1 The Intertextual Aeneid

Scholarship on Latin intertextuality has long been at a standstill, and it has found it hard to extricate itself from this stagnating mire. We may ascribe this outcome to what remain to this day two of the undeniably most influential works on the topic: Gian Biagio Conte’s The Rhetoric of Imitation, translated into English in 1986 as an abridged collation of two earlier Italian studies, and Stephen Hinds’ Allusion and Intertext, published in 1998 as part of a conscious effort to reflect on critical methodologies in the Cambridge UP series ‘Roman Literature and its Contexts’, the fate of which has recently received by Erik Gunderson a witting ‘comi-tragic retelling.’¹ Both books, still widely used as starting tools on undergraduate and graduate syllabi, took Virgil’s inherently intertextual and self-referential Aeneid as a seminal text to reflect on issues of intertextuality and allusion, metapoetics, genre and imitation, the role of author and reader. And how could they not, we may ask, with Juno’s Aeneid, when the Aeneid is first and foremost a rewriting of Homer, and an ‘ethical’ rewriting at that?

Conte and Hinds both feature heavily as the backbone of Joseph Farrell’s reading of the Aeneid in the light of the rival textual and ethical models of Iliad and Odyssey—which is ultimately what Juno’s Aeneid boils down to. Although Farrell has his own scholarly disagreements with (the elder) Conte,² who came to deprecate his early enthusiasm with literary theory as little more than a ‘juvenile affliction,’ whose unintended outcome was to set off flocks of ‘polysemy-seekers’ free to distort the Aeneid in defiance of what we should recognise as the ‘duties’ and responsibilities of the (textual) critic,³ it would be hard to deny the profound influence that Conte’s earlier thought still exercises on Farrell’s positing of a ‘dynamic’ intertextual model that some may find more static and prohibitive than it declares itself to be.

Drawing on his original chapter on ‘intertextuality’ in the Cambridge Companion to Virgil,⁴ now substituted in anticipation of this very monograph with a study of ‘intertextual [authorial] personae,’⁵ Farrell posits that the Virgilian intertext is ‘dynamic’ insofar as it is “not a skeleton key that unlocks the secret meaning of a poem, but rather a device that discloses vistas of interpretative possibility that would otherwise remain unglimpsed and inaccessible.”⁶ He credits Alessandro Barchiesi (2015, originally 1984), Stephen Hinds (1998), and Edan Dekel (2012) for allowing a ‘dynamism’ and ‘multivocality’ to those Homeric intertexts that appeared as little more than ‘dormant’ loci paralleli in Georg Nicolaus Knauer’s relatively static intertextual approach, based according to Farrell on the notion of a ‘systematic intertext’ (1964).⁷ Yet the shift of attention towards the process of textual allusion against the static dimension of conceiving of literary genres in terms of ‘horizons of expectations’ was undoubtedly one of the main contentions of Conte’s Rhetoric.⁸ Despite this emphasis on ‘process’, however, Conte explicitly kept his texts tied to the static metaphor of a ‘grid’, positing the task of the reader or literary critic as that of a ‘deciphering’ of texts, the rules of whose game had been ‘established’ and ‘regulated’ by the Author all along: even for the early Conte, literary texts were always already prescriptive and predetermined entities, whose full potential was tightly dependent on, and could only be unlocked by, the supposed competences of their knowledgeable readers.⁹

Hinds’ Allusion and Intertext certainly improved on this, in making better use of the lessons of deconstruction by pointing intertextual readers towards a more ‘dynamic’, almost networking model of texts. Yet, as Victoria Rimell has recently argued, while there is a vivid, experimental thrust in Hinds that makes him “a voice of intellectually brilliant, third-way moderation in Classics,”¹⁰ Allusion and Intertext is also “fraught by philosophical contradictions.”¹¹ In particular, Hinds never lets go of conceiving of “tradition and creative evolution as a geometric, linear, Oedipal battleground,”¹² openly prescribing us to recoil from what he called “intertextualist fundamentalism—which privileges readerly reception so single-mindedly as to wish the alluding author out of existence altogether.”¹³ As remarked by Rimell, Hinds’ substitution of Conte’s ‘grid’ with the metaphor of the ‘billiard table’ for the intertextual event, where the movement of one ball ends up reconstituting the entire array on the board, is still a model “of solid entities clashing on a two-dimensional surface, observed from above … the components of which remain entirely unchanged in their form.”¹⁴ The results of this preponderant conception of intertextuality in Latin literary studies, which offers limited dynamism on a small scale, posits the necessity of at least one text in the network to be a stable entity so as to ‘enlighten’ the other(s), projects a ‘model’ reader—ultimately coincidental with the literary critic/philologist—as ‘intended’ by the text or author, and cannot finally disentangle these literary works from a Bloomian framework of emulation and competition, are all eminently on display in Farrell’s ‘new’ take on Virgil’s Aeneid.¹⁵

How far have Latin intertextual studies moved on from Conte and Hinds? And what are the ethical, existential, and political stakes for us as literary critics on what intertextuality is and can do in Latin literature? In the passage quoted above, Farrell substitutes a visual image of multiple ‘vistas’ for the more traditional auditory metaphor of those ‘further’ voices of the Aeneid that Alfonso Traina had already decried as having become ‘too many.’¹⁶ Here and elsewhere, moving between what he dubs the ‘dynamic’ and the ‘dialogic’ intertext, Farrell flirts with the option of ‘multivocality’: he grants that the Aeneid, “the ultimate meaning of which continues to prove so elusive,” may not present “a discursive universe of monologic character, but one in which every voice … was just one voice within a dialogic drama,” and does not shrink from asking the question of what our interpretative stance may look like if we were to posit that ‘dialogue’ and ‘polyphony’ ultimately mean that no single voice can in fact silence the others.¹⁷ And yet the message that this book seems to keep repeating is that we should shut the door to the liberating possibility of simply not having to choose, from either an ethical or an intertextual perspective, both of which are for Farrell intimately bound.

Farrell is right in pointing out that, ultimately, the kind of multivocality and polyphony that critics such as Alessandro Barchiesi have espoused can look like no more than another interpretative choice, coming with its own prescriptive interpretative baggage. When Barchiesi writes that “the Virgilian text … accepts the risk of being multivocal and even exploits, for precise communicative ends, the polyphony of its own cultural codes,”¹⁸ there is no denying that he is still ventriloquizing the text, if not the author, conjuring authorial intention at the expense of reader response criticism. After all, multivocality as an interpretative choice might not differ fundamentally in methodology from purely monovocal views; if that is the case, Farrell seems to ask, then why not take a clear interpretative stance, recognising plurality within monologism, in a model according to which the poem may diffract voices (and genres, and intertexts) just as “an evanescent possibility,” even though it “ultimately speaks with a single voice”?¹⁹

It is, as I see it, this move that allows Farrell to double down on what he thinks the Aeneid is ultimately about. Juno’s Aeneid is, in short, a poem that “presents its readers with a dichotomous perspective on the world.”²⁰ Just as Farrell’s Juno, at the beginning of the Aeneid, chooses to control a narrative that was verging towards Odyssean ends, attempting to twist it towards the world of the Iliad,²¹ so the poem that we as readers are presented with is “full of binary choices”²² of a both “ethical and existential” nature; these antinomic choices always lie “between two opposite, incompatible possibilities”²³ that metapoetically map onto the dichotomy of the Homeric masterpieces. There is nothing fundamentally new in a binary reading of the poem: at least since Viktor Pöschl’s reading of the Aeneid as ‘symbol’ (on which more below), later crystallized in Philip Hardie’s Cosmos and Imperium, we seem to have become accustomed to reading this epic as relying “upon a number of interconnected but distinct dichotomies (e.g. Jupiter-Juno, heaven-hell, male-female) that mark the poem as one of the milestones of Western dualism.”²⁴ What is new, however, in Juno’s Aeneid is the intertwining of this dualistic view with a reprise of Francis Cairns’ interpretation of the poem in the light of kingship theory.²⁵ This allows Farrell to focus upon the ‘ethical’ dimension of these binaries, in an interpretation of the Aeneid where both readers and literary characters assume authorial as well as interpretative roles, which force them to pick a side between Iliad and Odyssey, a binary that is mapped onto the literary, ethical, and historical dichotomy between ‘bad’ and ‘good’ leadership and governance.

In a nutshell, for Farrell, the Aeneid can be read “as a battle to determine which of Homer’s masterpieces it will become,”²⁶ or, as a corollary of this, as the story of whether Aeneas will turn out to be an Achilles or an Odysseus. The argument is so simple as to sound bizarre, though surely there is also something enticing in the lucidity of this interpretative take, and there is much in this book that can and should be enjoyed if we are primarily interested in Virgil’s relationship with Homer, and in the characterization of Aeneas, whom Farrell allows to emerge from this book as a complex and multidimensional but also fully fleshed out character. Starting from Juno, the characters of the Aeneid come to resemble Pirandello’s ‘characters in search of an author’ as they embody their full metapoetic potential and attempt to make choices and narrative interventions that align with either an Iliadic or Odyssean perspective, with a model of kingship exemplified either by Achilles’ (negative) intransigent ‘wrath’ and ‘strength’ or else by Odysseus’ (positive, though not so much in the Aeneid) ‘intelligence’ and ‘cunning’, in a never-ending battle between mênis and mêtis. The philosophical framework against which we should interpret these ethical models is borrowed, as in Cairns, from Philodemus of Gadara’s On the Good King According to Homer, undoubtedly a familiar text for Virgil. Farrell claims that his agreement with Cairns is limited to allowing for the reception of Hellenistic kingship theory to “loom large on the horizon of expectations of ancient readers of the Aeneid,” but that he does not “share the assumption that such factors actually determine what it is possible for the poem to mean.”²⁷ This is refreshing, but ultimately an excusatio non petita when the entire book proves to tell a different story, constantly pushing us to evaluate both the poem and Aeneas within a rather rigid ethical and literary grid.

Similar contradictions taint the use of the Homeric intertext in the poem. On the one hand, Farrell does not want us to read “the Homeric program of the Aeneid” as “a fixed structure with a definite meaning, but [as] something more flexible and provisional,”²⁸ and writes that “we may regard Homer’s meaning itself as not static but dynamic,” and that it is “in the collisions of meaning that take place between Homer’s text and its interpretation by ancient critics … in the intertextual jostling of one Homeric allusion against another within the Aeneid, [that] one gets the impression of something very different from Knauer’s fixed intertextuality.”²⁹ And yet throughout the book, both Farrell and his characters (especially Juno and Aeneas) need Homer’s text to mean something relatively fixed in order to push forward their narrative interventions and interpretations. When Farrell writes that “the reputation of the Aeneid is not founded on its simplicity. In comparison to Homer, both the poem and its hero have sometimes been admired as well as faulted for indirection, indecision, and diffidence,”³⁰ and when he sets out mapping such ‘indirection’ onto a binary choice between Homeric models (“Aeneas is in no way comparable to either Achilles or Odysseus, and this causes him worlds of trouble”³¹), how are we left any room to appreciate Homer’s own literary and ethical complexity?

Just as Virgil’s multivocality, so his complexity can, for Farrell, be flattened into neat patterns and be subsumed into simplicity. “Binary choices,” he admits, “are seductive in their apparent simplicity,” but “they also have a way of breaking down into pluralities.”³² Yet even these pluralities can ultimately be led back to the original Homeric binary. This is the message of the book’s longer middle chapter, ‘Third Ways’, which surveys the import of other texts, heroic models, and literary genres (the Epic Cycle, Apollonius’ Argonautica, Heracles, tragedy, Roman epic and historiography) only to demonstrate that all these ‘third ways’ “do not replace or detract from the Homeric antinomy, but actually reinforce its importance.”³³ Within this Homeric framework, even tragedy—whose influence in the Aeneid has long been emphasised by scholars—becomes just another lens to be applied to the Homeric poems, “always to be understood [with Aristotle] as existing in tension with comedy,”³⁴ and a voice that may exist as an ‘evanescent possibility’ but which must ultimately be rejected in light of the larger modello genere of Homer.³⁵ The same can be said, mutatis mutandis, about Aeneas’ most famously recognisably ‘Roman’ heroic trait: his pietas. Although Farrell does not deny that the Aeneid does at times display “a sacral dimension … that is entirely unlike anything in Homer,”³⁶ even Aeneas’ “signature quality” of pietas,³⁷ translated here as ‘sense of duty’, which is “the usual defense” for Aeneas to be “a new type of hero,”³⁸ must eventually measure up to Homeric standards, becoming a trait that makes him emphatically unlike Odysseus and pushes him closer to the Achillean model that he will eventually embody at the end of the poem.³⁹

Attending to these apparent ‘third ways’ in conjunction with Homeric intertextuality reflects a literary theoretical move away from the model of the ‘dynamic’ intertext towards what Farrell calls the ‘dialogic’ intertext. Farrell credits above all Damien Nelis’ book on the Aeneid and Apollonius’ Argonautica as by far the most influential contribution for this intertextual shift.⁴⁰ He lingers on the recollection of the feeling of shock and unease that Nelis’ work brought upon Virgilian scholarship, reporting James O’Hara’s words on the ‘disturbing sense’ that the Hellenistic intertext adduced onto the Homeric background of the Aeneid, as if suddenly a ‘second moon’ had appeared in the sky of the critic-as-helmsman.⁴¹ For Farrell, too, this moment marked the beginning of disconcerting possibilities, as in the future “one had to be prepared, at least in theory, for yet another moon to appear one day, and another and another.”⁴² He further advocates for his surrendering of “any residual allegiance … to the concept of authorial intention,”⁴³ claiming that this will not play a role in the analysis that the book has to offer, even though the argument seems to imply throughout that the binary choices that readers encounter in the Aeneid are there ‘by design.’⁴⁴

Farrell seems to prepare us, then, for a reading of the text of the Aeneid as placed in a dialogic network within an open-ended literary tradition. And yet, what this ‘dialogic intertext’ looks like, throughout the book (where Apollonius is subsumed into the Homeric binary, quite unlike in Nelis), is little more than the addition of another ‘moon’, or rather another ‘ball’ to Hinds’ billiard table of the Virgilian intertext. If there is one takeaway from the ‘third ways’ chapter, then it is that this table cannot in fact host more than three/four balls simultaneously (Aeneid, Iliad and Odyssey, Argonautica or Aetia or Heracles or Women of Trachis or Bellum Punicum or Annales), all remaining rather solid in their nature, and all arranged in a clearly determined hierarchical array in relation to one another. In Farrell’s analysis, this ‘dialogic intertext’ appears as no different from the relatively hackneyed phenomenon of ‘window reference’ or ‘two-tier allusion’⁴⁵—where the metaphor of the ‘window’ clearly implies a visual hierarchy between the ‘two’ texts that are being alluded to. Both Richard Thomas and Farrell, at least theoretically, do not shy away from making room for a model of open-ended textual ‘allusion’ in which an entire multi-layered tradition of literary and cultural references may reverberate: this has to be Farrell’s ‘combinatory allusion’, akin to what Thomas called ‘conflation’ or ‘multiple reference.’⁴⁶ Indeed, in many ways Farrell’s analysis is always geared towards reading Homeric intertextuality in Virgil not necessarily directly vis-à-vis the Homeric text, but against the entire, and in many ways slippery, tradition of Homer’s ancient reception as made up of re-writings, re-purposing, re-appropriations. And yet the methodological model of ‘combinatory’ or ‘dialogic’ intertextuality is also an obvious hindrance to the book’s main argument on how the intertexts of the Aeneid are always necessarily reoriented back to the ‘Homeric binary’ and its fixed ethical connotations. Within the strictly dichotomous world that Farrell posits in this book, there is ultimately no space for literary and interpretative ‘third ways’, let alone for other moons to appear in the sky.⁴⁷

2 At Least Double

It is worth taking a step back and reflect on how this model of intertextuality posited by Farrell—ultimately coinciding with that espoused in Hinds’ Allusion and Intertext—is not, as it were, an inevitability; rather it looks like a scholarly construction specifically located in the Italo-Anglophone history of Latin literary criticism, where certain methodologies seem to have prevailed at the expense of others.⁴⁸ For what it’s worth, at least Conte credited Julia Kristeva for coining the term ‘intertextuality’ when introducing the work of Mikhail Bakhtin to a Western audience. Yet her application of intertextuality, for Conte, had an “ideological import” that extended “too far for the concrete needs of the philologist, who is a less abstract analyst of text”; thus, “we should probably accept the term and seek to redefine it.”⁴⁹ Conte claims to be following Gérard Genette, whom he credits with “purifying [intertextuality] of the ideological overtones found in Kristeva’s work” and making it “a more neutral instrument to philological analysis.”⁵⁰

It would be hard to downplay the political import of Conte’s move, which bears on the entire conceptualisation of a discipline that is currently attempting to engage in a salutary process of self-scrutiny over what and who gets excluded from the field when philology, as Patrice Rankine puts it, “pretends to be a neutral and disinvested test of intelligence.”⁵¹ By positing that Kristeva’s conception of intertextuality had an ‘ideological’ (read not just deconstructionist, but specifically ‘feminist’) dimension from which it had to be ‘purified’ before being handled by the supposedly clean hands of the classical philologist, Conte propagates yet another long-lived ‘ideology’ of seemingly socially detached literary criticism, whose normative and conservative dimension could not be farther removed from the supposed apolitical stance that it pretends to profess. This is perhaps not surprising from a critic whose latest polemical pamphlet opens with a request to philologists to follow a ‘categorical imperative’ reminiscent not just of Aristotle, but of the 19th century German historian Leopold von Ranke, arguing that our duty as interpreters is to establish ‘how things actually are’, or were.⁵² What is instead surprising is that this ‘purified’, de-politicised, de-feminised version of ‘intertextuality’, reconfigured against a conception of literature as in a Bloomian framework of Oedipal (or else Phaethontian) ‘son vs. father’ competition, was eventually carried forward by Latinists who would never espouse the type of philological dogmatism that Conte ended up embodying.⁵³ As noted by Gunderson, both Kristeva and Bakhtin disappear from the archives of intertextuality in Hinds’ Allusion and Intertext,⁵⁴ and the whole Cambridge series ‘Roman Literature and its Contexts’, which started on the premise to showcase Latin scholarship’s innovative engagement with literary criticism, ended up pushing aside the deconstructive turns of Duncan Kennedy (1993) and Charles Martindale (1993) in favour of a Bloomian model of intertextuality as literary emulation in a “linear descent and orderly succession”⁵⁵ that can and should be “illuminated by the bright light of philology.”⁵⁶

Linearity and literary genealogy had no place in Kristeva’s conception of intertextuality as presented in her well-known (perhaps outside of Classics?)⁵⁷ essay ‘Word, Dialogue and Novel’, originally written in 1966 and translated into English in 1980.⁵⁸ Here, linearity gives way to intersection, imitation to repetition, challenge to dialogue. The antinomy assumed in the basic Saussurian dyad of signifier/signified is reconceptualised for a discursive model in which the poetic word, as the minimal unit of language, is not standing for anything outside of discourse, but is always ‘one and the other’, always ‘at least double’, acting not as a signifier for a fixed signified but as a ‘multi-determined peak’, which is in continuous dialogue and transformation with other such ‘peaks’, and in a constant flux where ‘meaning’ is necessarily multi-determined by the infinite array of combinations that the poetic word establishes with others, in and outside its text.⁵⁹ There is no room, in Kristevan literary semiotics, for the kind of dyadic Aristotelian logic that underpins much of Western scientific discourse: rather, she presents Bakhtinian ‘dialogism’ as “a logic of distance and relationship between … different units”, “a logic of analogy and non-exclusive opposition, opposed to monological levels of causality and identifying determination”, and she gestures towards the possibility of looking instead at the non-dualistic logics of Taoism as an alternative philosophical model for conceiving the workings of poetic language.⁶⁰

It is then apparent that the original Bakthinian/Kristevan conception of literary ‘dialogism’ has little to do with the kind of ‘dialogic intertext’ presented in Virgil’s Juno. Indeed, Farrell’s insistence on the importance of identifying and recognizing dyads, dichotomies, binaries, polarities as the backbone of the design of the Aeneid, and as our main thrust for interpretation, could not be further from the type of logic that poetic language is said to embody in Kristevan theory. To repeat from above, the disavowal of the original conceptualisation of intertextuality in Farrell’s book was never an inevitability. Farrell refers once, in a footnote, to what is to my knowledge the one theoretical essay by Duncan Kennedy (1992) that really attempted to put deconstruction at work in Virgilian literary criticism, when reacting to the kind of monological and dualistic thinking that underpinned both sides of the debate over the pro- or anti-Augustanism of the Aeneid. Yet, the scholarly reception of this important reflection on the workings of both poetic language and hermeneutic practice seems to have paid to Kennedy lip-service at best.⁶¹ While the piece has undoubtedly been influential in “unleash[ing] Virgilian studies free to investigate new shores”,⁶² its ultimate methodological message seems to have fallen on deaf ears. Similarly, Farrell makes frequent use, throughout the book, of the work of Don Fowler, but there is no sustained engagement here with the salutary theoretical challenges that Fowler famously posed to the Contian model in ‘On the Shoulders of Giants’ and elsewhere.⁶³

Conte is obviously not wrong in pointing out that there is an ‘ideological’, or we may less derogatorily say ‘political’, stake in deconstruction; just as there are different ideological, or political stakes in the kind of opposition that he offers.⁶⁴ After all, the Aeneid is a profoundly political poem, not least because of the long history of its political appropriations,⁶⁵ and the way in which we choose to read it has inevitably political repercussions. Although this is clearly not Farrell’s open, or even conscious, agenda, there are consequences in promoting the idea that pluralities must ultimately be reconfigured into binaries, especially at a time like the present, when the political battles of oppressed parties have long theorised the importance of deconstructing the oppressive hierarchies of much of (Western) binary logic and thought.⁶⁶ This is not to implicate this book specifically in such dynamics of oppression (Farrell is in fact quite vocal in his standing on the progressive side of both American politics and the reconfiguration of Classics as a discipline),⁶⁷ but the wider import of this mode of thinking cannot be understated, especially as it collides badly with the expectations of readers who, judging from this book’s focus on Juno, may expect it to engage in discussions of gender or feminism as pertaining to Virgil’s poetry.⁶⁸

3 Narrative and Imagery

Although Farrell professes a ‘formalist’ methodology,⁶⁹ Juno’s Aeneid, as its original subtitle announced (‘Narrative, Metapoetics, Dissent’), is in the end a book about ‘Narrative’, which after the first chapter showcases surprisingly little engagement with the more obviously formalist practice of ‘close reading.’⁷⁰ One may of course not fault Farrell for an emphasis on plot, structure and narrative when dealing with epic, which not just for Aristotle but even for Kristeva retains an ‘organization principle’ of ‘narrative infrastructure’ that is ultimately ‘monological’ and relies on a logic that is ‘causal and theological.’⁷¹ And yet, such a reading of the Aeneid downplays the central role that metonymy and free association play throughout a poem whose richness of imagery and cross-generic interaction, especially in the use of similes, often speaks against the linearity of epic and privileges intratextuality with Virgil’s previous Eclogues and Georgics at the expense of Homeric engagement. To me, this is a viable third way of reading the Aeneid outside of the Homeric binary and its antinomic heroic code if ever there was one, but in this book it appears as yet another possibility for diffraction of meaning, the door to which is left rather firmly shut.

Farrell’s approach to the Virgilian simile is presented early on in Juno’s Aeneid as underpinning the dyadic methodology of the book as a whole. He invites us to regard epic similes “not as overqualified or overdetermined but as overadequate to their specific purpose.”⁷² In Virgilian similes, Farrell claims that readers should concentrate on the basic ‘third point of comparison’ between tenor and vehicle: “one might then enjoy other details from an aesthetic point of view, but in cognitive terms one could disregard them as excess.”⁷³ It is in this eventual ‘disregard’ for what is presented as a purely pleasurable experience of aesthetic form, strangely disentangled here from ‘cognitive’ processes of interpretation, that Farrell’s applied methodology seems to be surprisingly at odds with the supposedly formalist approach that he professes to take shortly afterwards in the book, where formalism aligns solely with ‘narrative’ (structure) and ‘metapoetics.’⁷⁴

This methodology for reading the Virgilian simile is borrowed from David West’s now canonical 1969 article on Virgil’s ‘multiple-correspondence similes,’ which established a clear-cut dyadic division between narrative and simile as existing not in a permeable relationship, but on explicitly separate planes, interacting with each other according to a rigid grid of three possible correspondences (‘bilateral’, ‘unilateral’ and ‘irrational’), in a division ostensibly aimed at ‘containing’ or making sense of any possible spillage from the world of the vehicle into the world of the tenor.⁷⁵ For West, ‘bilateral’ correspondences involve explicit matching details between simile and narrative; ‘unilateral’ correspondences still indicate logical matches, but they require readers to “clearly supply in our imagination” the corresponding details from simile into narrative.⁷⁶ West includes a type of ‘irrational’ correspondence to indicate illogical and uneven matches, including details that resemble and echo one another, sometimes in an auditory form but not at the narrative level, providing shocking juxtapositions. Yet he insists that such disjunctions are also part of Virgil’s grand design. There is nothing truly irrational about West’s ‘irrational’ correspondences: just like bilateral and unilateral correspondences, they too are “not coincidence. [They are] engineering.”⁷⁷

In the article’s last sentence, Virgil is not only an engineer, but a ‘miniaturist’ and ‘a symbolist.’⁷⁸ This sounds like a clear nod to Viktor Pöschl, who suggested that the greatness of the Aeneid lies precisely in its conscious appropriation of the ‘symbol’, in stark opposition to Homer, who was, for Pöschl, symbolic ‘malgré lui.’⁷⁹ Interestingly, the German word used by Pöschl to describe the Aeneid is not just Sinnbild but Gleichnis: that is, a ‘simile’ (as well as ‘a parable’), as if the proliferation of similes and images presented throughout the poem were to be read as a pars pro toto, and meant to help in unlocking the hidden meaning of the masterpiece.⁸⁰ That such meaning could indeed be unlocked by the critic is implied in Pöschl’s simultaneous positing of the Aeneid as Deutung (‘interpretation’, ‘explanation’) for the historical context. The relationship between simile and narrative becomes synecdochic of the poem’s relationship to its historical context, in a poetic work that eventually provides its critics with the keys for its deciphering. We can say that both the Aeneid and its similes can be read with Pöschl as a Saussurian signifier/signified dyad, disabling the proliferation of meanings, intertexts, and images that readers may otherwise want to see as propagating in and outside the poem.

Neither Pöschl nor the other classic Brooks Otis—who emphasised the way in which Virgil’s ‘symbol-structure’ allows “symbols and motifs to ‘correspond’ to or be attached to affective or emotive elements of his narrative”⁸¹—gives us a working definition for the ‘symbol.’ Yet their dyadic usage brings to mind Carl Jung’s critique of Freud’s employment of the word ‘symbol’ rather than ‘sign’ to indicate an unconscious impulse, or image, that can be interpreted and made to ‘stand for’ something else in the conscious mind, against which Jung posited that symbols are instead inherently ‘ambiguous and bipolar.’⁸² Their very German word, Sinnbild, combines conscious sense, ‘meaning’ (Sinn) and unconscious archetype, ‘image’ (Bild),⁸³ indicating that the symbol is simultaneously grounded in the unconscious but moulded in and by the conscious mind.⁸⁴ For Jung, symbols are not explicable by terms other than their own: they are “the expression of an intuitive perception which can as yet neither be apprehended better not expressed differently.”⁸⁵ A lack of reflection on this inherently polyvalent, and inherently inexplicable value of the symbol leads Virgilian scholarship to interpret the Aeneis-Gleichnis and the similes included therein to function as ‘signs’ rather than symbols, which are employed by the poet, as Pöschl writes, ‘with conscious intention’,⁸⁶ as something expecting interpretation in the form of some kind of decoding. There is no room here for ‘irrational’ correspondences because there is no room for irrationality full stop.⁸⁷

Both the search for structural patterns and the interpretation of the symbol underpinned Bernard Knox’s 1950 article ‘The Serpent and the Flame’, a true milestone in the application of New Critical theory to the study of the Aeneid. West curiously embraced Knox on the matter of detail while differing significantly in approach.⁸⁸ In the simile of Book 2 comparing Pyrrhus to a snake which has just shed its skin (Aen. 2.469-275), West adopts Knox’s reading of the correspondence between the serpent casting off its slough and Pyrrhus/Neoptolemus as a sort of Achilles in a ‘new’ skin.⁸⁹ However, West’s need to label this correspondence as ‘unilateral’ and tie it to his schematic grid erases one of the central messages of Knox’s article, namely that Virgilian similes are characterised by their unpredictable and uncontainable permeability with the narrative.

In Knox’s reading, Pyrrhus’ snake simile comes as “the turning-point in the development of the dominant metaphor” that had already permeated the first half of Book 2.⁹⁰ Snakes acted implicitly as vehicle for the disguised Trojans in the simile comparing Androgeos to a man inadvertently stepping on a snake (Aen. 2.378-382), and they featured as monstrous actors in the narrative in the form of the serpents from Tenedos, which Servius already interpreted as standing metaphorically for the Greek ships.⁹¹ On top of these explicit references, Knox draws attention to the surprising and potentially ‘inexhaustible’ series of snake-related metaphors throughout Book 2.⁹² Metaphorical serpents are not just the treacherous Greeks hiding in Tenedos, and in turn the Trojans who ambush Androgeos with Greek helmets whose crests (2.412 Graiarum … iubarum) recall the bloody crests of the serpents (2.206-207 iubae … sanguineae),⁹³ but also the horse who stealthily ‘slides into the city’ (2.240 inlabitur urbi) on the ‘gliding of wheels’ (2.235-236 rotarum lapsus),⁹⁴ and the treacherous sleep that ‘crawls’ into the Trojans’ limbs (2.269 serpit).⁹⁵ And serpent-like is also the flame that ‘licks’ Ascanius’ hair (2.684 lambere flamma comas) and ‘feeds’ around his temples (2.684 et circum tempora pasci), just as flames had done throughout the city, in an echo of the serpents ‘licking’ their lips with their flickering tongues (2.211 sibila lambebant linguis uibrantibus ora) and ‘devouring’ the limbs of Laocoon’s children (2.215 miseros morsu depascitur artus).⁹⁶ This is a final twist in the story of the snake-metaphor that takes on a “plot of its own”⁹⁷ in Book 2 and undergoes a metamorphosis of signification: from destruction and death to metamorphosis and rebirth, indicating on the head of Ascanius that Troy will be reborn from its ashes, like a serpent sloughing its skin. Crucially, for Knox, any attempt to support such analysis “by an appeal to Virgil’s intentions would be barren and irrelevant.”⁹⁸ But no more convincing are, for him, appeals to the unconscious, as in Robert Cruttwell’s attempt “to probe Virgil’s mind at work”,⁹⁹ a task which lies beyond the powers of critical analysis. “A glimpse into the ‘chambers of [Virgil’s] imagery’”, writes Knox in his close, “reveals only further chambers beyond.”¹⁰⁰

Farrell discusses the snake simile of Pyrrhus in Aeneid 2 as part of his discussion of Aeneas’ transformation into a Homeric hero (the book’s third and final chapter), but his lens on characterisation keeps the focus sharp on how Homer’s binary heroic model turns Pyrrhus’ “combination of Achillean force governed by Odyssean cynicism” into “a nightmarish parody of the ethical heroism to which [Aeneas] aspires.”¹⁰¹ Despite the initial promise of a formalist as well as intertextual methodology, there is little interest here in the chains of associations that the snake simile unleashes in and outside this poem, beyond its immediate referent in the narrative. This is no doubt a very practical interpretative and argumentative choice, but it is not the only viable one when reading the Aeneid. The permeable relationship that Virgilian similes establish between simile and narrative, a relationship that is always necessarily co-created by us as readers in our interaction with the text, forces us to confront head-on the question of what philosophical worldviews are afforded by a poem that posits continuous analogies between the animate, the inanimate and the abstract. We may recall that in 1986, Philip Hardie’s Cosmos and Imperium imported into the Aeneid a Lucretian interpretation of Virgilian similes in support of a philosophical worldview according to which “the structure of things is homogeneous at all levels.”¹⁰² Just as, in Lucretius, “the ornamental comparison becomes a vehicle of intellectual revelation,” similes in Virgil also express “an underlying analogy or even identity in the nature of things” that continuously encourages us to “adjust our perception” of both tenors and vehicles, that is on both the structure of the phenomenological world and on the workings of the human mind.¹⁰³ This principle of analogy, of similes predicated on likeness rather than difference, on simultaneity rather than diachrony, on pleasure and repetition rather than competition and imitation, opens up a possibility for reading the Aeneid as a poetic text in accordance to that logic of “analogy and non-exclusive opposition” that Kristeva predicated.¹⁰⁴ The likeness of similes allows a whole other world, a whole other poem, a whole other story to sprout while temporality in the previous story is suspended as the simile unfolds: they point us towards the ‘ebb and flow’ of the poem, “an extension into universal time in their short-lived moment.”¹⁰⁵

Alice Oswald has recently spoken of a gendered view of similes as sites of pregnancy and gestation, generative and relational, which she finds akin to the “agitated or animated reflections” produced on the surface of water.¹⁰⁶ Oswald’s image brings to mind the simile of Aeneas’ mental and emotional state diffracting and moving like the light reflected in the water of a cauldron at the beginning of Book 8:

We can of course follow in Servius’ footsteps and emphasise the Apollonian intertext (Arg. 3.755-760), pointing out, with Farrell, similarities and differences between Aeneas and Apollonius’ Medea. Other critics may combine the Apollonian intertext with Lucretius’s DRN 4.211-213, where the floating reflection of the stars in the water corresponds to the floating rapidity of the appearances of things (DRN 4.210 rerum simulacra) as they come into contact with our eyes, and build up an argument on the philosophical and existential implications of the image. Either way, something may get lost in the intertextual game when we forget to focus on where and how these images take us as readers who read with our feeling bodies, enjoining us to share in the ‘magnified’ and already ‘boiling’ surge of emotions that is stirring up in Aeneas’ chest (Aen. 8.19 magno curarum fluctuat aestu) just as anger did, previously, in Dido’s (Aen. 4.532 magnoque irarum fluctuat aestu), and giving us full space for diffraction and free association in the surging of images—the swiftness of the deranged mind (mimicked by the elision and dactyls of line 8.20 atque animum nunc huc celerem), the sense of separation from oneself (8.20 nunc diuidit illuc), the turbinous carrying away of wave-like feelings (8.21 in partisque rapit uarias perque omnia uersat)—even before we are allowed to expand ourselves in the space in the simile, to become ‘like’ (8.22 sicut) ‘the water’s’ (8.22 aquae)¹⁰⁷ ‘trembling’ (8.22 tremulum) ‘on the lips’ (8.22 labris)?—ah no, this is the ‘trembling light in a vessel of bronze’ (8.22 labris ubi lumen aenis)¹⁰⁸—we’re repeatedly pushed, made to reverberate, by the sun (8.23 sole repercussum)—invited to re-read, re-peat, re-live—or (8.23 aut) we may be reflecting something equally radiant (8.23 radiantis), the image of the shining moon (8.23 imagine lunae), with this basic astral dyad fusing into one from one side to the other of the line (8.23 sole … lunae)—again perhaps joining Aeneas together with Dido in their previous similes as Apollo and Diana, the sun and the moon—,¹⁰⁹ and then we are invited, with Aeneas, to lift and fly around (8.24 omnia peruolitat late loca), across space at large (late), or more specifically across Lat-ium, a Lucretian flight of the mind perhaps,¹¹⁰ to move from water to air (8.24 sub auras), standing up tall for a time (8.25 erigitur), but eventually hitting a roof (8.25 summique ferit laquearia tecti), surprisingly finding ourselves secluded, or trapped—Latium spells ‘hiding’ (latere), not ‘widely’ (late), seclusion at the expense of extension—and not just any roof, not (yet) Latinus’ humble mansion (8.366 angusti … tecti) that will soon host ‘massive Aeneas’ (367 ingentem Aenean)—no, we are taken back to Dido again, to the lofty laquearia of her palace (1.726 laquearibus), under which a funeral pyre ‘stretched’ indeed ‘to the sky’ (4.494 tu secreta pyram tecto interiore sub auras / erige).

It is especially the emotional attachment that may get lost when we focus on intertextuality rather than intratextuality, on difference at the expense of analogy. For Farrell, the simile “represents the confusion of a king who is not at all in love, though he is supposed to be getting married, and whose people are under the threat of imminent war.”¹¹¹ But on what grounds is Farrell’s Aeneas ‘not in love’? What allows us to read Aeneas’ curae as a signifier for a stable signified, rather than the unstable referent to a wide range of feelings—among which we may list, in random and non-exhaustive order, love, grief, worry, apprehension, longing, desire, escapism, sense of entrapment etc.—, feelings which the simile allows us to put into words? And what difference does it make if we allow the simile to get under our skin, to read it with the fleshy, feeling body of someone who is experiencing this poetry, rather than attempting to explain it? It is only in the very conclusion of Juno’s Aeneid that Farrell gives us a glimpse of what his reading through emotions rather than narrative or intertextuality may look like: he sketches the poem’s progression through the four ecphrases that Aeneas is called to interpret in the poem not as a decoder of images-texts but as ‘an emotional reader’, whose emotional reaction to the images seems to become more violent just as his ‘literal comprehension’ decreases.¹¹² Farrell’s brilliant insights in the close of the book are well worth pursuing, I think, and that may well be where he is heading next, even if this book salutes us by reminding us that this was all about intertextuality, ‘through and through’, when telling us that “the poem ends as an Iliad, through and through.”¹¹³

4 Meaning

In a book, and a poem, that has been held hostage to the constraining interpretation of signs, similes may encourage us to read the Aeneis-Gleichnis and the analogies therein included not necessarily as a quest for meaning, but as vehicle towards experiencing what cannot otherwise be apprehended nor expressed. Farrell may well, and rightly so, profess agnosticism, over and over, on the possibility of finding an ultimate meaning to the Aeneid—but even entertaining the possibility that such a search may be part of the critic’s task betrays an allegiance to modes of reading that come at the expense of highlighting what the poem can do and can stir, what political and emotional attachments it can unlock or assuage, what connections it can establish, what worlds and viewpoints it can reconfigure.

For Farrell, as for many other Latinists, the dominant metaphor of criticism remains that of a search into the ‘depths’ of what is ‘concealed’ by the text, and which is up to the knowledgeable critic-as-archaeologist to ‘uncover’ and ‘bring to the surface’:¹¹⁴

We find throughout this poem that an apparently dominant surface meaning rests upon, but does not quite conceal, a riot of conflicting possibilities roiling in the depths below, some of which occasionally burst through the surface and present themselves for all to see. (p. 49)

Yet we are simultaneously encouraged to think of the poem as “more of a Rorschach test than a manifesto, something that reveals readers to themselves.”¹¹⁵ It is not obvious to me that, by the time we come to the end of this book, Farrell has properly laid his cards on the table and told us what his Rorschach test of the poem revealed about himself as a reader and critic. Perhaps this will be for his own critics-as-analysts to assess, but only if we keep being so invested in conceiving of reading as diagnosis when discussing the workings of poetry and literary criticism.