The Culex—the earliest and best attested of the purported minor works of Virgil, and the most outright in gesturing towards Virgilian authorship—poses a problem for modern classical scholarship.Footnote 1 Since at least the seventeenth century scholars have been preoccupied with the poem's authenticity.Footnote 2 Is it a piece of early Virgilian iuuenilia, as the ancient testimonies and mediaeval transmission of the text seem to assert, or a later production? If a later production, should we see it as a deliberate forgery, or as a poem severed in the course of transmission from its original author and helplessly swept up in Virgil's train?Footnote 3 The authenticity problem has proven persistent: as recently as the 1970s, scholars tried to claim the Culex for Virgil.Footnote 4 Even among those who think it non-Virgilian, the apparent consensus of anonymous late-Tiberian authorship has been contested by Otto Zwierlein's suggestion of M. Julius Montanus and Jean-Yves Maleuvre's, even more unlikely, of Augustus.Footnote 5
In more recent years, however, authenticity studies have changed focus. Attention has shifted from pure Echtheitskritik—the busy philological work of athetising supposed interpolations, reassigning stray adespota, tying up the loose ends of the canon—to the literary and cultural function of the inauthentic text.Footnote 6 Irene Peirano's game-changing 2012 monograph, in particular, recasts these troublesome works as literary artefacts worthy of investigation in their own right. Rather than indict their authors for not only malicious but—worse!—incompetent forgery, and condemn their readers for being duped by such transparent deceptions,Footnote 7 she argues that we should allow for rather more sophistication. Authorial impersonations like the Culex seem to function as ‘complex literary games of concealment and revelation in which readers are both teased into taking the text as authentic and simultaneously signalled its artificiality as a literary construct’.Footnote 8 In the case of the Culex, the prospect of a sixteen-year old Virgil dabbling in surreal neotericism is (just, almost, tantalisingly) believable.Footnote 9 The fun is in doggedly continuing to suspend disbelief even in the face of persistent and insurmountable challenges to that disbelief: a Virgil who sounds suspiciously Ovidian,Footnote 10 who hymns Octavius as a nine-year old puer (Culex 26, 37) but describes a gnat's tomb eerily reminiscent of Augustus’ mausoleum,Footnote 11 who writes a pastiche of his three mature canonical works at the very beginning of his career.Footnote 12
‘The problem of the authenticity of the Culex, like the corpse of its heroic flea, simply will not die’, Glenn Most remarked in 1987: ‘it returns to complain of ill-treatment and to haunt those who thought they had killed it.’Footnote 13 Thirty years after his piece so convincingly laid to rest any lingering belief in Virgilian authorship, and with the rehabilitation of pseudepigraphic and impersonatory texts well under way, the time is perhaps ripe to reassess how the poem self-consciously signals its peculiar form of overdetermined inauthenticity and simultaneous protestation of true-born Virgilian authenticity.Footnote 14 Taking Most's passing comment not just as a wry verbal flourish but as an insight into the relationship between the poem's contents and its nature, purpose and reception as a whole, I propose that the eponymous gnat and the poem's authenticity have more in common than their irritatingly recursive behaviour alone. From its opening salvo at the hypothetical hostile reader (inuidus, 5) and its exploitation of established metapoetic gestures in the proem to the narrative structures and metaliterary self-positioning seen throughout the rest of the poem, the Culex repeatedly stages scenes of critical reading and interpretation falling somewhere between truth and deception, face-value and falsity. Reconsidering the gnat's narrative as an embedded ‘poem within a poem’, complete with its own narrator, audience and reception, provides insight into how the Culex models processes of reading (or rather, misreading) within the fiction, and thus how it prompts its own readers to approach it as a piece of literature.
Authorial Games and Flyweight Critics
The first few lines of the Culex programmatically sketch out the simultaneous qualities of veracity and willing deception that characterise the poem as a whole. The first word is lusimus (‘I have played’), a verb frequently connected to poetic composition in ‘light’ or ‘low’ genres. A certain neoteric or Alexandrian allegiance is certainly appropriate to the Culex. The poet overloads the proem with poetological keywords, ‘all the tired clichés of Augustan neotericism’, as David Ross has it:Footnote 15 gracili modulante Thalia (‘while willowy Thalia played the accompaniment’) in line 1; tenuem (‘slender’), the Catullan araneoli (‘spiderlings’)Footnote 16 and formauimus (‘I have shaped’) in line 2; docta (‘erudite’) in line 3.Footnote 17 But the verb also refers to playing a part, or participating in a performance.Footnote 18 The poet declares that the conceit of Virgilian authorship is a temporary pose or persona, not true identity: ‘I have played a role’, that of Vergilius iuuenis (‘Virgil as a young man’). Already in the first lines of the poem the reader is encouraged to maintain a double vision: on the one hand, suspending disbelief, taking the proem as sincere metapoetic statement and reading the poem in line with established literary tropes; on the other, stepping back from participating in the poem's fiction and instead watching the authorial mechanisms at work behind the scenes. We must understand the quasi-titular lusimus as a conventional marker of generic affiliation and, at the same time, as an admission of authorial play, impersonation and even deception.Footnote 19
The Culex-poet's openly impersonatory use of ludere in fact paradoxically strengthens the effect of his Virgilian impersonation. Virgil uses ludere of his own composition of the Eclogues: ludere uersu (‘to play in verse’, Ecl. 6.1), carmina qui lusi pastorum (‘I who played shepherds’ songs’, G. 4.565). The verb was swiftly taken up as a code word for Virgil's earliest authorial efforts: within his lifetime, Propertius characterises the author's composition of the Eclogues and Georgics with ludebat (‘he played’, 2.34.85).Footnote 20 A few decades later, Ovid, justifying his own poetic transgressions, outlines Virgil's career:
Before this [i.e. before the Aeneid], when he was young, this same man played in bucolic measures the passions of Phyllis and tender Amaryllis.
But even in Virgil's original usages, the verb ludere already has impersonatory undertones: who speaks as narrator in Ecl. 6, Virgil or Tityrus (6.4)? Who sang of Tityre…patulae…sub tegmine fagi (‘Tityrus under the shelter of a spreading beech’, G. 4.566)—the Virgil who now authors the Georgics, or Meliboeus, the character who originally uttered these words at Ecl. 1.1? The Culex-poet's use of ludere reads Virgil's in terms of authorial uncertainty, and rewrites it in the context of a supposedly ‘early’ work of dubious authorship: in other words, an impersonation of an impersonatory stance.Footnote 21
At first sight, the tense of lusimus seems to mark a transition from playful iuuenilia to weightier genres, fitting in neatly with Virgil's own characterisation of his upwards poetic trajectory (cf. lusi, G. 4.565). The third line reinforces the break with the poet's past endeavours: lusimus: haec propter culicis sint carmina docta (‘I have played: because of this, let the song of the gnat be erudite’, Culex 3).Footnote 22 But the next line blurs the categories of ‘playful’ and ‘serious’, and muddles the distinction between past and present poetic activity. The poem's erudition, its learned nature, is to be achieved per ludum, ‘through playing’ (4). Of course, this commingling of ludus and doctrina in literary works is an old trope. As far back as Plato's Symposium Agathon could declare that he wrote a speech τὰ μὲν παιδιᾶς, τὰ δὲ σπουδῆς μετρίας, καθ᾽ ὅσον ἐγὼ δύναμαι, μετέχων (‘mixing, as far as I can, the playful with the measuredly serious’, 197e).Footnote 23 By the time of late Republican and early Augustan neotericism, the imbrication of the two poetic registers was near compulsory.Footnote 24 But even given this context, the Culex-poet's emphasis on doctrina is extreme. Not only does it receive further definition, but the coexistence of ludus and doctrina is vaunted as the sole criterion by which the poem should be evaluated:
Let the carmina culicis Footnote 25 be erudite, such that the whole structure is playfully consistent with traditional subject matter and the words with the known style of leaders in the field, even if a hostile critic is present. Let anyone who is ready to blame my jokes and my Muse be considered lighter in weight and reputation than even the gnat itself.Footnote 26
According to the proem, then, an ‘erudite’ quality in poetry consists of the adherence of the poem's structure and style (ordo, uoces) to existing literary tradition (historiae) and the familiar style of canonical authors (notitiae ducum),Footnote 27 and we are to judge the poem accordingly. And how do we judge it? Taking lusimus as an admission of impersonation, it turns out that in this definition of doctrina the Culex-poet has indeed sketched out the parameters of the poem's imitatio: the Culex follows not only a plot structure recognisable from literary historyFootnote 28 but also the ordo of Virgil's literary career, from bucolic to georgic to epic; its uoces are almost entirely echoes of Virgilian and Ovidian phrases and diction. ‘I have played a role, and so let the poem be erudite’: the Culex achieves doctrina per ludum.
But doctus is a slippery term too. As well as denoting a sincere scholarly learnedness, it overlaps with callidus: ‘clever’, yes, but tricky and sly too.Footnote 29 We might think of Plautus’ perpetual docti doli (‘cunning tricks’),Footnote 30 or Ovid—self-proclaimed doctus amator (Ars am. 1.1f.)—and his characterisation of both love and poetry as intimately bound up in strategies of deceit.Footnote 31 To be doctus per ludum is to be not only ‘erudite, albeit in light genre’ but ‘cunning through trickery’.
And looking back at line 3 we find another ambiguity, this time syntactical, at play in the term culicis carmina. culicis is usually understood as an objective or definitive genitive, denoting the theme or title of the work: ‘the poem about the gnat’, or ‘the poem The Gnat’.Footnote 32 But it can also be construed as a subjective genitive, used not for the subject of the work but for its author: ‘the poem of/by the gnat’.Footnote 33 Compare Martial's Capitolini…carmina belli (‘poems about the Capitoline war’, Epigr. 5.5.7) with opus Maronis (‘Virgil's work’, 3.38.8) or libris Ciceronis aut Maronis (‘books by Cicero or Virgil’, 5.56.5): in the first instance the genitive denotes the theme or title; in the others it indicates the author. The gnat is the only speaker in the poem other than the narrator; its enormous speech of 174 lines comes close to overwhelming the narrator's 240 (or 238, if the epitaph's tacita uox, ‘silent voice’, in 413f. is subtracted). Its narration is introduced with cecinit (‘he sang’, 209): this is unambiguously a carmen. Which song is it that the proem brands doctum, the Culex as a whole or the gnat's narration?Footnote 34 Can we trust the gnat any more than we can trust the elusive Culex-poet hiding behind the pose of Vergilius personatus? Is the inuidus, the hostile reader quick to censure, to be envisaged as an external reader with the Culex in his hands—or as a character encountering the culex within the poem?Footnote 35
Dangerous Misreadings
The proem invites us to equate the culex with the Culex and to keep our eyes peeled for deception and ambiguity; the remainder of the poem illustrates just how fraught an exercise ‘reading the gnat’ might be. At the first appearance of the gnat, the consequences of misreading are brought sharply home. Our protagonist, the hapless goatherd, is suddenly and painfully woken from his afternoon nap by the bite of a gnat (184–7). Failing to recognise the altruistic gnat's good intentions—an attempt to warn the goatherd of the imminent threat from a monstrous snake (157–84)—he immediately squishes the poor insect:
The goatherd then notices the snake on the verge of attack and kills it too (189–201), still dozy from his nap but therefore blessedly ignorant of the gravity of the situation: erat tardus somni languore remoti | nec senis aspiciens timor obcaecauerat artus (‘he was dull from the drowsiness of the sleep he had just escaped, and so terror at the sight had not yet numbed [lit. ‘blinded’] the limbs of the old man’, 198f.).Footnote 36 But peace still eludes him: his sleep that night is disturbed by the vengeful effigies…culicis (‘ghost of the gnat’, 208), who complains at length of his unfair death and underworld torments (210–383) and demands that the goatherd provide him with proper burial so he may find rest in the abode of the righteous (pia sedes, 375).
This part of the plot rests on misinterpretation. The goatherd acted according to his understanding of the situation—but it swiftly becomes apparent that this understanding was partial and mistaken (uix compos mentis, ‘scarcely in control of his mind’, 191), and his error leads to grave consequences for both the gnat and the goatherd. Not just the goatherd's cognition but his sight too has been compromised: even though terror had not yet ‘blinded’ him (obcaecauerat, 199), the gnat still bit him right in the pupil of his eye (184–7).Footnote 37 Here we see that long-established metapoetic turn whereby internal acts of interpretation or reading suggest modes of and models for external reading of the text itself:Footnote 38 what consequences await the reader of the Culex, when the goatherd's misreading of the culex is so grievous? Indeed, furibundus (‘beside himself with rage’, 187) might be a somewhat bathetic reference to another morally ambiguous murder, famously committed in a fit of rage prompted by an act of reading: Aeneas’ killing of Turnus goaded by his interpretation of the balteus (‘sword belt’). Where the goatherd is furibundus, Aeneas is furiis accensus (‘spurred on by rage’, Aen. 12.946); though the goatherd's sight is partial and Aeneas is able to ‘drink in deep with his eyes’ the sight of the balteus (oculis…hausit, 12.945f.), both react with immediate responses to their perceptions of the situation.Footnote 39 The Culex replays this climactic moment of Virgil's epic, accentuating the pitfalls and ambiguities of reading and interpretation by explicitly depicting it in failure.
Disappearing into Thin Air
The gnat's visitation as effigies to the sleeping goatherd (208f.) and its lengthy katabasis-narrative (210–384)—the culicis carmina—are particularly important for the poem's thematisation of its own deceitful nature in the form of the gnat. True, the ghostly gnat is not suspected of falsity or deception by its internal audience: the goatherd is emotionally struck by the pathos of the gnat's tale (nec tulit ultra | sensibus infusum culicis de morte dolorem, ‘nor could he bear any longer the sorrow for the gnat's death that flooded through his senses’, 386f.), and answers its implicit request for burial (modo sit dum grata uoluntas, | exsistat par officium, ‘let there only be a grateful heart, an equal service rendered’, 230f.) by building it a grandiose tomb (385–414).Footnote 40 Yet various elements of the gnat's apparition and subsequent narrative certainly lead a more sceptic listener to question its veracity.
What kind of ghost is the gnat? Any attempt one might make towards applying a typology of apparitions to the Culex is foiled by the fact that the gnat's ghost—in both appearing to the sleeping goatherd and also narrating its journey through the underworld—seems to conflate two types of encounter with the dead commonly found in classical literature. First, the apparition or dream, in which a living character encounters some immaterial semblance of another; second, the nekyia or katabasis, in which the living protagonist goes to a ‘place of the dead’ and encounters and speaks with the ghosts of the dead.Footnote 41 Both of these categories are characterised by uncertainty, unreliability and instability, in historical and religious accounts as much as in narrative literature.
Dreams and apparitions are difficult to pin down for precise categorisation: the encountering character might be awake, sleeping or in a liminal state between the two;Footnote 42 and the apparition might be of a living or dead character or one of ambiguous vitality.Footnote 43 Dreams and apparitions, as well as the messages they convey, can be true, false or misinterpreted.Footnote 44 There is substantial ambiguity in the Latin terminology of phantoms and ghosts, too.Footnote 45 While effigies (Culex 208) and simulacrum refer primarily to false or fictional apparitions and manes (Culex 214) and umbra to the posthumous shade of one who was once living, overlap or unclear usage is frequent. A clear example of conflicting terminology reflecting the apparition's ambiguous state occurs at Aeneid 2.772, where Aeneas reports seeing infelix simulacrum atque ipsius umbra Creusae (‘sad phantom and ghost of Creusa herself’). Is Creusa dead or alive? Aeneas (like Virgil) avoids clarifying.Footnote 46 It is characteristic of such apparitions that both their ontological status and their messages lie somewhere between truth and fiction, and a substantial number are actively deceitful.Footnote 47 Significantly, literary treatments of such apparitions frequently foreground questions of narrative reliability, a tendency noted already by Cicero and compared to the actual effect of dreams outside literature: haec, etiamsi ficta sunt a poeta, non absunt tamen a consuetudine somniorum (‘these examples, even though they have been fabricated by the poet, are not that far from the usual nature of dreams’, Div. 1.42).Footnote 48
Not only dreams but underworld encounters in general in ancient literature are fundamentally and overtly fictitious, and rest upon the inherently ambivalent act of narration.Footnote 49 Odysseus’ nekyia (Od. 11), the archetypal underworld encounter of the classical tradition, is notorious for the doubtful status of its truthfulness within the poem: it is among Odysseus’ various conflicting and in some cases overtly fictitious narratives within the poem, and its wider context, the narration to the Phaeacians (Od. 9–12), is especially marked by fantasy.Footnote 50 The Aeneid redoubles this narratorial instability: the katabasis of Book 6 is variously propelled, interrupted, extended, echoed and prematurely curtailed by a host of different narrators and interlocutors, with fluctuating degrees of reliability and forthrightness.Footnote 51 Although Aeneas’ visit to the underworld is ostensibly an uncovering of hidden knowledge—a ‘cheat sheet’ to eschatology, the nature of the universe, the mysteries of past, present and future—any revelations are in turn matched by reconcealments, praeteritiones and aporetic obscurity, and any epistemological certainty within the episode is disrupted on multiple levels.Footnote 52 Internally, the revelations granted to Aeneas, like the underworld's geography itself, are labyrinthine, riddling and partial:Footnote 53 Dido's magnificently disdainful silence (Aen. 6.469–74), for example, or the contradictions between Palinurus’ and the narrator's accounts of his death (5.835–71, 6.337–83). Such contradictions are frequent in post-Homeric underworld narratives, perhaps in imitation of the prominent inconsistency in Book 11 of the Odyssey whereby Odysseus begins on the surface of the earth (nekyia) but by the end is walking in the land of the dead (katabasis).Footnote 54 Crucially, in the Aeneid these contradictions also highlight the fact that the katabasis-narration is a retelling, a re-narration: for the first time, Virgil returns to events that have already been described in the poem, and the discrepancies (and distortions) between versions told in different voices illustrate the dependence of the story on the teller.Footnote 55
Particularly significant for this article is Virgil's characterisation of Aeneas’ katabasis not only as narrative—self-reflexive, self-conscious, self-critical—but also as dream. Aeneas departs from the underworld via the Ivory Gate of falsa insomnia (‘false dreams’, Aen. 6.896), and subsequently fails to recall any details of his katabasis. This peculiar feature has led several critics to suggest that the entire underworld sequence is a falsum insomnium, destabilising the entire episode's veracity within the poem and the reader's trust in the narrator with it.Footnote 56 Sergio Casali goes further: in his reading, the dreams’ falsity spills out into the rest of the poem, staining not only the underworld of Book 6 but the poem as a whole.Footnote 57 The characterisation of the underworld as a dream or apparition goes the other way, too: the ‘true dreams’ that slip through the Gate of Horn to the world above are referred to as umbrae, ‘shades’—or rather, ‘ghosts’ (6.894).
The gnat's apparition to the sleeping goatherd and its katabasis-narrative participate in and develop this tradition of epistemological, ontological and narratorial instability. It is particularly interesting that the gnat—a dream-apparition, and a dead shade in the underworld asking for burial—is promoted to become narrator of its own katabasis. This is almost unparalleled in ancient literature: just as dreams are reported to others by the one who experiences them, not by those who appear within them, so too are katabaseis typically narrated by the living protagonist who returns to the upper world, or by an omniscient narrator.Footnote 58
The gnat's narration, moreover, begins and ends with indications of the specifically fashioned or fictive nature of both the ghostly speaker and its speech. Before the gnat begins, the narrator introduces the ghost as an effigies…culicis (208). The term effigies, derived from effingo, frequently denotes something not only artificially created but spurious, imitative, ersatz—Andromache's pitiful faux-Xanthus (Aen. 3.497), for example, or Circe's incorporeal mirage of a boar (Ov. Met. 14.358f.).Footnote 59 The term is not inappropriate for a ghost or dream apparition, but connotations of deceit are not far from the surface.
Nearly two hundred lines later, the gnat closes its speech with something of a poetic cliché: at mea diffusas rapiantur dicta per auras (‘but as for my words, let them be carried off into the far-spreading breezes’, 384). Servius tells us that the image of ‘scattering to the winds’ is a standard idiom for paying no attention to something, which is clearly the primary sense here: the pessimistic gnat concludes that its words are in vain and its plea will go unanswered.Footnote 60 A few lines before, it characterised the goatherd's imagined response in similar terms: haec immemor audis | et tamen, ut uadis, dimittes omnia uentis (‘you hear these words, unmindful, and still as you go you will abandon everything to the winds’, 379f.).Footnote 61 But there are several further points of interest in the gnat's closing words. First, a minor point: the gnat is immediately proven wrong, as the goatherd is deeply affected by the gnat's narration and straightaway sets about making amends. Far from being immemor (379), he is emphatically memor, even painstakingly so (memor, 394; assiduae curae memor, 398). That the gnat's prediction turns out wrong does not, of course, mean that it is an unreliable or deliberately mendacious narrator: its despondency is a perfectly legitimate reaction to its ill-treatment and suffering, and it is a pleasant surprise that the goatherd does in fact see his error and react accordingly. But if Virgil's underworld narrations have taught us anything, it is that discrepancies between a katabasis or dream and the external narrative are worth further attention.
More significantly, the gnat has chosen here a phrase with a rich and suggestive poetic heritage. The formula diffusas rapi per auras goes back to Homer as a striking image of evanescence and immateriality, particularly applied to phantoms and shades of the dead. The prototypical example occurs in Odysseus' nekyia, when he thrice tries to embrace the shade of his mother, but each time her immaterial ghost slips away ‘like a shadow or in a dream’ (σκιῇ εἴκελον ἢ καὶ ὀνείρῳ, Od. 11.208).Footnote 62 In this epic context the formula in itself does not provoke doubt as to the phantom's ontological status; by the time it reaches Virgil, however, some more complexities have accrued from its reuse in philosophical writing. Plato's Cebes uses the image to express the impossibility of life or consciousness after death, when the soul dissipates ‘like breath or smoke’ (Phd. 70a: ὥσπερ πνεῦμα ἢ καπνóς), and Lucretius takes up the same image with gusto, using it to illustrate the disappearance of souls after death, not to the underworld but out of existence altogether: et nebula ac fumus quoniam discedit in auras (‘since mist and smoke disperse in the breezes’, 3.436), or dissolui…omnem animai | naturam ceu fumus in altas aeris auras (‘the whole nature of the soul is dissolved, like smoke, in the uppermost breezes of the aether’, 455f.).
Virgil repeatedly uses the motif in his protagonists’ sorrowful encounters with phantoms of the dead, rehabilitating it from its appropriation in philosophy to its original epic context: Eurydice (G. 4.498–502), Creusa (Aen. 2.790) and Anchises twice (Aen. 5.740, 6.700–2). But despite the Homeric modelling of these scenes, in each case the phrasing is markedly Lucretian,Footnote 63 undercutting the scenes’ sentimentality and (in the case of the Aeneid) the legitimacy of the phantoms’ advice with an intertextual paradox: how can these phantoms be genuine, when they are described in terms that elsewhere indicate the impossibility of any form of posthumous existence?Footnote 64 As ever in Virgil's engagement with Lucretius, ‘to repress by giving voice to the repressed necessarily lays the groundwork for its unheimlich return.’Footnote 65 Virgil's systematic ‘remythologising’ of Lucretius, restoring passages such as these to epic contexts, always runs the risk of the reverse process, ‘demythologising’ the Aeneid.
With this in mind, it is striking that the gnat uses this motif not only of its speech, but of its own nature. The gnat's words are to be ‘swept away through the far-spreading breezes’ (diffusas rapiantur per auras, 383); it narrates its own death and arrival in the underworld in synonymous terms, rapior per inania uentis (‘I am carried on the winds through empty space’, 212). The parallel of speaker and speech, author and work, is standard: after the gnat finishes speaking the narrator immediately recapitulates the same conceit, commenting dixit et extrema tristis cum uoce recessit (‘it spoke, and sadly drew back along with its last words’, 384).Footnote 66 But the precise motif used to create this parallel is uncomfortable. Much as the gnat encourages its audience to put credence in its actual bodily experience of the underworld—note mea uiscera, ‘my vitals’, in the next line (215)—the insubstantial, inconsequential, even ontologically impossible characterisation of the speech carries over to the gnat itself.
Indeed, the description of the gnat's descent to the underworld, rapior per inania uentis, seems to invoke another element of Lucretius’ denial of the traditional conception of death and the underworld.Footnote 67 There is very little explicit discussion in ancient literature of how the souls of the dead were supposed to reach the underworld,Footnote 68 and therefore very few parallels against which to measure the Culex’s description.Footnote 69 It is striking, then, that not only does per inania carry specifically Lucretian connotations,Footnote 70 but that a phrase very similar to rapior per inania uentis appears in Lucretius during a remarkably apposite discussion of commonly held suppositions regarding the underworld: [aues] uacuum proper iam per inane…dispergunt animas (‘[birds] disperse their souls through this almost empty space’, DRN 6.838f.). Lucretius here addresses the common etymological topos regarding Auernus, namely that no winged creatures can fly over it: as if Auernus were derived from Ἄ-ορνος and cognate with auis.Footnote 71 He offers two scientific explanations in place of the myth that Avernus is the gateway to hell: either the noxious exhalations from the lake poison the birds, or the movement of the vapours displaces the air and creates a vacuum, through which the birds fall as their anima escapes through their pores (DRN 6.738–839). Either way, though, the lake is just a lake. There is no heaven and no hell, for Lucretius, and neither is lying under Avernus.
Virgil nods to Lucretius’ account in his narration of Aeneas’ approach to the underworld, substituting the Lucretian etymology for the suppressed name Auernus: quam super haud ullae poterant impune uolantes | tendere iter pinnis (‘over which no flying creatures could wing their way in safety’, Aen. 6.239f.).Footnote 72 As at the exit, so at the entrance: Virgil's framing of Aeneas’ katabasis destabilises it, leaving the reader unsure of the truth of the episode. Did Aeneas visit the underworld?—or did he merely dream that he did?—or is it purely allegorical? The Culex’s phrasing, in turn, recontextualises the Lucretian-Virgilian trope and redoubles its self-contradictions. Lucretius says there is no underworld and no flying creatures can fly over Avernus; Virgil says that Aeneas visits the underworld, but no flying creatures (haud ullae…uolantes) can fly over Avernus; the gnat, however, a flying creature par excellence, says that it passed through the Lucretian vacuum (per inania) and visited the Virgilian underworld itself. As in the Aeneid, the Culex’s narrative is in direct conflict with its allusive gestures; the reader is left in doubt as to the degree of reality or reliability in this episode.
Fluid Geographies and Watery Underworlds
Let us put aside the question of the ghostly gnat's ontological status and turn to the katabasis-narrative itself. It rapidly emerges that, as a narrator, the gnat leaves much to be desired. Despite its didactic assertions of autopsy in pre-emptive defence of its report's reliability (uidi, 216, 227; cerno, 259),Footnote 73 certain aspects of its narrative seem even deliberately confusing. The underworld's topography is especially hard to envisage, even in comparison with Virgil's renownedly unmappable underworld: though at times it seems vastly simplified,Footnote 74 consisting solely of Tartarus and Elysium with none of Virgil's subdivisions, as the speech progresses the underworld begins to seem ever more recursive and tangled. The gnat—who somehow gains entrance to the entire underworld though unburied and not yet judged (374–7)—passes from Tartarus (216–58) to Elysium (258–371) and then (back?) to Dis (372–7), a route unparalleled in ancient daytrips to Hades.Footnote 75
Admittedly, ancient underworlds not only vary from text to text but are necessarily murky, confusing and characterised by spatial fluidity:Footnote 76 within texts such as Odyssey 11, Plato's Myth of Er and Aeneid 6, their topographies are an eschatological secret revealed only partially and to a select few. Even considering this generic lack of fixity, though, the Culex-author hijacks the ontological and epistemological uncertainty of the underworld—most importantly, of course, the Virgilian underworld—and marks it in specifically narratorial ways, deploying it as comment on the issue of reliable authorship and authenticity. For example, there is a sense of both geographical and temporal disorientation in the Culex’s underworld, especially in the narration of the Trojan War (304–57) prompted by the sight of the heroes in Elysium (sede piorum, 295).Footnote 77 It is unclear whether the tale is simply a discursive aside on the gnat's part or if, in a psychological turn, the heroes are themselves remembering their deeds (and, if the latter, how the gnat discerns this): the gnat regularly shifts tense, sometimes employing the historic present and sometimes the past, and also oscillates between narrating the literary past and describing the heroes as it sees them in the underworld. A particularly ambiguous moment occurs when the gnat is describing Odysseus:
The son of Laertes keeps his face turned away from him [Ajax]. And now, as victor over Strymonian Rhesus and Dolon, congratulating himself over the Palladium, he rejoices, and then again trembles. He, formidable though he is, shudders now at the Cicones, now at the Laestrygonians. Ravenous Scylla, surrounded by Molossian hounds—the Cyclops of Etna—fearsome Charybdis—the gloomy lakes and dismal Tartarus—all terrify him.
The last line is misleading. In the context of the wider katabasis, it is natural to read it as a description of Odysseus’ current state (as at 327), as if the gnat sees him frightened by his underworld surroundings.Footnote 78 But Odysseus is in the Elysian Fields, not pallentesque lacus et squalida Tartara, as we are reminded soon after: hic alii resident pariles uirtutis honore | heroes (‘here abide others of equal reputation for valour, all heroes’, 358f., cf. sede piorum, 295). Line 333 is therefore a reference to Odysseus’ earlier experience of the underworld, a continuation of his adventures in the Odyssey as narrated here by the gnat (330–3). The present-tense terrent muddies the waters: is it a historic present, referring to Odysseus’ emotional state in his nekyia? Or is there a transference of emotion, so that traumatised Odysseus, lost in the memory of his earlier unpleasant underworld experience, believes himself there again?Footnote 79 If Odysseus confuses his imagined and actual surroundings, though, so does the gnat, so that we are uncertain (at least momentarily) precisely where its narration has taken us. Indeed, the Culex here reverses the roles taken by Ajax and Odysseus in the Homeric underworld, where it is Ajax who silently turns away from Odysseus (Od. 11.563–7).Footnote 80 Is this truly a reversal?—or has the gnat betrayed, once again, its imperfect recall of what it saw in the underworld?Footnote 81
One further aspect of the katabasis’ topography again flags up the dubious reliability of the gnat's narration, this time resting on another kind of fluidity. As in previous underworlds, the principal divisions of the underworld are marked by bodies of water: Lethaeas…per undas (‘through Lethaean waves’, 215), ad Stygias…aquas (‘to the Stygian waters’, 240), Elysiam…ad undam (‘to the Elysian water’, 260), Ditis opacos | …lacus (‘murky lakes of Dis’, 372f.), uastum Phlegethonta (‘huge Phlegethon’, 374).Footnote 82 The gnat's narration, though, seems preoccupied with water above and beyond the traditional wateriness of the underworld:Footnote 83 as well as the underworld rivers and marshes it passes, it also tells of Tantalus in the waters (240–2), Orpheus’ rivers (278), Simois and Xanthus (307), the Sigea or Rhoetea litora (‘Sigean or Rhoetean shores’, 307f., 313), a vivid river simile (318–20), Odysseus’ previous visit to these pallentes…lacus (333), the Hellespont's waves (338) and a shipwreck (344–57). Virgil's Sibyl speaks of ‘swimming across’ underworld rivers (transnauimus amnes, Aen. 6.671)—but the gnat does it twice (transnare, 215; tranandus, 260). The gnat's parting wish for the goatherd's continued life and happiness prioritises water (fontes, ‘springs’, 381) over other conventional elements of a pastoral idyll (uiridis nemorum siluas et pascua, ‘leafy forest groves and pastures’, 382), and contrasts this imagined scene of the goatherd settled among sacred springs (tu cole fontes, ‘you dwell among [or ‘worship’] the springs’, 381) with that of the gnat in murky subterranean waters (ego Ditis opacos | cogor adire lacus, ‘I am forced to go to the dark lakes of Dis’, 372f.).
The Culex’s first description of the gnat is umoris…alumnus (183), sentimentally rendered ‘little Noursling of the humid Air’ by SpenserFootnote 84 but in ancient zoology a strictly literal description of the origin of gnats, born from acescente umore (‘brackish water’).Footnote 85 Like Aeneas’ visit to Anchises in the Virgilian underworld, the gnat stages a return to the generative waters from which it originated.Footnote 86 What's more, its watery interests and haphazardly zig-zagging motion across the underworld's terrainFootnote 87 remind us that, despite the seemingly human underworld it traverses, it is something other than human. The Culex-poet's self-comparison to a spider in the second line of the poem (ut araneoli) has primed us for insect-like singers; the carmen culicis brings us a narrator who is an insect, on more than the level of simile alone.Footnote 88 Its narration betrays its different, limited and biased perspective, calling into question the seeming omniscience of our katabasis-narrator.Footnote 89 The gnat's narrative is unreliable, not just through a grandiose poetic and philosophical tradition of unreliable narration, but because it is a gnat. The poem holds so fast to its fantastical premise as to highlight its surreal nature in the first place.Footnote 90
Indeed, perhaps we, along with the goatherd, ought to shake ourselves out of sleepy acquiescence to this fiction of a talking, thinking, underworld-traversing gnat. The most frequent complaint regarding gnats, both in antiquity and today, does not concern their necrological abilities but rather the fact that they ‘produce a high-pitched whine or buzz or hum which can keep a restless individual awake for hours’:Footnote 91 as insomniac Horace complains, ‘the damned gnats drive sleep away’ (mali culices…auertunt somnos, Sat. 1.5.14f.).Footnote 92 The goatherd's interrupted sleep may have an altogether more prosaic cause.
Conclusion
I hesitate to claim outright that the gnat's effigies is necessarily a falsum insomnium, or that it is lying about its katabasis. Much of the humour of the poem derives from the contrast between the self-evident surrealism of the situation (gnats do not, as a general rule, perform heroic deeds, let alone posthumously inform the beneficiaries of their actions about poetic eschatologies) and the seriousness with which it is developed. The bathetic pathos of the goatherd's response would be lost, were the katabasis obviously deceitful or imagined. Yet the poem highlights and develops certain established literary tropes of specifically narratorial uncertainty and dishonesty: its allusive tactics betray a certain dissimulating relationship with the master-text of Virgil's canonical works; doubt is raised regarding the identity of its embedded narrator and how that identity affects the tale told; and the heart of the poem is a katabasis-narrative, a prime conduit for discussions of truth, fiction and inauthenticity in the ancient world. As the Culex both professes Virgilian authorship and self-consciously undermines that veneer of authenticity with regularly placed inconsistencies and metapoetic admissions of impersonation, so too the culicis carmen simultaneously asserts and undermines the reliability of its narrative and authorship.
