It is well known that Hellenistic and Roman poets playfully hid signatures in their works and that ancient readers in turn were conditioned to look out for them.Footnote ¹ Nicander (Ther. 345–53) and Baebius Italicus (Ilias Latina 1–8), for example, spell their names in acrostics, and Cicero (Diu. 2.111–12) relates that Ennius did the same. Modern scholars have identified subtler signatures in acrostics, puns, acronyms and other cryptograms in authors such as Aratus, Lucretius, Propertius, Virgil and Ovid.

These readings have met with varying degrees of acceptance and resistance, owing to the difficulty of distinguishing the intentional from the incidental. Discussing acrostics in general, Cristiano Castelletti provides criteria for more secure identification: a meaningful relation between the acrostic and its context, signposts placed nearby to get readers’ attention, and allusions that embed the acrostic in a tradition of such wordplay.Footnote ² When signatures take the form of cryptograms other than acrostics, they are less likely to jump off the page, and identification is harder. For these cases Matthew Robinson notes the additional importance of position: ‘These signatures are often found at key structural moments in a text’, especially beginnings and endings.Footnote ³

In another important article, Robinson problematizes the notion of authorial intention inherent in many discussions of acrostics.Footnote ⁴ He argues for a reader-centred approach that comes closer to the hermeneutics of allusion as it is now practised in classical scholarship. That is, since we recognize that ancient readers were on the lookout for acrostics, we can profitably examine how such acrostics enrich a reading experience without demanding indications of authorial intention. Robinson recognizes none the less that the impression of authorial intention was important to these same ancient readers and that signposts could ‘reassure readers of the significance of the acrostic’.Footnote ⁵

With such principles in mind, I would like to consider a possible authorial signature in the opening lines of Lucan’s Bellum Ciuile, a prime location for readers in search of the author’s name (Marcus Annaeus Lucanus). The signature is arranged so as to allow for reading in a few different ways, each beginning with an abbreviated form of Lucan’s name. Signposts and references to key texts in the history of poetic cryptograms invite readers steeped in this tradition to look for the signature, letting their eyes roam across the field of letters to consider the various possibilities. Conversely, these same signposts and references may provide confirmation to readers who have found it.

First, let us examine the following intext, or mesostic, whose central position and straightforward message help it stand out:Footnote ⁶

Hidden in adjacent letters across three lines, connected in the style of a modern ‘word search’, is the sentence LVCAN(us) SVM (‘I am Lucan’), or LVCANI SVM (‘I am Lucan’s’). The latter sentence strikingly parallels owners’ marks on objects such as drinking ware, speaking inscriptions consisting of the owner’s name in the genitive followed by sum and sometimes accompanied by a warning to thieves—not an inapposite association for an authorial signature, which could similarly serve to protect a text against plagiarism.Footnote ⁷ At the same time, we may find in these connected letters the syllabic acronym LVC(anus) AN(naeus)—or, without the C, LV(canus) AN(naeus), for when poets’ signatures take the form of syllabic acronyms, nomen and cognomen generally divide cleanly into separate lines of verse.Footnote ⁸ The letters of Lucan’s name hidden in the first two lines could stand out more to readers because of their emphatic metrical position: the words pLVs and CANimus fall immediately after the caesurae of their respective lines, beginning the lines’ second halves.Footnote ⁹ Moreover, in intersecting with CANimus, the signature is appropriately anchored in Lucan’s first reference to himself and the act of poetic composition.

Depending on the script, this signature might appear on a diagonal, which would not be unprecedented: Mathias Hanses has identified a diagonal intext in Aratus’ Phaenomena, in a passage (Phaen. 778–87) which also contains a gamma acrostic of the same word (ΛE;ΠΤΗ) and which became a touchstone for poets playing with cryptograms.Footnote ¹⁰ After Lucan, Valerius Flaccus incorporates a diagonal signature into his imitation of this Aratus passage, accompanied by a diagonal intext NOMINI in the same lines (2.368–72).Footnote ¹¹ Later still, diagonal intexts are among the many textual figures in Optatian.Footnote ¹²

The famous passage of Aratus and its ancient reception may have additional relevance for Lucan’s signature. Those lines of the Phaenomena concern the light of the moon, and Leah Kronenberg shows that Lucretius alludes to Aratus by incorporating a LVCE acrostic into his own discussion of the moon (5.712–15).Footnote ¹³ As well as reflecting the content of the passage, Lucretius’ acrostic is a window reference to Aratus’ model, the Homeric acrostic ΛE;ΥΚΗ (Il. 24.1–5), and a play on his own name, LVCretius. In the latter capacity, it anticipates a syllabic, backward acronym of his name incorporated into the text a few lines later: CAndenti LVmine TInctus (5.721) = CA(rus) LV(cretius) TI(tus). By hiding an acrostic and a signature in the passage, Lucretius pays double homage to Aratus, who had hidden his signature at the beginning of the Phaenomena, punning on the word ἄρρητον (‘unspoken’, Phaen. 2).Footnote ¹⁴ Lucretius leverages the form of his name, with its resemblance to words for light and to the original Homeric acrostic, to connect these two famous instances of Aratean wordplay. When Lucan then signs his own name, also beginning with LVC, in a hidden intext, he inserts himself into this particular tradition of cryptograms and name play. Ancient etymologists did, in fact, connect the name Lucanus with lux, and both Statius and Martial pun on the resemblance in poems about Lucan.Footnote ¹⁵

If we interpret the first letters of the cryptogram as LV(canus) AN(naeus), Lucan places his cognomen before his nomen, adopting the same reverse order as we have seen in Lucretius, an order used also by Virgil in several of his signatures. In these comparanda, we find signposts marking the inversion of naming conventions. To my knowledge, Lucretius’ signposts have escaped scholarly attention, but they function in a similar fashion to those which critics have noticed in Virgil. Lucretius embeds his reversed tria nomina in a discussion of the revolutions of the moon, with several mentions of ‘turning around’, all prominently placed at the beginning of lines: uoluier (5.716), uersarique potest (5.720), uersandoque (5.722). The signature, with its component parts ‘turned around’, falls in the line between these last two. The verb uertere performs a similar function in Virgil’s signature at G. 1.1–2. He hides MAR(o) VER(gilius) in the phrase terRAM | VERtere, with the cognomen spelled backwards and preceding the nomen.Footnote ¹⁶ This signature involves a host of playful verbal effects based on Aratus’ signature at Phaen. 2, but for our purposes suffice it to say that uertere can be read as a signpost of the inverted order of Virgil’s names.Footnote ¹⁷ Later in the book, in his imitation of Aratus’ moonlight passage (G. 1.425–37), Virgil again signs his name in reverse order, this time in an alternate-line acrostic MA(ro) VE(rgilius) PV(blius).Footnote ¹⁸ Again, he signals the inversion, in this case with the phrase reuertentis cum primum (1.427).Footnote ¹⁹

It is worth noting, then, that Lucan’s signature connects with the word conuersum, ‘turned around’. Cognate with Lucretius’ uersari and uersando and with Virgil’s uertere and reuertentis, the word indicates to readers that Lucan’s names can be read in backwards order and marks his intertextual engagement with Lucretius’ and Virgil’s similar play. Earlier, I took SVM as part of Lucan’s signature, but we might also construe the M as his praenomen, so that we get LV(canus) AN(naeus) M(arcus):

Read this way, Lucan’s signature anticipates Valerius Flaccus’ diagonal signature FL(accus) VA(lerius) CA(ius), which similarly intersects on its final line with a signpost indicating its reverse order, the word reuoCAntes (2.372).Footnote ²⁰

Expanding outward on the page from the LVCANI SVM or LV(canus) AN(naeus) M(arcus) intexts, one finds that they intersect with another potential signature, similarly arranged diagonally in a stair-step pattern:

The message reads LVCANI M. VIS CERA DEXTRA, ‘the power of M. Lucanus with his right hand on the wax’, with cera, as often, referring to the wax tablet used for poetic composition (OLD s.v. 4). Meanwhile, the message can also be rendered ‘the power of M. Lucanus on the wax on the right-hand side’, which describes the position of the signature: it occupies the right side of the page, starting at the centre and moving to its right margin. (The other variations on the signature discussed in this article are similarly located on the right side of the text.) In other words, dextra may serve as a signpost, directing readers’ attention to the presence and position of the message, much as we saw with conuersum.

There are comparanda for the use of dextra to show readers where to look for a cryptogram. Castelletti has argued that Aratus uses δεξιά (Phaen. 6) and ἀρίστη (Phaen. 7, cf. ἀριστερά) to indicate in which direction to read a boustrophedon acrostic, starting from the right and then proceeding from the left.Footnote ²¹ In a passage of the Aeneid (6.641–57) that features a previously identified syllabic acrostic signature,Footnote ²² I propose that Virgil indicates the presence of an additional signature as follows: conspicit ecce alios dextra laeuaque, ‘he looks—behold!—at others, to the right and the left’ (Aen. 6.656). If we look to the right- and left-hand side of that line and the next, lo and behold, we find herbAM | VEscentis, another signature on the model of terRAM | VERtere at G. 1.1–2.Footnote ²³ A comparable instance of dextra appears in a later text, an acrotelestic inscription erected to celebrate the domain of a fourth-century North African lord, Sammac.Footnote ²⁴ The second line boasts that Sammac protects the place ‘on all sides with his right hand’ (undique dextra), a phrase with a possible double meaning signalling the presence of cryptograms on both sides of the text, including the less commonly used right edge. In that inscription, the word dextra is tellingly placed at the rightmost end of its verse, intersecting with the telestic, toward which it draws readers’ eyes.

Another signpost is worth mentioning in connection with Lucan’s CERA DEXTRA. In Aratus’ moonlight passage, he instructs the reader to look at the moon at each of its ‘horns’, the ends of its crescent: σκέπτεο δὲ πρῶτον κεράων ἑκάτερθε σελήνην (Phaen. 778). Observing that the word κέρας can refer more broadly to the edge or side of anything, Feeney and Nelis interpret Aratus’ instruction as a prompt to the reader to look at the ‘edges’ of the lines and find the acrostic.Footnote ²⁵ With slight variations, others have similarly read the word κέρας as a signpost of the acrostic play in the passage.Footnote ²⁶ In Virgil’s reworking of the passage in the Georgics, he places his acrostic signature between two instances of the word cornu, a nod to this use of κέρας in his model.Footnote ²⁷ I suggest that Lucan too picks up on this aspect of Aratus’ κέρας, incorporating it into his own signpost through a bilingual pun with cera (‘wax’).Footnote ²⁸ Where Aratus used the word to direct readers to the one side of his text, Lucan uses its sound-alike cera (≈ κέρᾳ), paired with dextra, to do the same. Notwithstanding the gender mismatch, the juxtaposition of these two words helps bring out the bilingual pun, since κέρας was regularly paired with adjectives for left (λαιόν) and right (δεξιόν) to describe the left and right wings of an army, just as its Latin equivalent cornu was regularly paired with sinistrum and dextrum to do the same (OLD s.v. 8).Footnote ²⁹ Lucan’s signature occupies the right-hand side of the wax, with a gesture toward the right ‘wing’ of the ranged letters. The echo of military terminology in this nod to Aratus’ horns of the moon underlines the transposition of cryptogram conventions from didactic to martial epic, a theme to which we will return below.Footnote ³⁰

In its reference to the materiality of wax-tablet composition, the signature LVCANI M. VIS CERA DEXTRA harkens back to another such cryptogram, the boustrophedon acrostic at Aen. 1–4 A STILO M. V. (‘from the stylus of M[aro] V[ergilius]’).Footnote ³¹ Lucan’s proem is, of course, closely engaged with that of the Aeneid in general.Footnote ³² To name just a few formal connections: his bella mirrors Virgil’s initial word arma; he too lays out the topic of the poem in a single sentence of exactly seven lines; and his canimus is preceded by two direct objects, like Virgil’s cano. Might his detail-oriented doubling of the first lines of the Aeneid have extended to Virgil’s signature? To my knowledge, Lucan’s and Virgil’s signatures would be the only two to mention instruments of wax-tablet writing. Castelletti suggests that Virgil’s stilus is reminiscent of a weapon and thus an appropriate emblem of his composition as he turns to martial epic.Footnote ³³ As author of an even bloodier epic, Lucan makes this conjunction of the violence of writing and the writing of violence more prominent, for the word uis is used both of vigour of expression (OLD s.v. 6d) and of martial violence. The reference to uis upon the wax assimilates the incision of Lucan’s signature on the tablet with the forceful style of his writing and the violent nature of its content. On the level of form too, Lucan’s signature bears a resemblance to Virgil’s: he similarly puts his name in the genitive case, partially abbreviated to M. and in reverse order. (Again, the nearby presence of conuersum could be taken as a signpost of this inversion in names.)

The intexts examined thus far do not exhaust the possibilities. Additional signatures ramify in different directions from the shared core of Lucan’s name, extending to other metapoetic complements that suit the proemial context. Beyond the mention of his power upon the wax, we find LVCANI VERSV (‘with the verse of Lucan’):

And LVCANI MVSIS (‘with the Muses of Lucan’):Footnote ³⁴

The latter message is appropriately related to the word in the surface text that forms its centrepiece, canimus. In epic, the Muses are typically presented as agents behind the song and invoked in the proem. Lucan does not mention them in the surface text, but the intext brings them in, connected with his singing through shared letters.Footnote ³⁵ These shared letters also include animus, which Lucan uses as a substitute for the inspiration of the Muses later in the proem, on the model of Ovid (fert animus causas tantarum expromere rerum, ‘my mind drives me to lay out the causes of such great matters’, Met. 1.67).Footnote ³⁶ Lucan’s name, his Muses, his animus and his epic song are thus interwoven in a dense cluster of overlapping words with proemial import—the agents of the song all extend from, or inhere in, the verb for singing. The play on Lucanus, canimus and Musae is the same as we see in Statius’ poem celebrating Lucan’s birthday: Lucanum canimus, fauete linguis; | uestra est ista dies, fauete, Musae … (‘We sing of Lucan. Let all hold their tongues! This day is yours, Muses; be favourable …’, Silu. 2.7.19–20).Footnote ³⁷ It is tempting to see Statius’ lines as reception of Lucan’s play on words between the intext and the surface text, but at the least it shows another poet seizing upon this potential in Lucan’s name.

My argument is one of accumulation. The fact that Lucan’s name appears in adjacent letters and can be supplemented by references to poetry (VERSV), the Muses (MVSIS) and the writing tablet (CERA) makes each possibility seem less fortuitous. On the composition end, one can imagine Lucan arranging this connected array of intexts intentionally, choosing words in the surface text that facilitate the reading of various hidden signatures. Regardless, on the receiving end, readers attuned to looking for such signatures could find a number of them in this prime location at the beginning of the poem, each initiated by Lucan’s centrally located name. The multiplicity of options suits Lucan’s tendency to outdo previous epics by exaggeration. This tendency is embodied in the word pLVs in the poem’s first line, from which the plurality of signatures all take their starting point.

One more signpost remains for us to consider. Among the markers with which authors signal their cryptograms, we unsurprisingly find terms for signs, σῆμα in Greek and signum in Latin. Aratus and Virgil cleverly make their references to heavenly signs in the Phaenomena and the Georgics double as pointers to the words hidden in those passages.Footnote ³⁸ When Virgil embeds his MA(ro) VE(rgilius) PV(blius) signature in his moonlight passage (G. 1.427–39), he first signposts the acrostic in more obscure ways as he describes the moon’s more obscure light. Then he retrospectively illuminates it more clearly when he turns to the topic of the sun, employing two forms of signum in quick succession: sol … | signa dabit; solem certissima signa sequentur (‘the sun will give signs; the surest signs will follow the sun’, 1.438–9).Footnote ³⁹ Damien Nelis sees Ovid following suit at Fast. 1.2, using a reference to sunken signs arising (lapsa … orta … signa) to signpost a hidden anagram of Aratus’ name in the line, so that it can ‘rise’ to the reader’s attention.Footnote ⁴⁰ I would add that a lengthy acrotelestic in the Metamorphoses (12.235–44), discussed by Gregor Damschen,Footnote ⁴¹ coincides with the description of an embossed krater, whose decorations stand out (signis exstantibus, 12.235) just as the letters of the acrotelestic stand out to readers at the ends of the lines of verse. A variation on this theme is found earlier in the Metamorphoses, in the Deucalion and Pyrrha episode: Julene Abad Del Vecchio shows that Ovid simultaneously signposts the telestic SOMATA (‘bodies’, Met. 1.406–11) and flags its resemblance to σήματα (‘signs’) when he writes that the nascent human bodies, still unclear (non manifesta, 1.404), are similar to statues (simillima signis, 1.406) that are not yet finished.Footnote ⁴² In all these examples in Virgil and Ovid, then, the word signa is accompanied by an adjective (certissima; orta; exstantibus; manifesta) that indicates the signs’ degree of clarity, with a secondary reference to the clarity of the cryptogram lying nearby.

I suggest that we see Lucan riff on this tradition in his description of battle standards ranged against battle standards (obuia signis | signa, 1.6–7), placed shortly after the signature in the same seven-line introduction of the poem’s theme. In addition to its significance in the surface text, obuia can mean ‘visible’ or ‘easily grasped’ (OLD s.v. 4), approximating the sense of the modifiers Virgil and Ovid use with their metatextual signa. The resemblance to Virgil’s use of signa in the Georgics is particularly close (Luc. 1.5–7):

Like Virgil, Lucan uses two forms of signum in close proximity, and he too makes the wordplay visually prominent by locating one of them at the beginning of a line (and, in his case, putting the other at a line-end). Virgil’s surest signs (certissima signa) are twice reflected: not just in Lucan’s ‘clear signs’ (obuia … | signa) but also in the certa that begins the word certatum, vertically aligned with the signa that follows two lines below. For ancient readers attuned to seeing forms of signum as indicators of cryptograms, the double use of the word at the extremes of consecutive lines is a cue to look at the surrounding context more carefully and spot the certa as well as the signature that lies above. The echo of the Georgics—as well as Aratus’ Phaenomena and Ovid’s Fasti—points up generic differences: the signa that double as signposts of cryptograms in didactic epic are the signs of heaven; in Lucan’s martial epic they become the battle standards. In a didactic context, it is natural to discuss the clarity of signs; Lucan cleverly transposes this convention to the world of martial epic by using a word for hostile engagement (obuia) that can also indicate intelligibility, much as his allusion to the Aratean horns of the moon is overlaid with a suggestion of the wings of an army.

In the context of an authorial signature, signum has another relevant meaning, as the term for the personalized wax seal by which writers marked documents as authentically theirs. Irene Peirano shows that poets mobilize the metaphor of the seal, and its attendant vocabulary of signum and signare, in liminal contexts, as they ‘seal off’ their books at the edges.Footnote ⁴³ With these uses in mind, we might see a particular double meaning in Lucan’s obuia signis | signa, describing the signature and its annotation: ‘authorial stamps, visible by means of signposts’. Drawing in the certa that is vertically linked to signa, this can be extended further: ‘reliable authorial stamps, visible by means of signposts’. Two-factor authentication, if you will. The connection between an etched (OLD s.v. signo 1) signature and the wax stamp that it recalls metaphorically is made closer by Lucan’s earlier evocation of the wax material of the tablet, inscribed by his right hand.

Two more potential meanings of the pregnant phrase obuia signis | signa are worth considering. We saw earlier that Lucan’s text is arranged in such a way that readers may see his signature in multiple forms, all starting with his name. On the page, then, signs ‘come to meet’ signs, signature intersects with signature. Or we might read the phrase as an intertextual annotation, given Lucan’s close engagement with models, especially Virgil’s: ‘signatures (Lucan’s) ranged against signatures (Virgil’s)’. The aggression in the surface meaning of the text may be relevant here. ‘Battle standards ranged against battle standards’ describes the mirroring inherent in civil war, as like faces off against like. Meanwhile, like faces off against like on an intertextual level as Lucan mirrors the formal structures of Virgil’s poem, against which he is about to do battle. The signature is one more aspect of this formal doubling with which Lucan competes with his canonical predecessor. I do not mean to privilege one of these readings over the others. Rather, the use of signa in Lucan’s models and in close proximity to his signature invites a range of metatextual and metaliterary interpretations for readers familiar with the conventions of such wordplay.

If we return to Castelletti’s criteria for identifying cryptograms with which we began, we can see that Lucan’s signature would qualify. It has a meaningful relationship to its context—in this case the commencement of a new work. It is signposted, by conuersum, by cera dextra, by plus and by the double use of signa that follows. And it engages intertextually with a specific tradition of such wordplay, alluding to a range of texts, not least to the two most famous signatures of Virgil, Lucan’s alter ego and foil.