have become such a common element in historiographical narratives that later, in the mid-first century ce, Seneca could open his satire Apocolocyntosis, which narrates the attempted deification of the emperor Claudius, with a mock version of this topos: as evidence for Claudius' path to heaven the poet points to a possible witness, who allegedly saw the limping Claudius go to heaven in his characteristic gait (Apoc. 1).

When Naevius mentions within the epic that he participated in the war described, this means that he tries to impart credibility to the narrative according to the model of historiography. Indeed, surviving fragments confirm a rather factual narrative, for instance with reference to military details.Footnote ¹⁸ Other fragments show that Naevius' epic included a section that told of Aeneas' leaving Troy and his subsequent adventures; this probably provided the background to the war in Naevius' time.Footnote ¹⁹ Irrespective of whether the story of Aeneas leaving Troy and coming to Italy is regarded as a ‘myth’ or as part of ‘Roman prehistory’, this is a period that Naevius cannot have experienced himself. Nevertheless, these parts are not completely ‘fictional’, in the sense of ‘invented by the author’ as an alternative world, since the poet follows established narrative traditions; at the same time, these stories cannot be ascertained as ‘facts’. Naevius seems to have used a more poetic narrative for those passages, and this may indicate an admission of their less ‘factual’ nature.

These pieces in particular were influential in the Roman poetic tradition, as far as it can be determined given the fragmentary state of Naevius' epic. The late antique author Macrobius often indicates that sections in Virgil's Aeneid have been adapted from Naevius, including the following case:

This statement implies, as directly transmitted fragments confirm, that Naevius' epic about historical events had active gods. This might look like a contradiction for modern minds, but may not have been a problem for contemporary audiences. Even if gods mainly appeared in the ‘mythical part’ of Naevius' epic (as it is sometimes called), the fact that Jupiter reveals the future to Venus means that a divine plan is outlined for the entire course of history, possibly covering later and thus ‘historical’ times. It is also likely, though not certain, that Naevius' epic included an encounter between Aeneas and Dido and that this was presented as the ultimate cause for the wars between Rome and Carthage.Footnote ²¹ The inclusion of such elements could be an attempt at explaining historical causality within an epic framework, which leads to a combination of literary elements and historical facts according to modern categories. For Roman readers, such features may not have been at odds with a basically historical narrative: the historiographical work by Fabius Pictor also included the figure of Aeneas and, according to Cicero (Cic. Div. 1.43), narrated a dream by Aeneas in which he saw his entire history (F 1 FRHist).

The historical epic Annales by the slightly later Republican poet Ennius (239–169 bce) is the first work in Latin to narrate the whole of Roman history from the beginning up to the author's own time. Ennius placed himself within a refined poetic and literary tradition, and he self-confidently voices literary aspirations for his work rather than emphasizing a truthful and factual record. In the opening of the epic, he presented himself as a ‘reborn’ or ‘second Homer’, a ‘Homer for Rome’;Footnote ²² and in Book 7 he claimed that earlier poets were ‘Fauns and Seers’, while he was the first dicti studiosus (that is, philologos, ‘interested in language’) in Rome (Enn. Ann. 207–9 Sk. = 232–4 W.).

Ennius seems to have used a more poetic method of presentation, even in undoubtedly ‘historical’ sections, as opposed to the narrative of the more ‘mythical’ beginnings of Rome.Footnote ²³ For instance, again according to Macrobius (Macr. Sat. 6.3.2–4), Ennius describes the fighting of a tribune in the Istrian War metaphorically by comparing an attack by a large number of javelins to a rainstorm and by vividly illustrating the effects on the warrior (Enn. Ann. 391–8 Sk. = 409–16 W.); thereby he takes up a similar presentation of a fighter employed by Homer in the Iliad (16.102–11) and later adopted by Virgil for the Aeneid (9.806–14). The fighting of this tribune is obviously presented in Homeric fashion. Because the war took place, because the narrative illustrates a possible way of hard fighting in distress, and because the description has been adapted to Roman conventions (such as the weapons used), a ‘factual’ basis for this poetic presentation is suggested, so that the boundaries between ‘fact’ and ‘fiction’ become blurred.

Elsewhere Ennius gives a portrait of the so-called ‘good companion’, a trusted friend of the nobleman and politician Servilius Geminus (consul 217 bce). Since a suggestion by the scholar Aelius Stilo in the first century bce,Footnote ²⁴ many readers have regarded this portrait of a man of high moral and intellectual qualities as an indirect portrayal of Ennius. While Servilius Geminus is a historical figure and may have had a loyal confidant, the identity and the character portrayal of this individual cannot be corroborated. Regardless of a possible connection to the ‘real’ Ennius the scene is ‘non-factual’, since a historical individual is given a companion, who, in this idealization, is a ‘fictional’ figure. However, the lack of absolute factuality in this section does not invalidate the overall historical narrative. The idealized portrayal of an aspect of Roman life adds atmospheric background to the historical facts and helps to characterize the protagonists.

Ennius' Annales was followed by a number of historical epics in the Late Republic and early Augustan periods, of which mostly just names and titles survive. The most interesting example, about which a little more is known, consists in Cicero's writings. In addition to the piece about his fellow-townsman Marius, Cicero composed, or at least planned, two epics about his own experiences during his consulship (De consulatu suo) and his exile (De temporibus suis) in the context of a wider publicity campaign in the years after his consulship of 63 bce. As fragments from De consulatu suo reveal, Cicero presented himself as the peaceful saviour of Rome in high epic style and basically aligned himself with the founders of Rome (F 11, 12 FPL ⁴). Both epics were meant to include a divine council, where Cicero's policies were justified and praised from on high: Cicero seems to have exploited the gods of the epic tradition for self-justification.

Interestingly, even in antiquity Cicero was subject to criticism because of his epics. For instance, the Pseudo-Sallustian invective mocks the fact that Cicero uses the divine council to legitimize his actions ([Sall.] Inv. in Cic. 2.3). This comment cannot have been directed against the use of this epic element, since the combination of contemporary historical events and divine actions was nothing unusual in Roman epic. However, in this case the author applied the convention not only to a living character but even to himself, and this stretched the limits of probability (or literary conventions and decorum) even for contemporary audiences. There may have been unwritten rules for the use of such ‘literary’ elements in ‘historical’ narratives; they seem not to have excluded the insertion of these features, but rather set limits to their use.

When one follows the development of historical epic in the Imperial period, the next clear example is Lucan's (39–65 ce) piece on the Roman civil war in the 40s bce, written in the time of Nero in the mid-first century ce and almost completely extant. Lucan is well known to differ from other Roman epic poets by having abolished the divine apparatus; his epic does not include anthropomorphic gods. One might think that this contributes to a more ‘factual’ narrative. But Lucan attributes an important role to the supernatural forces of an impersonal divine being, as well as to ‘fate’ and ‘fortune’, whose impact on the lives of humans is considered (for example, 2.1–15); he has a long scene with the witch Erichtho (Book 6); and in a debated passage in the proem he talks about Nero being deified (1.45–59). Hence, there is in fact a similar tension between ‘factual’ events and a supernatural level.

Lucan's subject matter consists of events that happened almost one hundred years earlier. In shaping the material, he reduces details drastically, especially those concerning military tactics and strategy. He also deviates from the historical record, which is probably not a sign of carelessness, however (as it was sometimes seen in the past), but rather of poetic intent. This applies, for example, to the remarkable instance of a speech that Lucan has Cicero deliver before the battle of Pharsalus (7.68–85), in which the speaker urges war and the elimination of Caesar, although it is known that Cicero was not present at Pharsalus at this point and had a different political attitude. However, the speech is meant to be a representative statement of the position of the senatorial faction prior to the decision, and it gains weight if put into the mouth of an eminent orator, as the introduction to the speech suggests (7.58–67).

Such a scene goes beyond adding a supernatural level or poetic embellishment to basically historical events, by introducing an incident that contradicts what is attested elsewhere. This might be obvious to readers who know the historical circumstances, though they may still accept it, realizing the poetic intention and thus assigning a different status to it. Readers who have no other information about the historical details will note the purpose of the speech, but they will be unable to determine whether a historical set-up has been exploited or whether the scene does not have a historical basis. In historiography too, again since Thucydides (1.22.1), it was acknowledged (and accepted in modern scholarship despite differences in the detailed interpretation of this passage) that the precise wording of speeches might not be historical but shaped by the historian. Yet Thucydides takes care to stress that he has individuals say what they could have said, and he probably would not put them into geographical locations where their presence is not attested.

Thus, ultimately, Lucan's work is not so different in the combination of ‘factual’ and ‘fictional’ features from other epics including anthropomorphic gods. However, because the influence of these forces is presented more indirectly, there was some ambiguity about the generic classification of his poem in antiquity. For instance, towards the end of the first century, the epigrammatist Martial has Lucan state in an epigram (14.194): ‘There are some who say I am not a poet; but the bookseller who vends me thinks I am.’Footnote ²⁵ This couplet reveals criticism of the classification of Lucan as a ‘poet’; at the same time it shows Lucan's popularity with readers and indicates the appreciation that Statius, Quintilian, and Tacitus confirm. However, Quintilian recommends Lucan to orators rather than to poets for imitation and thus indicates hesitation about his status as a poet, though this seems mainly due to his style (Quint. Inst. 10.1.90). Late antique commentators agree that Lucan does not deserve to be counted among poets because he has basically written a history.Footnote ²⁶

The most famous reaction is a comment on the epic narrative mode by his contemporary Petronius, which is generally regarded as referring to Lucan although his name is not mentioned. In his satirical novel Satyrica, Petronius has one of the characters, the poet Eumolpus, claim that the task of a poet of a historical epic is not to clothe historical facts in verses, which is the job of historians, who can do that much better, but to tell the story by means of a divine apparatus and fabulous elements (Sat. 118).Footnote ²⁷ This description is followed by the speaker's own mini-epic on the civil war, as an example of how an epic should be composed; this version gives ample room for action on the divine level. Essentially, a mixture of ‘fact’ and ‘fiction’ is explicitly demanded, though there is no statement on the required accuracy of ‘historical facts’.

The ambiguity about the generic classification of Lucan's piece and the unease with the character of the work may indicate that, for at least some ancient Romans, the involvement of gods was a decisive characteristic of epic. Lucan's abolition of the traditional activities of anthropomorphic gods in Homeric fashion and the relatively straightforward narrative argue against a classification as epic in the minds of critics. The alternative option, to define the piece as historiography, which may sound strange according to modern criteria, is also problematic for the ancients because of the poetic form, but is less so as regards the narrative style when historiography, too, is based on a literary presentation.

Unfortunately, we cannot trace the subsequent stages of this discussion. There were a number of historical epics, particularly on contemporary history, after Lucan, but of those only titles are known. What seems certain is that the genre developed with the times: while in the Republican period the deeds of the Roman people or achievements of representatives of the people were praised, in Imperial times the actions of the emperor are glorified. Most of the surviving epics of the period are mythical and thus avoid commenting directly on the emperor, apart from obligatory statements in the introduction.

The only extant historical epic of the late first century, the Punica by Silius Italicus (c. 25/30–101 ce), who reintroduces active gods, deals with an earlier period.Footnote ²⁸ In the proem (1.1–16) he announces his theme: he will talk about the second of the three Punic Wars (218–201 bce); in this war ‘each nation strove to destroy and exterminate her rival, and those to whom victory was granted came nearer to destruction’ (1.12–14).Footnote ²⁹ This presentation includes clear allusions to the opening of the third decade of the historiographical work by the Augustan historian Livy, one of Silius Italicus' sources (21.1.1–3): Livy also makes the points that the Second Punic War was the greatest and even the winners came close to ruin. At the same time, Silius Italicus calls the Romans ‘Aeneadae’ (‘sons of Aeneas’) and the Carthaginians ‘the people of Cadmus’ (1.2; 1.6): he thereby alludes to the mythical founding fathers of the two nations, adding the dimension of the mythical background.

Immediately after the proem, Silius Italicus moves on to revealing the causes of the war and gives reasons on what could be called divine, mythical, and historical levels, by referring to the influence of the goddess Juno, to the legendary Carthaginian queen Dido, and to the character and upbringing of Hannibal. This is perhaps in line with the suggestions made in Petronius, but from a modern perspective it may give the narrative an ambiguous status between ‘fact’ and ‘fiction’.

Such an ambiguity characterizes the entire existence of Scipio, who gains prominence as the protagonist on the Roman side in the epic's second half. He is introduced with a double parentage, with both a mortal and a divine father (Jupiter). Perhaps the fact that Scipio also has a mortal parent (and not just a divine parent on one side, like Aeneas) is a concession to the historically attested father, whose death is a motivation for the epic Scipio to fight the war energetically. At the same time the epic ends with his triumph and a statement confirming his divine origin (17.651–4).

In Book 4 Scipio is presented as saving his mortal father in battle, by carrying him off on his shoulders in the manner of Aeneas saving his aged father from burning Troy (4.454–71). The story that Scipio rescued his father is also found in other authors.Footnote ³⁰ Livy, however, preserves two alternative traditions (Liv. 21.46.7–8), including another one according to which a Ligurian slave was the rescuer (Liv. 21.46.10). Hence, the idea that Scipio saved his father may not be ‘factual’; a ‘fictional’ dimension clearly applies to the detail that Scipio acts like Aeneas, which does not appear in other accounts: this feature is meant to characterize Scipio as a dutiful son by an allusion to a literary motif and an element of Roman tradition.

That, out of grief at the deaths of his father and his uncle, Scipio engages with the underworld in Book 13, is an evidently ‘non-factual’ element, with literary precedents in both Homer's Odyssey and Virgil's Aeneid. Yet neither Odysseus nor Aeneas are figures on the same level as a Roman politician with reference to their ‘real’ existence. The first of his relatives to whom Scipio talks in this sequence is his mother, Pomponia (13.615–49). She is introduced as Scipio's mother by Jupiter owing to divine machinations; in the meeting with Scipio she talks about her encounter with Jupiter. The narrative continues straight afterwards with Scipio meeting the spirits of his father and his uncle (13.648–53). The two versions are thereby immediately juxtaposed. Apart from the fact that stories of double parentage also existed for mythical characters in the ancient world, especially for Hercules, it is understood that they operate on different levels, and therefore they do not contradict each other completely (as with instances of double motivation on a human and a divine level in epic more generally). It is left open which story is more plausible until the final lines of the poem.

There are, however, indications that particularly obvious clashes between the ‘factual’ and the potentially ‘non-factual’ worlds are avoided in Silius Italicus: there are a number of divine scenes with gods interacting with each other and making speeches in the epic tradition, but they seem to take place in indeterminate locations. One gets the impression that the gods are not presented as anthropomorphic entities, but that they rather serve to add a further layer to the narrative and to provide an explanation for events and their future ramifications and thus to endow them with universal truths.

The divine sphere is not the only ‘non-factual’ level by which Silius Italicus adds another dimension to his historical epic: when he narrates battles in Sardinia in Book 12, he describes how the poet Ennius fights successfully as a Roman centurion; Ennius is supported by Apollo, and the god of poetry predicts Ennius' literary achievements (12.387–419). According to other ancient sources, Cato brought Ennius to Rome from Sardinia,Footnote ³¹ and it is often assumed that Ennius was in Sardinia because of military service. Whether more details were known in the first century ce, when Silius Italicus was writing, is uncertain. In any case, Ennius' appearance in Silius Italicus' work is in the context of T. Manlius Torquatus confronting a rebellion in Sardinia in 215 bce, and Cato is said to have brought Ennius to Rome from Sardinia in 204 bce.

Even if Ennius participated in this military confrontation, the details mentioned are unlikely to be ‘historical’ facts, but will rather be based on poetic fiction. However, the scene will not strike readers as completely implausible when it is known that Ennius was in Sardinia. Still, it is improbable that Ennius, who cannot have been a Roman citizen at the time (though he was later granted citizenship), had the rank of a Roman centurion in this war. This status, attributed to him by Silius Italicus, is rather meant to enhance his standing. Again, because it is known that Ennius supported the Romans, and it is presupposed that a great poet is also a great fighter, readers can accept this version. The involvement of gods in the scene further reduces its ‘factual’ appearance. Yet Apollo's speech makes it possible to name characteristics of Ennius' poetry relevant for Silius Italicus and the Roman epic tradition, namely the introduction of the hexametric metre and the presentation of Roman wars in verses. Ennius is thereby turned into a direct predecessor of Silius Italicus. This might be one of the reasons why the scene is shaped in this way, and readers will subconsciously notice that there is a metaliterary layer, which has necessitated ‘non-factual’ elements.

Ennius is not the only poet to appear in Silius Italicus' narrative.Footnote ³² In Book 13 an encounter with Homer is included in Scipio's engagement with the underworld (13.778–97). In this way, Silius Italicus manages to integrate the pioneer of epic poetry into a historical epic on the Second Punic War. In contrast to his encounters with other ghosts, Scipio does not talk to Homer. The poet remains a distant figure, and Scipio is informed about his identity by the Sibyl, who tells him that Homer ‘raised the fame of Troy to heaven’. Scipio's scenario that Homer could praise the achievements of Rome, obviously with the same effect, can be read as a metaliterary statement on Silius Italicus' own poetry. The ‘non-factual’ setting in the underworld makes it possible to present Homer's Iliad and Silius Italicus' epic on an event from Roman history on the same poetic level. In this case, the segregation of the ‘real’ and the ‘non-factual’ worlds is the vehicle for the poet's specific message.

Another interesting mixture of historical and literary elements is the story of Scipio at the crossroads: in an allusion to the tale of Hercules at the crossroads (told by Prodikos), Virtus and Voluptas appear to Scipio (15.18–128). Because of the personified figures of Virtus and Voluptas it is likely that the scene is poetic fiction, arranged according to a literary model.Footnote ³³ At the same time, readers familiar with epic conventions will recognize the poetic way of narrating how a decision is formed within a protagonist's mind. If the particular shape of the scene is accepted as a form of presentation sanctioned by the poetic conventions of the literary genre rather than a record of actual events, it will enrich the narrative and not contradict the historical basis.

Silius Italicus' epic turns out to be particularly nuanced in the combination of ‘factual’ and ‘non-factual’ elements: it demonstrates that a historical epic may not only include literary, ‘non-factual’ motifs from the epic tradition, such as the divine sphere or the underworld, but that this mixture can also be used to convey metaliterary statements or to illustrate internal processes.

III

When, at the end of this overview of key historical Roman epics in the classical period, one returns to the original question of the combination of ‘factual’ and ‘fictional’ elements in Roman epic in relation to contemporary historiography, it is clear that the distinction is not straightforward, because ancient historiography, in contrast to modern generic expectations, may include ‘non-factual’ features and poetic embellishment. The issue becomes even more complex when one takes the genre of drama into account, since poets in Rome developed a separate genre of serious drama on events from Roman history (called fabula praetexta), as opposed to tragedies on stories from Greek myth. These historical dramas could present incidents from the recent past using the narrative modes of Greek tragedy, which means that actions and utterances of historical figures were displayed with the means of literary, ‘non-factual’ narrative, and that historical characters were impersonated by actors.

Obviously, because of the epic or dramatic form and the corresponding literary conventions, ‘non-factual’ additions, insertions, or narrative modes are not felt to be incongruous in such genres, apart from exaggerated instances such as Cicero's case, while poetic presentation of events is explicitly demanded for epics in Petronius. In comparison with historiography and historical drama, there seem to have been specific narrative modes for epic, established by tradition, to such an extent that they are not only used in mythical epics but can be transferred to historical epics without affecting the plausibility of the narrative. As well as the metre, these include the poetic language; scenes going beyond a sketch of the mere sequence of events, such as visits to the underworld; and motifs, such as the determination of events or decisions of human characters by divine intervention. On this basis, further ‘non-factual’ details such as the self-presentation of a poet may be added.

Such elements that are recognized on the basis of narrative conventions can be described as ‘fictional’, whichever definition of ‘fiction’ one adopts. Yet one has to bear in mind that ‘conventions’ may vary according to readers' experiences, and that divine scenes might therefore have seemed less ‘fictional’ to ancient audiences than they do to modern ones, since at least their basis had a connection to ‘real’ life. Moreover, these literary conventions and ‘fictional’ elements create another world of their own, violations of which will be noted and which has a ‘historical’ dimension in a further sense, since it provides insight into the collective thinking of a society. The conventions of ‘epic’ narrative give historical epics an ambiguous status, more so than historiographical accounts: it is clear that the narrative partly relies on ‘facts’, but is also partly ‘fictional’ in the sense that it cannot be strictly based on eyewitness accounts or reliable sources.

As modern literary theory has noted,Footnote ³⁴ it is not possible simply to juxtapose or combine ‘factual’ and ‘fictional’ elements; instead ‘fictional’ additions will make the entire set-up ‘fictional’. A famous example from a more recent period is the ‘fictional’ figure of Sherlock Holmes living at 221b Baker Street in London: obviously London and Baker Street exist, and readers will have an idea of these, and thus visualize what is sketched in the narrative accordingly. However, the London described in books about Sherlock Holmes is not the real London, because in real London there is no 221b Baker Street at which Sherlock Holmes lives. Similarly (mutatis mutandis), there is no historical Punic War in which Scipio talks to the shades of dead human beings; therefore the introduction of such a scene makes the entire series of events ‘fictional’. The combination of elements ‘factual’ in origin and their ‘fictional’ elaboration results in ‘fictional’ texts on a ‘factual’ basis, which creates the appearance of even the ‘fictional’ aspects being ‘factual’.

Nevertheless, because readers know that the narrative is based on historical events, they do not normally reflect on the tension between ‘fact’ and ‘fiction’ and thus may accept the story as a version of ‘history’. The ‘fictional’ additions often serve to provide reasons for and background information to events. If one reads these epics with modern literary theory in mind, one will become aware of the literary techniques used to present historical events in poetic form and can assess how readers' impressions are manipulated, since conveying an interpretation of events or ‘universal truths’ may be more important than a mere record of incidents. This might explain why the effort was made to produce historical epics: as Scipio's comments on Homer in Silius Italicus suggest, a poetic and partly ‘fictional’ presentation enhances not only the value of the poetry but also the standing of what is narrated, namely the history of Rome.