1 Embedded Narratives in the Punica
The time that a study on Silius Italicus’ Punica had to start with an excusatio explaining why on earth the author had undertaken the effort of studying the allegedly most boring extant epic from Latin literature is long gone. Nor it is necessary to quote from Pliny’s obituary the infamous words maiore cura quam ingenio (‘with more care than talent’, Ep. 3.7).1 Since Scaliger this phrase has served as proof of Silius’ mediocre abilities as a poet (if he even deserved that title).2 Numerous articles, monographs, conferences, commentaries on single books, and an inevitable companion that have been published in the past two or three decades show that the Punica is a text that deserves serious scholarly attention and is able to evoke a lively debate. It is perhaps an exaggeration to speak of an aetas Siliana, but this millennium certainly has seen an enormous boost in Silian studies—in the slipstream of the renewed interest in the Flavian literature and age in general, which has been going on for somewhat longer.3
The Punica has often been read as a ‘revisionist’ epic—a return to a supposed Virgilian optimism after Lucan’s dire epic on the civil war Bellum Civile or Statius’ horror-epic Thebaid. How the message of Punica should be understood is an ongoing debate, some allowing for a more positive interpretation (e.g. Feeney and Vessey), others rather highlighting negative or ambivalent sides of the epic (e.g. McGuire, Dominik, and Tipping). The same, of course, applies to the debate on Silius’ most important model-text, the Aeneid, in which both “optimistic” and “pessimistic” voices have been detected.
This study aims to shed more light on the complexity of the Punica by investigating four embedded narratives, demarcated passages that at first sight may seem to be at the fringe of the epic or that are at least less obviously connected to the main narrative of the Second Punic War.4 The narratives under investigation are, however, not merely loosely ‘embedded’ in the epic, but foreshadow or look back on elements that are found elsewhere in the main narrative. In this way, they serve as ‘mirrors’ of the main narrative, in that they reflect upon certain aspects of it. It is of course never a one-on-one reflection (that would mean that the embedded narrative would be an exact copy of the main narrative),5 but it is clear that they resemble other parts of the epic or even the epic as a whole. These connections between the main narrative and embedded narratives are forged by an intricate poetics of intratextuality, which functions in a comparable way as intertextuality does in Latin epic poetry: marked words or phrases can activate an intratextual link and subsequently invite a comparison.
At the same time, these embedded narratives are in dialogue with a plethora of other texts—they are as ‘hyper-allusive’ as the rest of the Punica.6 These intertexts give shades of meaning to the embedded narratives, and by proxy also to the main narrative. In this thick forest of allusions, it is easy to get lost. I hope that I will be able to guide my narratees so that in the end they still can see the wood for the trees. Since the texts of this study are narratives, which themselves are embedded in a narrative, I found it necessary and useful to combine my intra- and intertextual readings with narratological vantage points. Sometimes these different methodological outlooks may overlap, for example when the narrator of the embedded narrative is a character in the main narrative or when he recalls a narrator from another text.
Embedded narratives are a standard feature of epic since Homer—think only of the long narrative of Odysseus on his adventures (Books 9–12 of the Odyssey) or that of Aeneas on the fall of Troy and his journey across the Mediterranean (Books 2 and 3 of the Aeneid). In these narratives, the protagonists (Odysseus and Aeneas) and the major themes (the journey of the hero) coincide with those of the surrounding main narrative. Other embedded narratives deal with other persons and events, and thus are not directly connected with the characters and plot of the epic, although there may be thematic connections. A good example is the song of Demodocus on Ares’ adultery with Aphrodite (Od. 8.266–366). It is this type of ‘free’ embedded narratives that I will be examining. In fact, all embedded narratives in the Punica are of this ‘free’ kind; there is no example of an embedded narrative so strongly connected with the main narrative as those by Odysseus and Aeneas. This does not mean, however, that these ‘free’ embedded narratives are totally disconnected from the main narrative; the connections are simply less obvious. The nature of these embedded narratives fits into a larger pattern of digressions from the main narrative which abound in the Punica, such as aetiologies and ecphrases. This episodic structuring of an epic is reminiscent of Hellenistic literary preferences. It especially recalls the concatenation of stories in Ovid’s bold epic experiment Metamorphoses. Wilson even contends that “there are many digressions along the way, often inspired by Ovidian prototypes, so that the Livian Punic war narrative can also be read as a unifying principle for a diversity of aetiological, mythological, creative-epic inventions, not dissimilar in concept to the Metamorphoses.”7
It is thus only to be expected of my hypothesis that these embedded narratives, although they are no part of the plot, are both intricately connected with the main narrative and central to the understanding of the Punica. An increased interest in these narratives over the last years shows their critical potential. Next to articles on single embedded narratives, we have the survey of Schaffenrath on embedded narratives in the Punica and Walter’s monograph on narrators in Flavian epic.8 This study builds on their research, but aims to provide a fuller exploration of embedded narratives in the Punica. The larger part of this book deals with the four case studies in which I provide close-readings of embedded narratives. They show how embedded narratives in the Punica operate in practice. In this introductory chapter, I first briefly discuss embedded narratives in general and in the Punica more specifically. Then, I give a first taste of my combined methodology of narratology, intertextuality, and intratextuality by applying it to the embedded narrative of Proteus. It shows that a narratological vantage point strengthens both intertextual and intratextual readings.
2 Embedded Narratives and Their Functions
The simplest definition of an embedded narrative, or tale within a tale, is “a narrative that is embedded in the main narrative, either told by the primary or a secondary (or tertiary etc.) narrator.”9 Simple as this definition is, it contains an important nuance to other definitions which suggest that embedded narratives are merely situated at a secondary narrative level.10 Embedding is, however, not necessarily a change in narrative level. This is the case when the primary narrator gives an external flash-back or flash-forward, narrating events which fall outside the time limits of the main narrative. The two examples of this type that I discuss, the narratives of Falernus and Anna, are distuinguised from the main narrative by separate prooemia. Other examples where the narrator remains the same, show a change in the nature or ‘universe’ of the narrative. This is the case, for example, in the narration of dreams, in which events “take place in an alternate universe created by a character’s mind rather than being physically carried out in the spatial-temporal universe of the rest of the narrative”, as Nelles explains.11 Examples of this type in the Punica are the dream of Hannibal (3.183–213) and the aetiologically motivated narrative of Pyrene, a nymph that was raped by Hercules in the mythological past (3.415–441). Both of these embedded narratives take place in another ‘universe’, but are told by the primary narrator.
Embedded narratives can have several functions: they can be explanatory, predictive, thematic, persuasive, distractive, and delaying.12 When they serve as flashbacks (analepses) of events inside (internal or intradiegetic) or outside (external or extradiegetic) the main narrative, we call this the explanatory function. An example from the Punica is the embedded narrative of Dido’s sister Anna in Book 8, whose journey to Italy and transformation in a water nymph shed light on her role as instigator of Hannibal in the main narrative (see Chapter 4). Other embedded narratives are flash-forwards (prolepses) and can be called predictive. This is the case in prophesies of several gods, such as the oracle of Jupiter Hammon in Book 3 that is reported to Hannibal through an intermediary (see Chapter 1). In other cases, such prophesies are mainly intended for the primary narratees, as is the case in the embedded narrative of Proteus in Book 7, whose story is not heard by any character from the main narrative except the water nymphs, his direct narratees. Third, embedded narratives can have a thematic function if they share one or more themes with the main narrative. The longest embedded narrative of the Punica, about Regulus in the First Punic War, is a good example, as it deals with uirtus (‘virtue’ or ‘heroism’), patientia (‘perseverance’), fides (‘loyalty’) and perfidia (‘perfidy’), key-concepts of the epic as a whole (see Chapter 2). A fourth function is persuasive: this is the case when the embedded narrative is intended to influence the course of the main narrative. An example is Cilnius’ embedded narrative on the bravery of the three hundred Fabii (7.34–68). Cilnius, a prisoner of war, relates this story in order to arouse Hannibal’s anger and so to get killed. The primary narrator explicitly states this goal in the introduction to his speech: ‘this man longed for an end to his misery and to break his life’ (hic ardens extrema malis et rumpere uitam, 7.33). After Cilnius has finished his story of the Fabii, Hannibal, his direct narratee, sees through this plan and refuses to give him what he wants: ‘ “In vain you arouse my anger, fool,” he said, “and seek to escape from your captive chains by dying” ’ (‘nequiquam nostras, demens,’ ait, ‘elicis iras | et captiua paras moriendo euadere uincla’, 7.70–71). In this case, the persuasive function of the narrative is annulled, as Cilnius is forced to stay alive. The fifth function is distractive, when the narrative is primarily told to entertain. In the Punica, the second song of Teuthras in Book 11 is an example. When he is asked to entertain Hannibal and his entourage, the bard carefully ‘selected out of many the following song as the most graceful for the dinner’ (haec e multis carpsit mollissima mensae, 11.439).
Needless to say, embedded narratives can have more than one of these functions at the same time. In such cases, it is often useful to make a distinction between the function on the level of the secondary narratees (argument function) and on the level of the primary narratees (key function). The second song of Teuthras, for example, is distractive for the Carthaginians (argument function). The primary narratees, however, are able to see a persuasive function of this narrative that is lost to both Teuthras and his audience (key function). Venus, as the primary narratees have learned shortly before, has ordered her Cupids to enfeeble the Carthaginians with Wein, Weib und Gesang (11.405–409); Teuthras song fits perfectly into this divine scheme and manages to achieve the desired effect on the Punic army, as the primary narrator remarks when the bard has finished his story: ‘So then the Pierian Teuthras was breaking the breasts of these men, hardened by wars, by a Castalian song’ (sic tunc Pierius bellis durata uirorum | pectora Castalio frangebat carmine Teuthras, 11.481–482). From this point onwards, the strength of Hannibal and his army starts to wane. The hibernation in Capua amidst all sorts of luxury has affected the war spirit of the Carthaginians, as the primary narrator stresses when they set out from the city on a new campaign (12.15–19). Later in Book 12, Hannibal’s attack on Rome is checked by the gods, which marks the decline of Carthaginian dominance in the war and the epic and eventually leads to their defeat. The entertaining song of Teuthras therefore plays a significant role in changing the course of the main narrative, without the Carthaginians having been aware of it.
3 Embedded Narratives and Tactics of Delay
The Punica is full of sideways and digressions. The epic abounds in ecphrases, etymological vignettes, and embedded narratives. This is a feature that the Punica shares with other imperial epics. “The expansive, digressive nature of these epics has always been seen as one of their worst features”, as Fowler observes.13 For a long time, digressions have been viewed as redundant or even disruptive for the unity of the epic.14 At best, they were seen as artful interruptions of the main narrative. Steele, for example, argues that variety was the most important function of the embedded narrative on Regulus: “Silius tried to vary the monotony of historical narrative by the introduction of a hero of the First Punic War.”15 Ribbeck, who in general is not unsympathetic about such digressions in the Punica, also stresses their distractive function. The aetiological narrative of Falernus in Book 7 is a “gemütliche Episode”, and “die hübsche Schilderung” of Pan in Book 13 is “[ü]berraschend und wahrhaft erquicklich”.16 Stärk, in an article on the prophesy of Proteus, looks for the motivation of such digressions in the biography of Silius, whom Pliny in Ep. 3.7 had labelled as an excessive
The view on these digressions vis-à-vis the epic as a whole has drastically changed, especially in the last few decades. This study aims to show for the embedded narratives how intricately they are connected with the main narrative. This said, there is of course a sense of ‘separateness’ to these embedded narratives, precisely because they are embedded in a larger whole. Due to their nature, they ‘pause’ the course of the main narrative and contribute to a sense of delay. Masters has convincingly argued that the delays in Lucan’s Bellum Civile should be seen as conscious acts of the narrator to procrastinate the continuation of the civil war. “[P]owerless as Lucan may be to prevent the final catastrophe, he has at least the power, as poet, of delaying it within his poem; we can conclude, then, that Lucan is anxious to display his reluctance to allow the action to proceed, and he achieves this by erecting barriers that are at once literary and artificial. But again there is more. Although Lucan is reluctant, he does yet continue the action; and in writing the poem he is allowing the civil war to be re-enacted, he is re-enacting the war.”20 Mutatis mutandis, this can also be said of Silius and his Punica. The narrator is consciously creating a re-enactment of the greatest defeats from Roman history, culminating in the Battle of Cannae that occupies the central books of the epic.21 It has been suggested that the 17 books of the epic refer to the duration of the war: 17 years (218–202 BC).22 The years of the war are, however, not evenly distributed over these books: the first two years up to the Battle of Cannae take up Book 1–10, whereas the other fifteen years are dealt with in the seven remaining books.23 Compared to Livy, who treats the same events in his third decade, Silius is delaying the build-up to Cannae. As Fucecchi points out, Punica 1–10 correspond to Book 21 and 22 of the Ab urbe condita, while the remainder of seven books compresses the material of Book 23–30.24 It is as if the narrator on the one hand is delaying the zenith of Hannibal’s campaign, whereas on the other he is speeding up its nadir at Zama. Embedded narratives form an important means to bring about such retardation of the main narrative. It should come as no surprise, then, that most of these embedded narratives are clustered in the first eight books of the Punica, including the two longest ones, the story of Regulus in Book 6 of about 500 lines and the story of Anna Perenna in Book 8 (almost 200 lines). Ribbeck already suggested that there is a broader connection between the Regulus episode and the surrounding main narrative: “Durch die Ernennung des Fabius kommt der Krieg zum Stehen. So füllen das sechste Buch wesentlich episodische Rückblicke auf den ersten Krieg.”25 In Book 6, the narrator copies, so to speak, the delaying tactics of the Roman general by inserting a flashback of the First Punic War. This poetical copycatting continues in Book 7, where we find the embedded narratives of Falernus and Proteus. They momentarily manage to pause the narration of the war in Campania at the same moment that Fabius tries to delay the war in the main narrative.26 Interesting, too, is that these latter two narratives display generic modes other than ‘essential epic’, such as aetiology, theoxeny, elegy, and comedy.27 So, on a generic level, too, the embedded narratives in Book 7 form part of an anti-martial, anti-epic agenda.28 But in spite of all these delaying tactics by both Fabius and the narrator, the martial exploits of Hannibal are unstoppable.29
4 Embedded Narratives as Mise en Abyme: The Example of Proteus
Embedded narratives are fertile ground for literary self-reflexivity. As miniatures of storytelling they are often mirroring (parts of) the main narrative, the primary narrator, and/or its primary narratees.30 In this way, they can represent the epic as a whole and shed light on the poetics behind it. In such cases, we can speak of an embedded narrative as a mise en abyme, to use the famous term coined by Gide.31 This reflexive aspect of embedded narratives is especially foregrounded when the secondary narrator is a bard, the traditional symbol of epic poets themselves. As Fowler puts it, “[a] poet within a poem is the most obvious form of mise en abyme.”32 Prophets are the other favourite surrogates for the poet, as both are uates that are imbued with prophetic powers.33
As Schaffenrath observes, mise en abyme “ist eine bei Silius sehr beliebte Technik.”34 With regard to the Punica, the phenomenon has most extensively been treated by Deremetz.35 He applies the theory of Dällenbach to the two narratives of Teuthras in Book 11, arguing that they operate as mises en abyme on different levels: they reflect parts of the main narrative, the act of narrating, and the literary history that Silius tries to reconstruct.
I want to illustrate this phenomenon with the embedded narrative of Proteus (7.409–493), one of the examples of mise en abyme in the Punica that are listed by Deremetz.36 As in the song of Teuthras, we can discern multiple aspects of mise en abyme. The most obvious one is the prediction of the course of the Hannibalic war. The Roman defeat at Cannae, Hannibal’s attack on Rome, and the Roman victories at the Metaurus and at Zama that are predicted by Proteus will all appear to be major events in the ensuing main narrative of the Punica (Book 9, 12.479–540, 15.493–823, and 17).37 Proteus can be viewed as mirroring the primary narrator. This becomes especially clear in the introduction to his narrative:
tunc sic, euoluens repetita exordia retro,incipit ambiguus uates reseratque futura (7.435–436)
Then the ambiguous prophet starts, unfolding the origins thatagain are repeated, and reveals future events.
These lines introduce Proteus and his narrative, but could well have been said of the primary narrator himself, who also intends to uncover the origins of the war in his Punica. Compare the introduction of Proteus’ narrative with the last line of the epic’s prooemium: iamque adeo magni repetam primordia motus (‘and now I will bring back the origins of this great upheaval’, 1.20). In addition, exordia and incipit signal an allusion to the very first word of the epic (ordior, ‘I begin’).38
Proteus’ narratees, too, can be interpreted as mise en abyme for the primary narratees of the Punica. When a Carthaginian fleet lands at Caieta, they might wonder whether this is a reversal of Aeneas’ arrival at the same port, or even whether the Punica brings about a reversal of the Aeneid.39 Cymodoce, the oldest of the Nereids, fires a barrage of questions at Proteus. These reveal that she is not only a spokesperson for the other nymphs, but also for the primary narratees: quid Tyriae classes ereptaque litora
5 Theory and Method
The embedded narratives in this study will be examined on a narratological, intertextual, and intratextual level. It is my firm belief that the combined application of these three theoretical frameworks can yield rich and meaningful readings of embedded narratives in Flavian epic.40 In this section, I will in brief compass review the ideas and terminology that I use from these theories, and in passing sketch the possibilities of these theories for embedded narratives. I will illustrate this, again, with the narrative of Proteus in Punica 7.
5.1 Narratology
The study of embedded narratives is inextricably bound up with issues of narrative levels; in many cases, not the primary narrator (the ‘poet’), but a character takes over the role of (secondary) narrator; in turn, they often introduce speaking characters of their own, with which we descend to the tertiary level. In order to say something about the meaning and interrelation of the main and embedded narrative and their narrators, it is necessary to keep clear track of these different levels of narration. The foundation of thinking about narrative levels is irrefutably the work of Genette;41 his terminology, however, is somewhat arcane for the uninitiated reader. Following De Jong, I will therefore use ‘primary’, ‘secondary’, and ‘tertiary’ to refer to narrative levels instead of ‘extradiegetic’, ‘intradiegetic’, and ‘metadiegetic’.42 When narrators play a part in their own story—which is often the case in embedded narratives in the Punica—I speak of ‘internal’ narrators; when they do not participate in their narrative, I call them ‘external’ (as for example the primary narrator of the Punica).43 By analogy I will speak of primary, secondary, and tertiary narratees. To avoid needless confusion, I have refrained from adducing the concept of the implied author/poet, or implied readers/addressees. I do, however, allow myself the use of ‘Silius’ as an occasional metonym for the primary narrator of the Punica, especially in places where narratology is more in the background of the discussion. Since De Jong’s application of narrative theory on the epics of Homer, narratology has become one of the mainstream methodologies for studying ancient narrative texts.44 Most of its terminology such as analepsis, prolepsis, and metalepsis, will by now be familiar to most classical scholars.45
When we apply the narratological terminology to the Proteus episode, we can speak of the divine prophet as the external secondary narrator. The nymphs that come to him in distress and ask about the future are the external secondary narratees of his story. He addresses his daughters explicitly (at uos, o natae, 7.479), when he warns them to stay clear from the Adriatic Sea near the mouth of the Aufidus, because the Battle of Cannae will stain these waters with blood.
His narrative contains both analeptic and proleptic elements, as the introductory lines make clear: he will both reveal the origins (exordia, 7.435) of the current war and disclose its future course (futura, 7.436). Indeed, Proteus starts with an external analeptic story on the Judgement of Paris, the Trojan war, and the quest of Aeneas (7.437–475). He then continues with an abstract, a general prophecy of Rome’s empire (7.476–478), followed by an internal prolepsis of the Second Punic War (7.479–491). The end of Proteus’ prophecy exceeds the boundaries of the main narrative: the last two lines are an external prolepsis about the Third Punic War and Rome’s final victory over Carthage (7.492–493).
In his longest analepsis, the Judgement of Paris, Proteus introduces Venus as a tertiary narrator (7.449–457). She mirrors Proteus as a narrator, in that she addresses her children, the Cupids (alloquitur natos, 7.449). Like Proteus, she is able to foresee the future and predicts her own victory in the beauty contest (uictoria nostra, 7.455). This is an internal prolepsis in Proteus’ narrative, as the sea-god will narrate Venus’ triumph only a little later.
Proteus’ narrative remains unobserved by any other character on the secondary level. This is emphasized when Proteus has finished his speech and the primary narrator makes a transition back to the main narrative:
quae dum arcana deum uates euoluit in antro,iam monita et Fabium bellique equitumque magisterexuerat mente ac praeceps tendebat in hostem. (7.494–496)
While the prophet revealed these secrets of the gods in his cave, the master of the war and the knights [i.e. Minucius] had already put from his mind the warnings of Fabius and was pressing head over heels towards the enemy.
The secrets of the gods (arcana deum) are revealed to the nymphs and the primary narratees, but the agents of war in the main narrative are unaware of these warning prophecies and continue their doomed warfare.
5.2 Intertextuality
Intertextuality has proven to be a particularly meaningful way of approaching Latin poetry in general and Flavian epic in particular, and therefore hardly needs justification.46 The Punica is one of the highly intertextual epics from this period that plays an intricate game of imitatio et aemulatio with the earlier literary tradition, both Greek and Latin.47 This is especially the case for the earlier epic tradition and Virgil in particular.48 The first two words of the epic (ordior arma) already signal this; Barchiesi reads these words as ‘I am beginning: here is a poem in the Virgilian tradition’.49 From the outset it is clear that the Punica is a continuation of the Aeneid, a realization of Dido’s curse in Aeneid 4.621–629 and a resurgence of Juno’s wrath that was seemingly put aside in Aeneid 12.841–842.
The Punica is also a continuation of the epic predecessors of the Aeneid. This is made explicit when Ennius is staged as a soldier in 12.393–414. Apollo as god of poetry prevents the poet-warrior from being killed in battle. The god predicts that Ennius ‘will be the first to sing of Italian wars in noble verse and will raise their commanders to heaven’ (hic canet illustri primus bella Itala uersu | attolletque duces caelo, 12.410–411). This is of course an allusion to Ennius’ Annales, which among other events deals with the Second Punic War. Apollo’s words also recall the first line of the Punica: ‘I begin the war, by which the fame of the Aeneadae was raised to heaven’ (ordior arma, quibus caelo se gloria tollit | Aeneadum, 1.1–2). The implication seems to be that Silius is a true successor of this first epic poet to write in hexameters of Roman wars (not Greek wars, like Livius Andronicus); Silius can be said to pose himself as an epic poet in the true Ennian tradition of historical epic, as authorized by Apollo himself.50 Upon reflection, the opening word ordior (‘I begin’) seems, however, to challenge the ‘first-ness’ of Ennius (primus): Silius begins to narrate the Hannibalic war. In this time of narration, the Annales still have to be composed, and their composition is only predicted. This makes Silius’ Punica ‘first’.51 Homer, too, appears as a character in the epic. In Book 13, Scipio meets the bard’s ghost (13.778–797). The Roman hero wishes that the great bard of the Greek would still be alive, to sing of his deeds: ‘If fates would permit that this poet would sing of Romulean deeds throughout the world, how much more would these same deeds impress posterity if he testified to them!’ (si nunc fata darent, ut Romula facta per orbem | hic caneret uates, quanto maiora futuros | facta eadem intrarent hoc … teste nepotes!, 13.793–795). Because Homer cannot fulfil this task, the wish paves the way for Silius to become a ‘Roman Homer’ for Scipio.52 In this way, Silius’ intertextuality creates its own “immanent literary history”: the position that the Punica should have in literary history is shaped by the Punica itself.53
Another example of this ‘do-it-yourself literary tradition’54 is the way Proteus puts the Punica in line with the Iliad and the Aeneid. As we have seen above, Proteus can be seen as a mirror of Silius, and his prophecy of the course of the war as a mise en abyme for the ensuing books of the Punica. Before the prophecy proper, the god traces the origin of the war back to the Judgement of Paris (7.437–471). His account is conspicuously long, especially when compared to the short references of Homer and Virgil to this event, both consisting of merely two lines (Il. 24.29–30 and A.1.26–27).55 Proteus thus highlights a scene which is only mentioned in passing by the greatest epic poets. Next, he reduces both Homer and Virgil to footnotes in his account of the epic past, permitting only four verses for an epitome of the Trojan war and travels of Aeneas (7.472–475).56 Right after this summary, Proteus starts his prophecy of the Punic wars, for which he reserves much more space (7.476–491). This representation simplifies epic literary history to a succession of Iliad, Aeneid, and Punica. Silius’ poem is framed as the logical successor of Homer and Virgil, and at the same time as the culmination of the epic tradition (yet in the making). It should be no coincidence that it is Proteus, a prophet who featured prominently both in Homer (Od. 4.349–570) and Virgil (G. 4.387–529), who constructs this honourable position of the Punica in the literary tradition.
The Punica also alludes widely to post-Virgilian epic. I have already pointed to the importance of Ovid’s Metamorphoses as model for many ‘digressive’ elements in the epic, such as embedded narratives.57 Again, the Proteus episode may serve as an illustration. Proteus, who since Homer is famous for his ability to change shapes, is an Ovidian figure avant la lettre. The changeable god plays only a marginal role in the Metamorphoses, also in the capacity of a seer (Met. 11.221–223; 11.249–256). Nevertheless, he is presented in the Punica as an Ovidian character. When the nymphs, in distress about the arrival of the Carthaginian fleet, come to Proteus’ cave to ask about the future, the god duly shows his tricks, changing into a snake and a lion: ‘he eluded them by taking various shapes’ (per uarias lusit formas, 7.423).58 Littlewood rightly tags lusit as a metapoetical marker;59 in addition, I would suggest that the uarias … formas are a transformation of and an allusion to the mutatas … formas of the opening line of the Metamorphoses. This prepares the narratees for the ‘Ovidian’ narrative of Proteus, which has been viewed as a confrontation of ‘epic’ and ‘elegiac’ elements, especially with regard to the story of the Judgement of Paris.60 Negotiations between epic and elegy are of course an important feature of Ovid’s poetic programme in the Metamorphoses,61 and embedded narratives such as the Proteus episode are likewise the main generic battlegrounds in the Punica.
Another important intertext is Lucan’s Bellum Civile, the only other extant historical epic from the first century AD. The Punica, with a reintroduction of the divine machinery and mythological elements, has often been viewed as a restoration of the Virgilian epic principles that Lucan had expelled from his epic on the civil war. The Punica positions itself as an epic in the middle: the medium bellum from the prooemium (1.12) is not only a reference to the Second Punic War as the middle of the three consecutive wars against Carthage, but also suggests the epic’s intermediary position in between the Aeneid and the Bellum Civile. The Hannibalic war of Silius is as it were a prequel of the civil war between Caesar and Pompey of Lucan, just as it is the sequel of the Aeneid.62 The Punica refers to future events (Roman civil wars), while alluding to a work from the literary past (Lucan’s Bellum Civile).63 That the Punica is foreshadowing the Bellum Civile can, again, be shown with Proteus’ narrative. After the epitomes of the Iliad and Aeneid, Proteus makes a statement about the rule of the Romans:
dum ceteponto innabunt, dumsidera caelolucebunt, dumsol Indo se litore tollet,hic regna et nullae regnis per saecula metae. (7.476–478)
As long as sea-animals shall swim in the sea, as long as stars shall shine in the sky, as long as the sun will rise from the Indian shore, [their] rule will be here and there will be no limits to [their] rule over the centuries.
These lines have unsurprisingly been read as a confirmation of Jupiter’s prophecy in Aeneid 1 that Rome will rule the world without limits.64 At first sight, these lines are indeed meant as a reassurance for the nymphs who have come to their father Proteus in distress about the Carthaginian advance: his message is that, in the end, the descendants of Aeneas will rule the world. A closer look at these lines reveals, however, that this statement is perhaps less reassuring. Proteus is truly an ambiguus uates (7.436). First of all, it is not made absolutely clear, whose regna (note the plural) he means. In the context, a ready assumption is that he means of the Romans, but in fact he only states that in Italy (hic) there will be empire(s) (regna) over the centuries (per saecula). This allows for the possibility that Roman rule will eventually come to an end. And even if regna refers to eternal Roman rule, the plural might be more than just a poetic variation of regnum. Does Proteus make a veiled prophecy of Roman civil wars in the centuries to come with various Romans striving for regnum, or ‘tyranny’? Here Lucan enters the intertextual game. In his first book, the poet states that Rome will continuously be haunted by civil discord (in fact, Rome came to being as a result of civil war, Romulus having killed his brother Remus). Lucan, like Proteus, uses a dum-clause to stress the perpetuity of Roman civil wars:
dum terrafretum terramque leuabitaer et longi uoluentTitana laboresnoxque diem caelo totidem persigna sequetur,nulla fides regni sociis (Luc. 1.89–91)
As long as earth supports the sea and air the earth, as long as Titan revolves in his lengthy toils and in the sky night follows day through all the constellations, there will be no loyalty between associates in tyranny.
Proteus employs similar cosmic images to stress the eternity of the regna: the sea, the sun, and the stars. As long as the cosmos exists, (Roman) rule(s) over Italy will exist. If we accept the Lucanian intertext as relevant for Proteus’ prophecy, the regna might be taken as a reference to the civil wars that will plague Rome in the centuries to come.
This cycle of civil war will, however, come to an end. By means of an analogy Lucan uses the image of an apocalypse that will force ‘so many centuries of the cosmos’ (saecula tot mundi, Luc. 1.73) to collapse into chaos. Therefore, the limitless rule per saecula that Proteus speaks of is limited after all. When we take into account these allusions to the Bellum Civile, the prophecy of Proteus is in line with Jupiter’s prophecy in Punica 3. There, the supreme god, addressing his daughter’s worries about the Carthaginian successes, predicts a long rather than eternal rule for her descendants: ‘your blood is holding the Tarpeian citadel and will hold it for a long time’ (tenet longumque tenebit | Tarpeias arces sanguis tuus, 3.572–573).65
A vexed problem is the relation between the Punica and the other Flavian epics. Valerius Flaccus’ Argonautica is often dated in the Vespasianic era and is therefore regarded as a predecessor of the Punica, but the chronological relationship between Statius’ Thebaid and Silius’ epic is subject to debate.66 Although it is certain that Statius finished his epic project earlier, this does not automatically cast Silius into the role of the alluding poet and Statius in the role of the poet alluded to. In fact, it is well possible that these poets influenced each other, having heard/read parts of the other’s poem while composing their own. Therefore, it is hard, if not impossible, to pin down who alludes to whom.67 In spite of these chronological issues, Flavian intertextual readings of the Punica can yield interesting observations and interpretations.68
So far, I have merely discussed Silius’ intertextual engagement with other epics. Their prominence in my own and other intertextual readings of the Punica is understandable enough: they share both a comparable generic outlook (including generic crossovers) and a similar language (including vocabulary and metre). As we have seen above, Silius positions himself in this epic tradition. This focus on epic intertexts runs the risk of developing blind spots for intertextual connections with other genres, be it poetry or prose. Silius, in fact, uses a plethora of intertexts, as I hope to show in my case studies; I do not pretend, however, to be all-encompassing. It is impossible to detect, describe, and interpret all possible allusions in a highly intertextual epic like the Punica: “The critic, like the poet, can only bring finite resources to the infinity of discourse.”69
A name largely absent in this book is Livy. The influence of the third decade of the Ab Urbe Condita on the Punica can hardly be underestimated, but the embedded narratives in Silius’ epic deal with stories that mostly cannot be found in Livy’s work. An exception is Regulus’ encounter with the giant snake, but Livy’s account is unfortunately only transmitted in excerpted form (Liv. Per. 18).
The praxis of a modern intertextualist includes a digital component, such as the use of Tesserae, Diogenes, and the digital collection of the Packard Humanities Institute for checking or hunting down allusions. These digital tools have made it possible to track references that would otherwise perhaps have escaped notice and make ‘distant’ readings of texts possible that can hardly be performed by a human.70 It makes one wonder how ‘natural’ this digital reading is, especially compared to an average reader from Silius’ own time. But perhaps we should not worry to much about this artificiality: in the end, it is up to the reader, whether using digital tools or not, to accept a certain intertext and to interpret it.
It is also good to be aware of blind spots of digital intertextual methods.71 A potential danger is a strong focus on verbal allusions, because, for example, thematic similarities are not easy to trace digitally. Intertextual studies of Roman literature have traditionally concentrated on verbal markers of intertextuality, and my study is no exception to that rule. Nevertheless, I hope to show that the allusive practice of Silius also manifests itself on the level of sound, metre, word order, narrative techniques, and themes. When it comes to terminology, I have allowed myself to freely use ‘allusion’, ‘reference’, ‘echo’, ‘influence’ when describing intertextual links. It goes without saying that I do not think we can know the (intertextual) intentions of the author, but I do think that the alluding poet is a “good tool to think with”.72 I hope that the reader can forgive my lack of philological fundamentalism in this matter.
Allusions in Silius are frequently signposted: markers in the text alert the reader that a reference to another text is made. One can think of ‘Alexandrian footnotes’, such as fertur or dicunt, words that denote repetition (iterum, repeto, re-, rursus), or old and new (antiquus, nouus). As they draw attention to the process of poetical creation, I usually refer to them as (metapoetical) markers or signposts of intertextuality.73
5.3 Intratextuality
The third critical tool that I use is intratextuality. Although this term is not as widely used as its far commoner pendant intertextuality, I think of it as a useful means of looking at texts, especially for my purpose of relating embedded narratives to the epic as a whole.74 Intratextuality looks at chunks of texts from the same text and the relations between them. As Sharrock puts it: “Reading intratextually means looking at the text from different directions (backwards as well as forwards), chopping it up in various ways, building it up again, contracting and expanding its boundaries both within the opus and outside it.”75 The last two words betray the affinity with intertextuality. Tellingly, Fowler, in an article in the same volume, easily crosses boundaries between the two in his discussion of the Virgilian episode of Nisus and Euryalus: “With this we move on from intratextuality to intertextuality, but the two phenomena are of course constantly involved with each other.”76 My contention is that intratextuality in the Punica roughly works in the same way as intertextuality does. It is often through verbal allusions that a link is forged between two passages from the same work, inviting a comparative reading. As is the case with intertextuality, it is up to the reader/narratee to decide what to do with the allusion: do the passages confirm each other, or does the allusion rather highlight a contrast? And these are of course only two interpretative paths that one can take.
Intratextuality is for me not a tool to uncover a supposed unity of the text. Just as intertextuality, as developed by Barthes and Kristeva, allows readers to freely associate, intratextuality is an act on the part of the readers. They should themselves decide whether a bit of text is a subdivision in the first place and whether there is a relation with other bits, the whole or not at all. In the words of Sharrock: “Intratextuality is about how bits need to be read in the light of other bits, but it is also about the bittiness of literature, its uncomfortable squareness-in-round-(w)holeness.”77 There is no need to see intratextuality as a totalizing quest for unity and coherence—nor does it exclude the possibility of reading unity and coherence. Of course, there is a big difference between intertextuality and intratextuality. The former offers a scope of possible associations that is in theory infinite, whereas the latter is restricted by what is regarded as the whole—in my case, the Punica.78
To illustrate intratextuality in the Punica, let us for the last time go back to the Proteus episode. A strong and well-recognized example of verbal intratextuality is the diminutive paruulus, which describes the size of Cupid’s quiver: ‘a tiny quiver and a golden bow glittered at his shoulder’ (paruulus ex humero corytos et aureus arcus | fulgebat, 7.443–444). This elegiac diminutive adds to the un-epic atmosphere of the arrival of Venus on Mount Ida and Judgement of Paris at large. The attentive narratee will signal its repetition in Venus’ speech to her Cupids in Book 11, where, again, the word is applied to their panoply: paruula nos arcu puerili spicula sensim | fundimus (‘we but gently launch tiny arrows from boyish bows’, 11.393–394).79 The strongest marker of intratextuality is of course the repetition of paruulus, at the beginning of the verse. In addition, paruulus is a repetition of a rare word, as these are the only two attestations in the Punica. But there are more echoes from Book 7 that support the intratextual reference: note the repetition of arcus ~ arcu, the assonance of the last two words
For that matter, paruulus is a marker of an intertextual allusion, too. It is an echo of the famous Virgilian hapax in Aeneid 4.328, where Dido muses on having a baby with Aeneas, a ‘tiny Aeneas’ (paruulus … Aeneas), who could have been a comfort (and substitute) for his father’s absence.80 Her daydreaming runs counter to the epic telos of Aeneas, as decided by Fate, and therefore her wish does not come true. The stay at Carthage as a whole is an ‘elegiac’ mora of the main narrative: the hero has to abandon Dido and return to his mission of reaching Italy. Paruulus stands for a world that is the opposite of this epic quest. These elegiac overtones are also present in Proteus’ and Venus’ descriptions of Cupid’s armoury. But ironically, the weapons of Venus and her sons do affect the epic narrative—more than that, epic depends on them: Proteus goes back all the way to Venus’ enchantment of Paris to explain the origins of the wars that are fought in the Iliad, the Aeneid, and the Punica.81 In Book 11, we can see that these ‘elegiac’ powers of Venus are still significant for the course of war: the goddess orders her Cupids to weaken the hearts of the Carthaginians and they duly obey to her words. After their stay in Capua, Hannibal and his men are no longer capable of gaining victories over the Romans.82
6 Scope of This Study
In the following chapters I will apply the combined method of narratology, intertextuality, and intratextuality to four embedded narratives from the Punica. The interpretation of narratological mechanisms, allusions, and intratextual references takes up quite some space. Of course, it would have been possible to deal with all of Silius’ embedded narratives, but that would have meant that I could only touch upon them superficially. Instead of being exhaustive in the number of embedded narratives that I discuss, I have therefore chosen to treat four of them in great depth.
The four case studies that I have selected are representative of the phenomenon of embedded narratives in the Punica. They show a variety both in function and narrative levels. Chapter 1 deals with Bostar’s report to Hannibal of his visit to Jupiter Hammon’s oracle and is an example of a flash-back with a flash-forward embedded (3.647–714). The long narrative of Marus on the Roman hero Regulus (6.62–551) is the subject of Chapter 2. This narrative is a flash-back or analepsis of events from the First Punic War; its historical subject-matter is very much in line with Silius’ own main narrative on the Second Punic War (although it includes, like the rest of the Punica, elements that are less historical). Marus’ narrative, containing speeches on the tertiary level of Regulus and his wife Marcia, alternates with interruptions of his direct narratee Serranus and the primary narrator. Chapter 3 discusses the story of Falernus (7.162–211), an example of a narrative with a strong aetiological nature, explaining the name and origin of Falernian wine.83 This otherwise unknown farmer is visited by the god Bacchus. Like the story of Regulus, it treats events from the past, but now from a rather mythological perspective. Besides a flash-back to a mythological past, it also contains a prolepsis of events to follow in the ensuing books of the epic. This story is told by the primary narrator, but is clearly demarcated as an embedded narrative. The last case study (Chapter 4) deals with the story of Anna Perenna, Dido’s sister (8.44–201). This analeptic story also deals with the mythological past and is at the same time a continuation of the Aeneid, like the Punica as a whole. It therefore forms an important link between both epics. The story is told by the primary narrator, but (unlike the narrative of Falernus) includes secondary narratives by Anna (8.81–103 and 8.116–159).
Another criterium for selection was length; I have included the three longest ones: Bostar (68 lines), Regulus (489 lines), and Anna Perenna (158 lines). All the other embedded narratives in the Punica cover less than 60 lines. Of these smaller ones, I have selected the story of Falernus, because with its 49 lines it is the longest of the subgroup of aetiologies. This does of course not mean that the narratives that have fallen out of my selection are unsuitable for my approach.84 In fact, I hope to show that my approach can yield interesting interpretations of all embedded narratives in the Punica, but for the time being I leave them for future research.
7 Relevance of This Study
In these times of ecological and climatological crises, geopolitical unrest, and existential uncertainty, one might ask why we should continue to study classical texts and Silius’ Punica in particular. Is there any relevance in reading them? And: does this text-oriented study in any way contribute to debates on issues in the real world? With some hesitation and at risk of sounding too arrogant, my answer to both questions is yes.
First of all, Silius and his narratees were people of flesh and blood. They, too, lived in uncertain times, full of political unrest, war, and natural catastrophes—think of the eruption of Vesuvius in AD 79. Although the Punica relates events from about 300 years earlier, they would have read (or rather heard) that past against their own times. It is not too daring to assume that they could perceive the characters and events in the epic as reflections of contemporary events, such as the civil wars in AD 68–69.
In turn, classical texts like the Punica can act as mirrors for us as readers in the twenty-first century. They can help to reflect on modern situations like (always recurring) wars and questions of all times regarding leadership, grief, and generational trauma, to name a few. The fact that these texts were created and situated in a distant past, creates a gap. This gap makes them sometimes difficult to understand, but also creates a certain distance that render them useful as starting points of critical thinking and discussion of perennial or modern issues.
The present study inevitably touches upon issues that are relevant for the modern reader. Inevitably, because these issues are simply there in the text of the Punica. They are, however, not the focus of this study, which is very much text-oriented. Nevertheless, this textual approach is in a general sense also relevant for a wider public than classicists alone. It arose from general questions such as how a text is put together and how we should read it. I hope to demonstrate that texts like the Punica are not so easy to interpret and often turn out to be ambiguous—and that this is exactly what makes studying them worthwhile. In a world where mankind is increasingly leaving reading and writing to AI, it is important to show that slow reading of a complicated text from a real person can be rewarding.85
This rather derogatory evaluation “puts Silius on a par with Callimachus, whom Ovid described as ‘strong in skill, if not in talent’ (Am. 1.15.14: quamvis ingenio non valet, arte valet)” (Pomeroy 1989: 139 n.78).
Scaliger on Silius: quem equidem postremum bonorum poetarum existimo; quin ne poetam quidem (De Arte Poetica, p.324, as cited by Conte 1994: 492).
It goes without saying that current scholarship on the Punica is greatly indebted to earlier studies, such as the trail-blazing monograph of Von Albrecht 1964 and the sometimes undervalued commentaries on the entire epic of Spaltenstein 1986; 1990. Dominik 2010 gives a good overview of modern Silian studies, as does Augoustakis 2014a. See also Ariemma 2000a and Schaffenrath 2010a: 9–10. A Forschungsbericht of the last decade is a desideratum. Recent monographs that concentrate solely on the Punica are Stocks 2014 on the portrayal of Hannibal, Haselmann 2018 on the motif of rivers and other water bodies, and Jacobs 2020, providing an introduction to Silius Italicus and his epic. Commentaries on single books or smaller units from the last decade include Bernstein 2017 on Book 2, Augoustakis and Littlewood 2022 on Book 3, Schedel 2022 on Book 4.1-1-479, Littlewood 2011 on Book 7, Lee 2017 on Book 8.1–241, Zaia 2016 and Bernstein 2022 on Book 9, Littlewood 2017 on Book 10, Telg genannt Kortmann 2018 on Book 12.507–752, Van der Keur 2015 on Book 13, and Roumpou 2019 on Book 17.341–564; more commentaries are expected to be published in the coming years. Augoustakis and Bernstein 2021 is a fresh English translation of the Punica.
With ‘main narrative’ I mean the events told by the primary narrator, except for external analepses and prolepses. This definition is based upon the term ‘main story’ in De Jong, Nünlist, and Bowie 2004: XVI.
“If they resembled each other completely, we would have identical texts”: Bal [1985] 20093: 60.
I borrow this term from Zissos 1999: 300, who applied it to Valerius Flaccus’ Argonautica. Wilson 2004: 248 aptly calls the Punica “the most intertextual of poems”. Modern scholarship does not question the intertextual richness of the Punica (see e.g. contributions in Coffee et al. 2020). This massive intertextuality has contributed to no small degree to reservations about Silius’ poetical abilities, cf. e.g. Haüßler 1978: 162 n.30 in a critique of Bassett 1955 and 1959: “so dünkt uns das mit minutiöser Geduld versammelte Material fast allzu üppig, das Netz der von Vers zu Vers sich rankenden subtilen Beziehungen und Anklänge allzu fein gesponnen: man staunt, wie Silius—gleichsam poeta doctissimus und universales Gedächtnisgenie—vor so viel Bäumen den Wald noch sehen und ein 17 Bücher langes Epos auf seine alten Tage hin zustandebringen konnte.”
Wilson 2004: 237. Conte 1994: 495 argues that the digressive nature of the Punica “seems to produce strong centrifugal thrusts within a structure whose unity shows itself to be ever more formal and less substantial.” For digressions as part of a ‘centrifugal’ poetics (as opposed to ‘centripetal’ poetics), see Heath 1989: 5 and 11 n.18. He concedes that Aristotle’s theory of unity (“single and complete”) implies that ornamental digressions can be part of a unity when admirers of this type of poetry expect it to be part of it (Heath 1989: 9 n.16). However, Heath dismisses ‘allegorical’ interpretations that explore thematic affiliations between digressions and the work as a whole. An interesting article that discusses Aristotelean ideas of unity is Friedrich 1983. For epic poetry he contends: “In epic poetry, the alternative is not unity or diversity; rather, the poet’s aim is unity in diversity.” (1983: 46).
Schaffenrath 2010b and Walter 2014. Schaffenrath discusses embedded narratives told by secondary narrators who take over the role of the primary narrator; he exemplifies his ideas with a discussion of Marus’ narrative on Regulus. Walter’s study, encompassing all three Flavian epicists, investigates both primary and secondary narrators and the interaction between the different narrative levels. With regard to the Punica, it covers the embedded narratives of Bostar, Proteus, Anna, and Teuthras. Another related study is Fucecchi 2008 on ‘epyllia’, discussing the Regulus and Falernus episodes. Recent articles on single embedded narratives include Stürner 2015 on Bostar, Vinchesi 2011 and Walter 2018 on Regulus, and Fernandelli 2009 and Chiu 2011 on Anna Perenna. Older scholarship has also shown a relatively large interest in this phenomenon. I mention here the seminal studies of Bassett 1955 on Regulus, Vessey 1973 on Falernus, and Santini 1991: 5–62 on Anna Perenna as notable examples.
This definition is taken almost literally from De Jong, Nünlist, and Bowie 2004: XV.
E.g. Bal [1983] 20093: 56–57.
Nelles 1997: 132–133.
This list is taken from De Jong 2014: 34–35. Cf. also De Jong 2004: 10.
Fowler 2000a: 299.
A good example is Legras 1905: 152, who deems the digressions in Statius’ Thebaid unnecessary and disruptive for the unity of the epic.
Steele 1922: 325.
Ribbeck 1892: 200–202.
Stärk 1993: 142.
Stärk 1993: 143.
Stärk 1993: 139. A similar view is taken by Perutelli 1997, who argues that the Proteus narrative is merely ornamental.
Masters 1992: 5–6. Delay is also structural to Virgil’s Aeneid, where the description of the war in Latium only starts in Book 7. On delay in imperial epic, see Fowler 2000a: 299–301.
For reflections on the Makrostruktur of the epic, see Kißel 1979: 211–217, T. Gärtner 2010, and Stürner 2011.
Zinn in Von Albrecht 1964: 133 n.35. Pace Kißel 1979: 217 n.18. The decisive Battle of Zama took place in 202 BC, although peace was only made in 201 BC. I consider the Punica as a ‘finished’ work, with Book 17 as its final book. The number 17 is, however, very unusual for an epic. For a convenient overview of the discussion on Makrostruktur and the number of books, see Schubert 2010: 22–23 with n.39 and also Augoustakis 2010a: 9–10. Wenskus 2010 relates the number of 17 to its importance in ancient medical thinking.
Kißel 1979: 213.
Fucecchi 2008: 41.
Ribbeck 1892: 193. For a metapoetic reading of Fabius and his delay of the war, see Tipping 2010: 131–137.
The landscape of Campania, too, helps to delay and obstruct Hannibal according to Biggs 2019: 211–212.
On the Falernus narrative, see Chapter 3. The term ‘essential epic’ is a coinage of Hinds 2000: 223. On this idea, see also Heerink 2015: 18–19.
For the generic variety in the Proteus narrative, see Perutelli 1997. A similar mechanism is found in Book 14, the self-contained episode of Marcellus’ campaign in Sicily. Marks 2017 argues that non-epic digressions in that book delay the epic narrative. In the case of Sicily, the delaying tactics of the narrator turn against the invasion of the Romans.
This is quite similar to the Nemean episode, the longest mora in Statius’ Thebaid. For all its Callimachean associations, this is the place and moment where the war narrative takes off. Cf. Soerink 2014: 47–56 (providing a bibliography on mora in Statius in n.228).
Fowler 2000b: 90.
The most extensive study of the phenomenon is Dällenbach 1977 (English translation Dällenbach 1989). He quotes Gide’s original passage (Dällenbach 1989: 7) and then gives the following definition: “ ‘mise en abyme’ is any aspect enclosed within a work that shows a similarity with the work that contains it” (Dällenbach 1989: 8). Heerink 2015: 29 n.30 gives a convenient overview of this metapoetic technique, with references to classical scholarship. Bal [1985] 20093: 62 suggests to use the term mirror-text instead of mise en abyme. De Jong 2014: 36–37 understands the term mirror-story as “[a]n embedded narrative that reflects the main narrative, as flash-back/flash-forward or thematically”, and considers it as a subtype of mise en abyme. Fowler 2000a: 301 points to the difficulties of relating the mise en abyme to the main narrative: “Often it will not be clear exactly how we are to relate the story to the rest of the work, and the interpretation of the stories may frequently be thematized into a more general hermeneutical problematic.”
Fowler 2000a: 29. I will therefore use the term mise en abyme not only for embedded narratives themselves, but also for secondary narrators that represent the primary narrator. Besides Fowler, Heerink uses the term in this way, for example when he discusses the secondary narrator Orpheus in the Argonautica as a mise en abyme of Valerius Flaccus (Heerink 2013: 274–276). Dällenbach, too, seems to allow for this ‘extended’ use of mise en abyme, when he includes “the story being told and the agent of the narration” (Dällenbach 1989: 43) as a possible object of reflexivity. Cf. also Deremetz 1995: 434–435 and 443–445.
The locus classicus for the poet as uates is Hor. Ars 391–401 (with Brink 1971: 391). This does not mean of course that every uates in Latin epic is a mise en abyme of the primary narrator / poet. See e.g. Lovatt 2007 for a nuanced stance on the role of uates in Statius. Walter 2014 pays quite some attention to the role of prophets as secondary narrators in Flavian epic and their relation to the primary narrator.
Schaffenrath 2010b: 119.
Deremetz 1995: 411–474, esp. 434–467.
Deremetz 1995: 446–447.
Deremetz 1995: 446.
Reserat is another wink to the prooemium. The same verb is found there of Scipio Africanus conquering Carthage: reserauit Dardanus arces | ductor Agenoreas (‘the Dardanian general opened the Agenorean citadel’, 1.14–15). The victory of Scipio is one of the future events that Proteus is about to disclose. Littlewood 2011: 174 and Walter 2014: 314–316 rather highlight the markers of intertextuality in this passage (e.g. repetita and retro), which signal that Proteus will engage with the literary tradition. Proteus’ prooemium evokes other epic prooemia. There is an echo of Ennius’ prooemium to Book 7: ‘we have dared to open ⟨the sources⟩’ (nos ausi reserare ⟨fontes⟩, Skutsch fr. 210); for this phrase and its reception in Virgil and Statius, see Suerbaum 1968: 275–280. It also interacts with the prooemium of Statius’ Thebaid, as Ruurd Nauta has suggested to me (cf. euoluere, primordia, and retro in Theb. 1.2–7). For retro and re
See Walter 2014: 307–308, who argues that the arrival of the Carthaginians at Caieta “für jeden Kenner Vergils ein Alarmsignal darstellen muss.”
To my knowledge the combination of these three theories has not been applied to the Punica (at least not explicitly). Walter 2014 comes closest; although she does incorporate inter- and intratextual readings, she gives precedence to a narratological methodology.
Genette [1972] 1980; Genette [1983] 1988. An overview of theoretical thinking on narrative levels is given by Pier 2014 in the online Living Handbook of Narratology.
Schaffenrath 2010b: 116 calculates that approximately 30 percent of the Punica consists of speeches on this secondary level. Tertiary narrators are most often encountered in embedded narratives (Schaffenrath 2010b: 117 n.24).
De Jong 2004: 1–4 and 2014: 19–20. Genette’s corresponding terms are homodiegetic and heterodiegetic. See also Schmid [2005] 2010: 67–70 for a simplification of Genette’s terminology.
Starting with De Jong 1987.
A convenient glossary of narratological terminology can be found in De Jong et al. 2004: xv–xviii, while examples from Greek and Latin literature are found in De Jong 2014.
For intertextuality and Latin poetry, see Coffee 2013; for intertextuality and Flavian epic, see Coffee et al. 2020.
For an introduction to intertextuality in Silius, see Von Albrecht 1999: ch. 12, who deems it “the guiding principle of invention” of the Punica.
Cf. e.g. Bernstein 2018: 249: “Intertextual engagement with Virgil’s Aeneid is one of the dominant compositional characteristics of Roman epic of the first century CE.” See also Hardie 1993 and Hinds 1998.
Barchiesi 2001a: 129. It must be added that ordior is also suggestive of Roman historiographical tradition; cf. the use of this verb in Liv. praef. 12 f. with Feeney 1982: 6–7. Saliently, Livy himself starts his Ab urbe condita with a hexameter and treats ‘poetical’ subject matter in his first decade.
On Silius Italicus and the tradition of Roman historical epic, see now Augoustakis and Fucecchi 2022.
Ennius, in turn, presents himself as the reincarnation of Homer. For this ‘genealogy’ of Homer, Ennius, and Silius, see Deremetz 1995: 470–474. For Ennius and Silius, see esp. Marks 2010a. Suerbaum 1968: 281 discusses Ennius’ use of the ‘primus motif’ in the prooemium to Annales 7.
Cf. e.g. Bernstein 2018: 248: “Silius accordingly creates an [sic] space for himself to be a new Homer, this time as the poet of Scipio’s deeds.” For the encounter of Scipio and Homer, see especially Van der Keur 2014.
For the idea of “immanent literary history”, see Schmidt et al. 2001. For allusion as a way of shaping literary history, see Hinds 1998 and Deremetz 2001: 147: “la tradition n’existe que mobilisée par des textes; chaque texte, en choisissant ses modèles, en respectant certaines règles de production, institue et désigne lui-même la tradition vive à laquelle il prétend se rattacher.”
I borrow this phrase from Hinds 1998: 123.
See Walter 2014: 309. The two lines from the Iliad are athetized by some.
Lines 7.472–473 highlight the arrival of the Greek army and the fall of Troy. These events, marking the start and end of the Trojan war, are properly part of the Epic Cycle. The story of the Iliad is glossed over and only implied by mentioning the two limits of the war. Of course, Homer’s epic contains references to the beginning of the war (e.g. the catalogue of Iliad 2) and foreshadows the fall of the city (e.g. Zeus conceding Troy to Hera in Il. 4.25–49, and Agamemnon stating that Troy surely will fall in Il. 4.163–168).
On Silius and Ovid, see especially Wilson 2004. Foundational studies that have explored Ovidian elements in the Punica are Bruère 1958 and 1959.
This is a nod to and an abbreviation of the literary tradition, only performing the first two of in total six metamorphoses listed in Od. 4.456–458. At the same time, this return to Homer is a ‘correction’ of Virgil, whose Proteus changes into fire, beast, and running water (G. 4.441–442). Pace Juhnke 1972: 393 and Perutelli 1997, who deny a direct influence of Homer in this instance. See Thomas 1988b: 219–220 for the Virgilian adaptation of the Homeric tradition in this passage. Littlewood 2011: 172 notes: “Silius’ friendly seer plays along, lusit (423) with a few showy transformations, hissing like a snake and roaring like a lion.” But lion and snake are two animals that are strongly associated with Hannibal in the Punica; this makes Proteus at first sight much scarier for the already alarmed nymphs (and also for the primary narratees) than Littlewood allows.
Littlewood 2011: 171–172. Pace Spaltenstein 1986: 474, who contends that “chez Sil., ces transformations n’ont pas de raison d’être, ce qui montre aussi le caractère convenu de ce passage.”
Walter 2014: 309: “Wie kaum ein anderes eignet sich dieses Thema für eine Erörterung der Frage, welcher Gegenstand epischem Erzählen angemessen ist und wo die Grenze zwischen Elegie und Epos verläuft.”
For Ovid and genre, see e.g. Harrison 2002.
“In singing of Roman history, Silius self-consciously constructs a continuation of the mythological time of Virgil’s Aeneid and Ovid’s Metamorphoses and the historical time of Lucan’s Bellum Civile.” (Bernstein 2018: 263). See also Ahl, Davis, and Pomeroy 1986: 2501–2502. For the relation between Silius and Lucan, see e.g. Meyer 1924; Brouwers 1982; Marks 2010b.
Barchiesi 2001a: 105–127 calls this type of allusion a ‘future reflexive’.
Compare for example the note of Littlewood 2011: 183: “The word regna conveys power and dominion, while meta and saecula signal allusions to Jupiter’s pronouncement on the eternity of Rome (Virg. Aen. 1. 278–279 his ego nec metas rerum nec tempora pono; | imperium sine fine dedi). The continuation of the Virgilian speech touches Silius’ epic closely.” See also Walter 2014: 316.
Contra Spaltenstein 1986: 478: “Le vers 478 contredit le vers 3,572 (où l’empire romain est dit limité dans le temps).”
For Silius and Valerius Flaccus, see e.g. Ripoll 1999 and Augoustakis 2014c; for Silius and Statius, see e.g. Ripoll 2015.
Agri 2020: 1–3 gives a convenient overview of the chronological debate. She herself only allows for mutual influence in Books 1–12 of the Punica, on the assumption that both Statius and Silius adopted a linear style of composition and completed approximately one book a year. For a similar ‘work in progress’ scenario in the Alexandrian Mouseion, see e.g. Heerink 2015: 15.
Examples are Lorenz 1968, Lovatt 2010, Heerink 2013, Soerink 2013, and Agri 2020. There is also interaction between the Punica and Statius’ Achilleid. Ripoll 2000a: 105 and 2015: 440–441 gives examples where Proteus’ account of the Judgement of Paris influenced Statius’ Achilleid. Walter 2014: 310–312 on the other hand contends that Silius alludes to Statius.
Hinds 1998: 52.
See e.g. Bernstein 2020 and Coffee 2020. An interesting comparison between Tesserae and philological commentaries on Lucan’s Bellum Civile 1 is given in Coffee et al. 2012.
Hinds 2020 provides some caveats of digital intertextuality, especially the trouble of detecting negotiations between Greek and Roman epic traditions. On Greek literary traditions in Flavian poetry, see Augoustakis 2014b.
Hinds 1998: 50, in his apology of the term ‘allusion’. See Edmunds 2001: 164–169 for a critique of Hinds’ alleged “nostalgia for the presence of the author”.
For such metapoetical signposts, see Wills 1996: 30–31, who calls them ‘external markers of allusion’, and Hinds 1998: 1–5, who discusses the ‘Alexandrian footnote’ together with other forms of ‘reflexive annotation’. For re- see Hardie 2013: 115 and Casali 2017: 100 on A. 2.3 (renouare); for rursus, see Barchiesi 2001a: 139–140 and Heerink 2015: 7–8 on V. Fl. 3.596–597.
An example of an intratextual approach of an embedded narrative in Statius’ Thebaid is Van den Broek 2016.
Sharrock and Morales 2000: 5. Their collected volume is (still) the most important contribution to the theorizing and application of intratextuality. A recent volume that collects papers on intratextuality in Latin literature is Harrison, Frangoulidis, and Papanghelis 2018.
Fowler 2000b: 91.
Sharrock and Morales 2000: 7.
This paragraph owes much to a discussion with Piet Gerbrandy about the (potential) implications and presuppositions of both intertextuality and intratextuality.
The intratextual echo has not escaped modern readers of the Punica: Spaltenstein 1986: 475, Barchiesi 2001b: 339 with n.37, and Walter 2014: 318–319. Littlewood 2011: 176 signals that the “neoteric atmosphere is heightened by the sentimentality of the diminutive paruulus”, but seems to have missed the intratextual significance.
Spaltenstein 1986: 475 and Barchiesi 2001b: 339 note the parallel.
Barchiesi 2001b: 339 n.37.
See Introduction, section 2 above. For the influence of Venus on the main narrative, see also Walter 2014: 318–319.
See especially Chapter 3, section 1 and 3. Two other (but shorter) examples of similar invented aetiologies on geographical names told by the primary narrator are the story of Pyrene (3.415–441), explaining the name of the Pyrenees, and the story of Thrasymennus (5.3–23), explaining the name of Lake Trasimene. In fact, all embedded narratives in the Punica contain one or more elements of aetiology. Bostar explains the name and workings of the mysterious Spring of the Sun (see Chapter 1, section 3.1 and 4.2), Marus introduces his narrative of Regulus as an explanation of the origin of his spear (see Chapter 2, section 4), and the narrative of Anna Perenna gives answer to the question why a Carthaginian goddess is worshipped by Romans (see Chapter 4, section 1). Besides embedded narratives, the Punica is brimmed with shorter aetiological explanations. Up to now there is no specific study of aetiology in Silius that I know of. On aetiology in Greek and Roman literature, see Walter 2019 and 2020.
The embedded narratives that do not receive an in-depth analysis in this study are the two aetiological narratives by the primary narrator on Pyrene (3.415–441) and Thrasymennus (5.3–23), Cilnius’ narrative on the Fabii (7.34–68), the prophecy of Proteus (7.409–493; discussed in this Introduction), Cinna’s narrative on Cloelia (10.478–502), Teuthras’ songs in Capua (11.288–302 and 11.440–480), Virrius’ account on the origin of Apollo’s temple at Cumae (12.88–103), and Dasius’ story of the Palladium (13.36–81).
I thank Piet Gerbrandy for encouraging me to address the wider relevance of this study.