Incontri di filologia classica 12 (2012-2013), 1-28 LUCA GRAVERINI Ovidian Graffiti: Love, Genre and Gender on a Wall in Pompeii. A New Study of CIL IV 5296 - CLE 950 The poem published by Augusto Mau in CIL IV Suppl. 2,5296 (= CLE 950 and Courtney 1995, 92) was scratched on the doorway of a house in Pompeii (regio IX, insula 9).1 It is a remarkable inscription for at least two reasons: it is one of the longest and most complex poetic graffiti found in Pompeii and, perhaps most importantly, it is apparently one of the few ancient documents that appear to preserve a female poetic voice. I will discuss these aspects and several others in the following pages; first of all, however, it will be necessary to closely consider the text, its problems and its interpretation. Below are the transcript and translation provided by Courtney 1995, followed by a reproduction of the original inscription from the CIL. o utinam liceat collo complexa tenere braciola et teneris | oscula ferre labellis. i nunc, uentis tua gaudia, pupula, crede. | crede mihi, leuis est natura uirorum. saepe ego cu(m) media | uigilare(m) perdita nocte haec mecum medita(n)s: «multos | Fortuna quos supstulit alte, hos modo proiectos subito | praecipitesque premit; sic Venus ut subito coiunxit | corpora amantum, diuidit lux, et se | paries quid ama. 5 10 Would that I might hold my (your) arms embraced around your (my) neck and give kisses with my tender lips. Go now, poppet, and entrust your joys to the winds. Believe me, men’s nature is fickle. When in my desperation I was lying awake in the middle of the night, often, thinking over things with myself, (I said) «Many whom Fortune has raised aloft, these she subsequently oppresses, suddenly hurled down headlong. Similarly after Venus has suddenly united the bodies of lovers, daylight separates them. The poem contains several ambiguities and obscure points, and is evidently incomplete since it lacks the ending; although nothing hints that the beginning might be missing as well, there is no proof to the contrary either. The metrical structure is irregular: only vv. 1, 5, and 8 are well-formed hexameters. 1 On the precise location of the graffito see below, section d. -1- LUCA GRAVERINI Fig. 1 CIL 5296 I will start with a close reading of the graffito and some comments on its literary texture, postponing the analysis of the metrical problems to section c. I assume from the beginning that the poem’s speaker is a woman who is addressing another woman; this claim will be substantiated during the discussion. a) Reading the text 1. Both the beginning and the ending of the hexameter use common expressions in elegiac poetry, although it is not possible to identify specific models2. As we will see in what follows, the whole poem – although it is not exactly a cento, as Mau suggested in the CIL3 – makes intensive use of elegiac topoi and standard phrases, and of specific quotations from and allusions to the main authors of the Latin literary tradition. 2. Courtney 1995 and others print label(l)is following Mau’s transcript in the CIL, but his supplementum at p. 705 rightly points out that «labellis scriptum est». The only other literary occurrence of the diminutive brac(h)iolum is in Catullus 2 The pathetic verse opening o utinam is frequently found especially in elegiac poets, although never at the very beginning of a poem (5 times Propertius, 4 Ovid’s Amores, 4 Heroides, 1 each Fasti and Tristia, 1 Tibullus; also 3 times in Ovid’s Metamorphoses, 4 Lucan). The phrase complexa tenere also has precedents in love poetry, starting with Catullus 11,18 quos simul complexa tenet trecentos; similar hexameter endings occur in Ovid, am. I 13,39 complexa teneres and III 11,11 complexa tenebas; Statius, Theb. XII 337 and Ach. I 253 complexa tenerem; Silius Italicus VI 132 amplexa tenebat; Ciris 309 amplexa tenebo. 3 «Est scilicet carmen centonis vicem e variis quae poetriae mentem subibant fragmentis consutum, omissis additisque nonnulis ut panniculi coirent». -2- OVIDIAN GRAFFITI 61,174. This is «no doubt» the model for our inscription according to Courtney 1995 and Copley 1939, 342; Goold 1998, 24-25 displays as much certitude in rejecting any direct Catullan influence and assuming that «by the year 79 his [Catullus’] poems had ceased to be read by the literate». Both statements are unwarranted: Catullus was certainly still read in the Silver age4, but it is far from certain that this single, de-contextualized word can be considered a quotation from or an allusion to his poem since we must also allow for a possible popular and familiar circulation of the term5. The phrase collo complexa tenere braciola is slightly ambiguous since it is not clear if the speaker wants to have her arms around her friend’s neck, or to have her friend’s arms around her own neck. Although the former interpretation is often chosen6, there is a good reason to prefer the latter: affective diminutives like braciola are in fact more naturally used of the addressee than of the speaker7. This slightly confusing phrase probably results from the conflation of two different expressions, both commonly used in love poetry to describe a hug: complexum tenere aliquem8, and another combining brachia and collo like Ovid, am. III 7,7-8 subiecit eburnea collo / bracchia9. Teneris… labellis is sometimes interpreted as ablative10, but the standard phrase oscula ferre usually governs a dative11. As with braciola above, it is also more likely that the 4 For a list of references to Catullus in ancient authors, including of course the Silver Age, see Wiseman 1985; on Catullus and Martial, see e.g. Lorenz 2007; Gaisser 2009, 168-174. It is true, however, that he is not among the most frequently exploited models in Pompeian inscriptions: for a discussion of some possible (and tenuous) Catullan influences see Gigante 1979, 158-159, with further literature. 5 It is just the diminutive of the very common noun brachium, and its rarity might be due to its mainly familiar and non-literary usage. Brachiolum was also used as a technical term by Vegetius and the Mulomedicina Chironis (see ThLL s.v.). 6 It is Courtney’s first choice, and the only one e.g. in Goold 1998, 19. 7 Cf. the discussion of teneris... labellis below. The very similar expression in Statius, sil. V 4,15-16 aliquis… puellae brachia nexa tenens might have the same meaning if puellae is a genitive depending on brachia, as the word order suggests, rather than a dative depending on nexa. The verb complector is usually transitive, but it governs the dative collo in our inscription. Acording to the ThLL (s.v. 2085, 39-44), this is the earliest of a very short list of examples of this construction (not surprising per se: the ThLL itself, s.v. 2081, 72-73 mentions the analogy with verbs like circumdare, circumfundere). Cf. also Plautus, rud. 1203 uxor complexa collo retinet filiam, where collo could be either dative or ablative according to the ThLL (2082, 11). 8 Cf. the passages quoted in n. 2. 9 Cf. also her. 5,48 ut tua sunt collo bracchia nexa meo; met. I 762 dixit et inplicuit materno bracchia collo; Catullus 64,332 leuia substernens robusto bracchia collo. 10 Courtney 1995, 99: ‘with my tender lips’; Goold 1998, 19: ‘with tender lips’. 11 Cf. Ovid, ars III 310 oscula ferre umero; pont. I 4,49 oscula ferre comis; Propertius II 18b,18 canae… oscula ferre comae; Tibullus I 9,78 aliis... oscula ferre. -3- LUCA GRAVERINI addressee’s and not the speaker’s lips are described as ‘tender’ and ‘small’12: all in all, there can be little doubt that a kiss on the mouth is described here. Both diminutives braciola and labellis can also be considered as a first hint that the addressee of these verses is a woman. Of course, diminutives in general can be used for men as well as for women, especially in erotic contexts; however, brachiolum is said of the arm of a woman in Catullus, and in description of kisses it is not uncommon to see labra and labella used respectively for the man and the woman13. This first impression about the addressee’s female gender will be confirmed by pupula in the next line. 3-4. i nunc (possibly followed by et: see below, section c) is another common beginning for an elegiac verse14. For the image of «trusting the winds with one’s happiness» cf. e.g. Vergil, Aen. X 652 nec ferre uidet sua gaudia uentos15. The vocative pupula clearly shows that the poem’s speaker is a woman (see below on v. 5) addressing another woman. In an implicit attempt to establish a more ‘normal’ gender relationship for what is usually viewed as a love poem (see the discussion below, section d), some scholars interpret these two verses as a soliloquy of the speaking woman16, while a man would be addressed in vv. 1-2. However, there is absolutely nothing in the text to suggest a change of addressee, and crede mihi at v. 4 makes it extremely unlikely that the speaker is addressing herself17. 12 Cf. above on braciola. Of course there can be exceptions, such as Catullus 8,18 cui labella mordebis and Propertius II 13,29 osculaque in gelidis pones suprema labellis where the poets use labella/labellis for themselves: however, in both cases the poet is self-commiserating and/ or adopting the perspective of his beloved one. See also the examples provided in the following footnote. I am commenting on the graffito here, but it is possible that the poem originally had labris and not labellis: see below, section c. 13 See e.g. Plautus, Mil. 1335 labra ab labellis aufer, nauta; Ps. 1259 nam ubi amans complexust amantem, ubi ad labra labella adiungit; Lucilius, sat. VIII 303 f. labra labellis... compono. The collocation teneris labellis also occurs in Plautus, Ps. 67 and Ovid, ars I 667. Again, note that the original text of the poem might have had labris (which bears no clear gender characterization) and not labellis: see above, n. 12. 14 Cf. e.g. Ovid, her. 3,26 and 17,57. 15 So Courtney 1995; further references in Harrison 1991 on Vergil, Aen. X 652 and Nisbet - Hubbard 1970 on Horace, carm. I 26,2. 16 An hypothesis by Baldi 1982 mentioned by Varone 1994, 100; cf. Buonopane 2009, 242. 17 Marco Fucecchi suggests to me that vv. 1-4 might all be interpreted as a soliloquy: vv. 1-2 would quote the sweet words pronounced by an undependable man, and vv. 2-4 would be a meditation by the speaking woman on men’s unreliability addressed to a generic ‘you’. While this interpretation is more likely than the other, it also heavily relies on a change of speaking mode (quotation in vv. 1-2, meditation in vv. 3-4) that would be very difficult to grasp by a reader in the total absence of textual hints - and, of course, of quotation marks. The considerations on -4- OVIDIAN GRAFFITI Pupula, in the meaning of ‘poppet, darling’ and not ‘pupil of the eye’, is rare but certainly not a «clear instance of Catullian practice» as Copley states18 since one should allow for the possibility that it is simply an instance of sermo familiaris19. Of course, the idea of a lover’s unreliability is very common in love poetry, and many parallels could be mentioned. However, we should note that woman almost always bears the stigma of being uarium et mutabile (Vergil, Aen. IV 569), rather than man – an inevitable consequence of the fact that almost all the poets we know of were men. This makes Copley’s usual claim of a Catullan influence more legitimate in this case: in Catullus 64, in fact, Ariadne has good reasons to lament Theseus’ unreliability, as she does at vv. 139-144: at non haec quondam blanda promissa dedisti uoce mihi, non haec miseram sperare iubebas, sed conubia laeta, sed optatos hymenaeos; quae cuncta aerii discerpunt irrita uenti. nunc iam nulla uiro iuranti femina credat, nulla uiri speret sermones esse fideles But these were not the promises you gave me idly of old, this was not what you bade me hope for, but the blithe bride-bed, hymenaeal happiness: all empty air, blown away by the breezes. Now, now, let no woman give credence to man’s oath, let none hope for faithful vows from mankind There are no clear verbal parallelisms with our graffito, but at vv. 142-143 the sequence of thoughts is remarkably similar: the image of the winds blowing away the hopes of a betrayed woman is closely followed by a statement of the unreliability of men20. the fickleness of Fortune and Love at vv. 6 ff., for example, are clearly marked as a soliloquy by meditans (for a discussion of a possible textual problem there see below, section c). Considering all this, the ‘plain’ interpretation of vv. 1-4 as an address to a female ‘you’ seems much more likely to me, despite the gender-related issues that will be discussed below, section d. 18 1939, 345-346; cf. Catullus 56,5 pupulum. 19 Cf. Apuleius, met. VI 16,3 mea pupula with GCA 2004, 488 ad l.; Callebat 1968, 373. It also occurs in CIL X 6009 (= CLE 56), and it is not unparalleled as a woman’s name (CIL III 804; VI 2760; XIII 5836; of a man in XV 953). Pupa for ‘girl’ is attested both in literature (e.g. Martial IV 20,1-2) and in Pompeian inscriptions (CIL IV 6842 = CLE 2057; CIL IV 1234 = CLE 232). 20 Copley 1939, 344-345. At p. 343 f. he also highlights the correspondence of v. 3 with Catullus 65,17-18 ne tua dicta uagis nequiquam credita uentis / effluxisse meo forte putes animo. However, «trusting the winds with something» (and not specifically joy) is a common expression and an easy metaphor: cf. Manilius I 77 uentis credere uota; Ovid, fast. II 453; Tibullus I -5- LUCA GRAVERINI The remark on the leuitas of men’s nature21 can also be considered a clue that the speaker is a woman (cf. below on perdita): a man would hardly be interested in pointing that out to a woman. 5. The sentence is incomplete, and lacks a main verb (see section c for a textual analysis). It marks a change of subject, from the leuitas of men (vv. 3-4) to a remark on the instability of Fortune (vv. 5-7); this part of the poem provides an introduction to vv. 8-9 (cf. v. 8 sic), which in turn focus on the instability of love. The wakeful night of the lovelorn lover is another well-worn topos of love poetry: again, many parallels could be mentioned but it is not possible to identify a precise model22. The words ego... perdita clearly show that the speaking-I is a woman. I only mention the fact that, for obvious metrical reasons, perdita cannot be an ablative with nocte because two scholars, evidently too reluctant to accept a woman as the poem’s speaker, either misinterpreted or ignored this word: Della Valle 1937, 139 paraphrased this line «perdo ore preziose della notte», Della Corte 1976 «su questo meditavo l’altra notte». 6-7. For the spelling supstulit cf. CIL IV 4456 (= CLE 929) supstenet23. The speaker reinforces her previous statement of the leuitas of men by confessing that she has had occasion herself to suffer because of it; love relationships are often affected by fickle Fortune, and nobody can escape its influence. The masculine multos is generic: the speaker includes herself in a long list of forlorn lovers, regardless of their sex. The idea of Fortune’s reversals is of course commonplace; for example, Ovid adopts a similar phrasing in tr. III 11,67-68 humanaeque memor sortis, quae tollit eosdem / et premit, incertas ipse uerere uices; slightly less close is Horace, c. I 34,14-16 hinc apicem rapax / Fortuna cum stridore acuto / sustulit, hic posuisse gaudet24. Despite the wide diffusion of the topos, the Ovidian parallel quoted above seems 7,20; Quintilian XII pr. 3; Seneca, Herc. f. 152; Ag. 443. 21 Possibly an elegiac variation on the common philosophical topos we find e.g. in Lucretius III 315: naturas hominum uarias. 22 Bücheler in CLE 950, following a suggestion by Leo, compares Euripides, Hipp. 375 ἤδη ποτ᾽ ἄλλως νυκτὸς ἐν μακρῶι χρόνωι / θνητῶν ἐφρόντισ᾽ ἧι διέφθαρται βίος; however, Phaedra’s words do not concern Fortune, nor men’s unfaithfulness in love. Copley 1939, 346 mentions Tibullus I 2,76 and 2,4,11; Propertius I 1,33 and III 15,1-2; Ovid, ars I 735; am. I 2,3; Sappho fr. 94 Diehl; Meleager, AP V 152,3 and 173; XII 72,1-2; Asclepiades, AP V 150; CIL IV 2146; CLE 943. 23 Courtney 1995. 24 See Luck 1977, 220 on the Ovidian passage and Copley 1939, 347 for further references. Same verse ending in Lucan IV 649 sustulit alte. -6- OVIDIAN GRAFFITI much less casual if one considers that the ending of v. 7, praecipitesque premit, echoes very closely the ending of v. 24 from the same elegy of Ovid, praecipitata premunt. The immediate context is not related to the subject of Ovid’s poem (vv. 23-24 are a consideration on the nature of true valour), but our unknown author might have been attracted to this Ovidian iunctura by the assonance of pr-, which the closeness to proiectos makes even more marked in the graffito. 8. the initial sic refers back to the previous remark on the instability of Fortune; the sentence is incomplete due to the fragmentary or unfinished condition of the inscription. The line is clearly modelled on Lucretius V 962 et Venus in siluis iungebat corpora amantum (Copley 1939, 348). 10. All the editors of the graffito lament that the last line does not make much sense. Mau (both in the CIL and in Mau 1889, 123) only offers a bare transcription, AARIIIIS QVID AAM, corrected in PARIIIỊS QVID AṂẠ by Marichal in Varone 2012, 2.437; Sogliano 1888, 519 and Bücheler in CLE 950 provide a more readable but inevitably incomplete paries quid ama. Della Corte 1976 tries to make better sense of these letters and prints Marius quidam (a sort of fake and humorous signature), which is hard to accept both because it is very difficult to read Marius in the graffito, and because of the poor meaning. For Copley 1939, 348 only quid is clearly readable; the first word might be paries, «but if so the line must have been written by a second hand, for entirely different forms are used for the letters P and A». Indeed, this is very likely the case: a second hand is also suggested by the shape of the q in quid, that has a longer stem descending from the left instead of the right side of the letter (as in quos and praecipitesque, vv. 6-7). The very long stem of the initial p is not surprising per se; however, it is significant that, while all the previous verses are carefully written leaving a small but constant margin from a decorative frame painted on the wall, the stem of the p deliberately crosses it. Whoever wrote this line was clearly less respectful of the wall’s decoration, and/or less interested in a coherent ‘typographical’ presentation of the graffito25. 25 For a picture of this inscription see Varone 2012, 2.437. Goold 1998, 27-28 also notes that at least two different scribes were involved in the writing of this inscription, possibly because the first one got tired as he approached the bottom of the wall and the work was taken over by «a smaller boy who, if more nimble in stooping down, was less accomplished in calligraphy and less orthodox in spelling». Now, although it is a quite possible that two different scribes were involved, it must be noted that in the space below our inscription several other short graffiti were found, written in bigger letters: the image provided by Varone 2012 includes CIL IV 5297, 5298 and 5299; 5300 is not reproduced there, but the CIL locates it to the left of 5297-5299. Although no information is available about the distance of these graffiti from the floor, it is clear -7- LUCA GRAVERINI The word paries is actually readable well enough, assuming that the hand which wrote this line was not only different but also uncertain: there is an extra vertical stroke26 in the letter e (Varone 1994 reads parees, but the first of the four strokes is isolated from the following three and must be read as i); the letter a (as well as the first one in ama) has an alternative shape, different from all the a’s in the previous lines but frequent in other graffiti, and an extra stroke perhaps due to an unsteady hand. The last letters of the line are not very clear either: nevertheless, ama is a very likely reading, again considering that the letters are differently shaped27 and there is another extra stroke under the m denoting some uncertainty in the ductus. All things considered, attributing a meaning to paries quid ama is much more difficult than reading these words in the graffito; I will discuss this issue in section e. b) Speaker, addressee, author and writer At this point, it is possible to make at least some general remarks on this text. The poem’s speaker, as we have seen, is clearly a woman. Ego… perdita in v. 5 is proof enough of this: therefore, we can easily dismiss nonsensical interpretations such as Della Valle’s, who attributed the poem to a «sentimental officer who served in one of the three cohorts of Pompeii»28. The vocative pupula in v. 3 and the diminutives in vv. 1-2 strongly support the idea that the addressee of these verses is also a woman. There are obviously various issues connected with these genre-related aspects of the inscription; I will postpone their discussion to section d, and will now deal with the problems of who composed these verses and who inscribed them on a wall of a Pompeian house. One of the first editors of our graffito implicitly claimed that the speaking voice and the author of these verses were one and the same person29; indeed, this is an easy assumpthat any assumption about the need of stooping down to complete the poem is unwarranted. Equally unwarranted is Goold’s second assumption, that the lack of space forced the graffito writer(s) to come to an abrupt conclusion of the poem (which would be therefore complete and, in his opinion, composed on the spot by those who have written it on the wall): if anything, the lettering becomes slightly larger and more airy in the last two lines of the poem, and that can hardly make us suspect that the graffito writer(s) were feeling constrained by the lack of space. In fact, even admitting that CIL IV 5297-5300 were already there when our poem was inscribed on the wall, there is space for several other lines between them and 5296. 26 A common mistake: in this graffito, for example, the final m in v. 6 mecum also has an extra vertical stroke. 27 Cf. e.g. the word amat in the first line of CIL IV 5095. 28 Della Valle 1937, 168-169. Della Corte 1976 claims that the speaker is a dishonest cynic who scoffs at the distress of a young girl he has just abandoned. 29 Mau 1889, 123 spoke of an «autrice». -8- OVIDIAN GRAFFITI tion in the case of a text written in the first person. Nevertheless, it is also obvious that there are a lot of exceptions to this ‘rule’ – think, for example, of Apuleius’ Metamorphoses or Ovid’s Heroides, both written in the first person by authors who could not identify with their characters, or could only do so to a very limited extent. The identity of speaker and author, therefore, cannot be taken for granted a priori, all the more so since we only have very scanty proof of women possessing enough literary education to produce poetic texts, either written or inscribed. Nevertheless, scanty proof is far from no proof at all. Some poetic inscriptions, and even a few of the more literarily ambitious ones, are or could be by women30; and since many Pompeian graffiti were addressed to women, of course we can safely assume that there was a literate female audience who could read them. One has to admit that, if we refuse to believe in female authorship in dubious cases only because of the scarcity of proofs for female authorship in general, we end up by dismissing potentially important documents only on grounds of circular reasoning31. In the case of our poem, we should also wonder for what reason a male poet would choose to impersonate a woman speaking to another woman, be they lesbian lovers or close friends (on this issue see below, section d). Of course, a poetic exercise of any kind is always possible; and we also have to take into account the fragmentary nature of this inscription, which might obscure a more complex narratological structure. Nevertheless, all things considered, a female author is certainly the most reasonable assumption for this poem in the form we have it. Although her work will not strike anybody as a sublime achievement, we must credit her with some learning: as we have seen, she can remodel verses by Lucretius and Ovid, and use several well-known conventions of love poetry. Can this author be directly responsible of the Pompeian graffito and of its various problems concerning metre and, from v. 5 onwards, also syntax? The answer to this question can only be speculative, but in my opinion there is a very good chance that she and the graffito writer were two different persons. Most of the language and prosody issues in these verses can be easily mended, and seem to be simply due to memory slips; as we are going to see in the next pages, they tend to cluster around the inscription line breaks, which supports the idea that the physical layout of the graffito (where line breaks never correspond to verse endings, except for v. 3) contributed to the writer’s memory slips and sloppy prosody. All of this, I think, allows us to at least suspect that the graffito contains mistakes made by a clumsy or careless writer rather than by a totally unskilled author. 30 See Stevenson 2002 and 2005, 49-58; Woeckner 2002; Buonopane 2009. More generally on women’s literacy and education see Harris 1989, 252, Cavallo 1995 and especially the thorough assessment by Hemelrijk 2004, 15-91. 31 This kind of reasoning is not unusual when scholars discuss the ‘real’ authorship of the few literary pieces usually attributed to women. For a convenient overview see e.g. Keith 2006 on Sulpicia. -9- LUCA GRAVERINI It was common practice for Pompeian wall-writers to write somebody else’s words: originality was clearly not a usual requirement for graffiti. Famous lines by the main Latin authors were repeatedly painted or scratched here and there32, but this is also true for verses of unknown origin. The famous but anonymous couplet admiror, paries, te non cecidisse ruina / qui tot scriptorum taedia sustineas (CLE 957 = Courtney 77) has been written, in slightly different forms, on the walls of the basilica Pompeiana (CIL IV 1904), in the amphitheatre (CIL IV 2487) and in the main theatre (CIL IV 2461). On the same wall as CIL IV 1904 we read CIL IV 4091 (= CLE 945, Courtney 88): quis] quis amat ualeat, pereat qui nescit amare, / bis tanto pereat quisquis amare uetat. Kristina Milnor points out that «the words quisquis amat are found in numerous other places in Pompeii, more than twelve times in various material contexts and attached to various subsequent words and lines»33. This quisquis amat poem is not attested in the corpus of extant Latin literature but was clearly familiar to Pompeian wall writers, who used and adapted it on several occasions. This process of rewriting and adapting often resulted in mistakes: for example, paries of the first couplet quoted above is written as pariens in CIL IV 1904; and the first words of the hexameter are written in the wrong order in CIL IV 2487 (admiror te paries: an obvious metrical mistake). The case of CIL IV 10241 uellem essem gemma hora non amplius una / ut tibi signanti oscula pressa darem is very similar: a fragmentary, but metrically superior and therefore probably original version of this epigram is CIL IV 1698 = CLE 359 gemma uelim fieri hora non a...34. It is unwarranted, then, to assume that whoever wrote our poem on the wall of a Pompeian house must also be its author: on the contrary, it is a distinct possibility that he or she was only reproducing from memory somebody else’s lines – and not making a very good job of it. If this is the case, one can certainly wonder what was the original and supposedly correct (or less faulty) form of this poem, and try to reconstruct it at least exempli gratia. This is what I am going to do in the next section. c) The hypothetical original text The poem might have been all in hexameters35, but it is also possible that one or two pentameters appeared at irregular intervals, as it happens sometimes in Pompeian inscriptions36. 32 A thorough overview in Gigante 1950 and 1979. Milnor 2009, 301; the couplet is also written on a papyrus roll depicted on a panel painting (CIL IV, tab. xviii.1). Cf. also Courtney 1995, 304; Gigante 1979, 210-211. 34 On this epigram see Gigante 1979, 88-99, with references to the previous literature of which I only mention Pepe 1974 and Lebek 1976. 35 So Copley 1939, 339. 36 For (partial) lists of parallels see Goold 1998, 25 and Courtney 1995, 27. They both mention also the two poems in Petronius 34,10 and 55,3; however, while the succession of hexam33 - 10 - OVIDIAN GRAFFITI Let us have another look at the text, this time concentrating on its metrical problems and their possible solutions. 2. the verse is metrically irregular, but it can easily be corrected to form either a hexameter or a pentameter. The CIL suggests that it would be a perfect pentameter if we read labris instead of labellis. Copley 1939, 339, who argues for an all-hexameter poem, suggests that a word such as rapta might have been forgotten by the writer at the end of the first line of the inscription, i.e. after teneris (the same might have happened at the end of the second line: cf. v. 4); other suitable options could be misera (nom., cf. perdita at v. 5), or other adjectives used by elegiac poets with oscula, such as multa, mille, cara37. If the inscription preserves the beginning of the poem, a pentameter is exactly what would be expected here. As we will see, the poem certainly did not have a regular elegiac structure, but it is likely that our unknown poet tried her best to have a regular couplet at least at the beginning. 3. this line becomes a hexameter if et is added before uentis (so Mau in the CIL); very possibly a mnemonic mistake by the writer. 4. this line lacks one foot (or a spondee/anapest before natura) to be a hexameter. Mau suggests that pupula, an emphatic repetition from the previous verse, might have fallen at the beginning – that is, after a line break in the graffito. It would be more difficult to make this verse into a pentameter, and no suggestions have been put forward to this end. eters and pentameters appears to be rather casual in the carmina epigraphica, the structure 2 hexameters + 1 pentameter used by Petronius is more regular and justified. He uses it wittingly, and his pentameters provide an unexpectedly humorous conclusion to the deeper thoughts elaborated in the previous lines. 37 Goold 1998, 26 also argues for a hexameter, but rather than integrating a word like Copley he suggests to scan oscula as a molossus instead of a dactyl. In my opinion, although scansion mistakes are not uncommon in Pompeian graffiti, an easy emendation is certainly preferable to the irregular scansion of two syllables (including the desinence) of such a common word in elegiac poetry. The parallel provided by Goold 1998, 25-26, otiosis scanned as a molossus in CIL IV 813 = CLE 333, is certainly not a mistake of equivalent importance, and not even of the same kind: on i before a vowel treated as a consonant in some inscriptions see Courtney 1995, 24 (conversely, oscula scanned as a molossus instead of a dactyl does not correspond to any of the common prosodical anomalies listed by Courtney 1995, 23-26). Goold rejects Mau’s correction of labellis into labris since the diminutive, which «nicely corresponds to braciola and pupula», must have been ‘purposely contrived’, but (especially if one assumes that who wrote these lines on the wall was not also their author) his reasoning can be easily reversed: the other two diminutives could have prompted the corruption of labris into labellis. - 11 - LUCA GRAVERINI If a corrupted hexameter is really what we have here, this verse is a good hint that the original poem could not have a regular elegiac structure: Copley 1939, 338 rightly notes that it would be difficult to suspect a lacuna and insert a pentameter after verse 3, which appears to be well connected to this verse. 6-7. these are by far the most problematic lines, both because of their prosody and their poor syntax; the writer’s memory was apparently very confused at this point, and all reconstructions can only be speculative. First of all, the written meditas is interpreted as meditans by most editors, an easy correction since the nasal m is also omitted twice in the previous line; cf. also coiunxit for coniunxit in v. 8. In fact, meditas (instead of the regular meditaris) would be very hard to accept for three reasons other than metaplasm, which would not be impossible per se in everyday language: a past tense and not a present is expected in conjunction with the previous uigilarem; the word order suggests ego of v. 5 and not an understood tu as the subject; finally, if this is a love poem (on this see section d), it is hard to reconcile the idea that the two lovers are meditating together (implied by mecum meditas), with the previous perdita (the woman who speaks is in despair, and therefore presumably not enjoying the company of her beloved in the night). As regards the metrical issues, Mau in CIL suggests omitting Fortuna in v. 6 (but this would leave supstulit without a subject) and subito in v. 7 (the adverb could be a reduplication from v. 8). Other options have been suggested for v. 6: Wick emended multos Fortuna to modo Fors,38 and Goold to Fors, quos modo39. These reconstructions fix the metre and restore the correlation modo... modo40. However, they do nothing to give the sentence a main verb41; and the expunction of multos is hard to accept, since it transforms a generalization, very appropriate in a proverbial expression or moral adage like this, into a hard and fast rule with no exceptions. In my opinion a better, if tentative, reconstruction of these verses would be: 38 Wick, quoted by Mau in his addendum to CIL IV 5296 at p. 705. Goold 1998, 27. 40 The correlation modo... modo, however, is not strictly necessary. Cf. OLD s.v. 6b, with a single modo meaning ‘at another time, then’; ThLL s.v. 1311,27ff. «i.q. postmodo, deinde, statim, paulo post»; Shackleton Bailey 1956, 32. Dominicy 1974 reviews all the passages listed by the ThLL and dismisses (not always convincingly) the possibility that they can support the meaning of modo = paulo post. However, at pp. 295-298 he also stresses that the first member in a modo... modo structure can be easily suppressed: the best example is Manilius I 869-871: sed trahit ad semet rapido Titanius aestu / inuoluitque suo flammantis igne cometas / ac modo dimittit. 41 Unless one considers meditans (eram) as equivalent to meditabar (so Goold 1998, 26, but without offering any parallel for this usage) or understand something like dixi (Courtney 1995; but meditans dixi does not seem to make much sense). 39 - 12 - OVIDIAN GRAFFITI haec mecum meditans: multos [Fortuna] quos supstulit ante Fors modo proiectos [subito] praecipitesque premit. The expunction of Fortuna and the emendation of hos into Fors are obviously related, and since both corrections are on the same line of the inscription they are easily explained by a single memory slip; the adverb subito in v. 7, probably reduplicated from v. 8, is also expunged. A main verb is still missing; one could correct meditans: multos into meditabar: eos, but again, in my opinion, multos is hard to do away with. Another option would be to correct v. 5 in order to have a main verb there: for example, saepe ego sub media uigilabam, perdita, nocte42 This would make medita(n)s in v. 6 perfectly acceptable, although at the price of two small corrections in one of the few metrically unproblematic verses in this poem.43 In the end, it is clear that we can only speculate about the original form of these verses. Both the difficulty of imagining praecipitesque premit as the second half of a hexameter (it would have a very harsh central diaeresis) and the fact that Ovid’s verse on which these words are modelled is a pentameter suggest that v. 7 should be treated as a pentameter; it is much more difficult to make it fit Copley’s assumption of an all-hexameter poem. Copley’s own speculative reconstruction of vv. 5ff. is haec mecum meditans: multos modo supstulit alte / fors, modo praecipites subito proiecit eosdem; after these lines another verse would be lost, containing premit and probably a main verb governing the whole sentence (p. 340f.). Ultimately, the difficulty of this emendation and the poor quality of its results44 are in my opinion a further argument for the presence of a pentameter here. 9-10. Be it the beginning of a hexameter or of a pentameter, v. 9 is irregular: diuidit is a cretic but should be scanned as a molossus. Bücheler suggests that it might be a mistake for diuellit; Copley 1939, 348 proposes diuisit, an easier correction although a present like diuellit after the perfect coiunxit is supported by the previous sequence supstulit… premit. He also suggests that se at the end of this line could be the beginning of seiunxit but makes no hypotheses as to what could follow. 42 For the expression sub media... nocte cf. e.g. Vergil, Aen. VII 16 sera sub nocte; Statius, Theb. II 102 nocte sub alta. 43 A further possibility, of course, is that the main verb was included in the last and fragmentary part of the poem, although the resulting sentence (starting with v. 5 saepe) would have been very long and probably cumbersome. 44 Of course, the Ovidian flavour of the ending of v. 7 is lost in Copley’s reconstruction: one should think that the poetic echo is only the casual result of a clumsy wall-writer’s work. - 13 - LUCA GRAVERINI Others try to make sense of this final part of the graffito by attaching v. 10 to v. 9. Varone 1994, 100 connects se to what he reads as parees in the following line in order to obtain separes from a 3-conjugation separo45, and reads what follows as qui amant. However, this would be the only case in this graffito where a word is split between two lines; most importantly, the syntax is still garbled by an unexpected change of subject from Venus to an understood (and not very understandable) tu, and a further correction of quid to qui is required; the meaning of the sentence is unsatisfactory even in Varone’s translation «in modo eguale, come Venere congiunse repentina i corpi degli amanti, così la luce li divide e il loro amor separerai». Goold interprets the last two lines of the inscription as an apostrophe to the wall46. He reads sic instead of se at the end of v. 9; what the writer really intended to write, however, would be dic, so that the final verse of the poem would be diuidit lux et dic, paries, quid aam? As for aam, it would be a phonetic rendering of agam, since colloquial speech is known to weaken the intervocalic g. This is rather hard to accept, since this spelling would be unparalleled; one might add that the intervocalic g is preserved in other words in the graffito, such as ego and uigilarem (the phonetic context is different, but ego → It. ‘io’ is exactly one of the examples mentioned by Goold to support his suggestion). As regards the meaning, I cannot but find it rather odd that a wall should be personified to the point that it can become a trusted interlocutor who can give advice to a desperate lover, although of course this is not altogether impossible in a highly rhetoricized context. In my opinion, any reconstruction of the missing three feet of v. 9 is highly speculative; it is safer not to venture into hazardous conjectures, and simply accept the poem’s fragmentary status. Goold’s reconstruction, although not satisfactory per se, shows us that any attempt to consider vv. 9 and 10 as the beginning and end of a single (and final) verse must overcome at least two serious difficulties: - it is not easy to understand why they appear to have been written by different hands; - the resulting verse makes for a very abrupt and ultimately unsatisfying conclusion to the poem. In this situation, I will leave v. 9 unfinished, and will try to provide a plausible explanation for the last line of the inscription, the enigmatic paries quid ama, at the end of this paper. To sum up, here is a possible reconstruction, exempli gratia, of our poem: o utinam liceat collo complexa tenere braciola et teneris | oscula ferre labris. 45 For the metaplasm cf. Statius, Theb. IV 481. Goold 1998, 27-28, apparently elaborating on a suggestion by Leo quoted by Bücheler in CLE («hanc Leo adlocutionem facit parietis»). 46 - 14 - OVIDIAN GRAFFITI I nunc, <et> uentis tua gaudia, pupula, crede: | <pupula>, crede mihi, leuis est natura uirorum. Saepe ego sub media | uigilabam, perdita, nocte haec mecum meditans: “multos | [Fortuna] quos supstulit ante, Fors modo proiectos [subito] | praecipitesque premit; sic Venus ut subito coniunxit | corpora amantum, diuellit lux, et se... | paries quid ama... 5 The suggested corrections only improve prosody and syntax, and none alters the meaning of the poem. Therefore, this reconstruction offers us a better (if hypothetical) idea of the poem’s author and of her literary skills, but its speculative nature does not affect the more general remarks I am going to make in the next section. d) Genre and occasion After Copley’s paper of 1939, this poem has often been labelled a paraklausithyron47. The position of the graffito, written on a doorway, has clearly influenced this generic designation more than its content48 since no mention is made in these verses of a door (or any physical obstacle separating two lovers) or of a plea for admission, which are the most distinctive features of this kind of love poem. The only possible textual hint is the mention of a paries in the last line of the graffito, apparently written by another hand. In my opinion, the suggestion of a paraklausithyron is very weak, and not only because, as I am going to show in the next pages, the last line of the inscription containing the decisive word paries does not pertain to the same poem as the previous lines: in fact, this interpretation of the graffito stems from a double misunderstanding about its physical context. Bücheler stated in CLE that our poem was «Pompeis in aedibus medici parieti inscriptum stilo», but he was wrong: both the excavation reports49 and the CIL make it very clear that the graffito is on the right wall of the entryway of another house, pertaining to the same insula 9 but much smaller than the rather grand «doctor’s house» which faces the «via di Nola». The door to this small, nondescript house (marked with an asterisk in Figure 1; the «doctor’s house» is marked with a diamond) opens on a narrow side alley between insulae 8 and 9 and it is the sixth on the left for those who 47 See Gigante 1979, 212-216, Goold 1998, and more obliquely Courtney 1995, 306 («it looks like a specimen of verses of a type of which we hear particularly in the elegists, those written by a lover on the obdurately closed door of the beloved»). For the formal elements typical of paraklausithyra see Copley 1956. 48 So admits Copley himself (1939, 337). 49 Sogliano 1888; Mau 1889. - 15 - LUCA GRAVERINI come from via di Nola; a short corridor leads to a central open area, flanked by only two small rooms. Bücheler’s mistake was repeated by all the scholars who support the interpretation of this poem as a paraklausithyron50. Even more important is the second misunderstanding: all these authors (except Bücheler) take it for granted that this graffito, located «on the doorway» of a house, was in a public space, where a lovelorn poet could entrust a wall with his feelings for his/her beloved who supposedly lived inside the house. Copley, who first proposed the paraklausithyron theory, drafted a list of Pompeian «doorway scribblings» that could create an appropriate context for this poem51, but he clearly mixed together inscriptions found in public and private areas. The CIL locates most of them «ad sinistram/dexFig. 2 Insula IX tram ostii X», that is near a door, outside a house and on the public side of the wall, where anybody who passed by could write; one, however, was clearly inside a house, «in quarti osti postis dexterioris parte interna»52. The CIL note for IV 5296 is not so specific, but it clearly implies that that the graffito was in the house’s entryway and inside the door, not outside it: «in ostii sexti a septentrione pariete dextro», confirmed with some more detail by Sogliano who locates the graffito «nel 6° vano, sull’alto zoccolo grigio dalla parte a dr. di chi entra»53. The graffito itself is no longer visible on the wall, whose plaster is now lost. A slab of it, however, measuring 74 * 38 cm. and containing CIL IV 5296-5300, was detached before the damage occurred, and is kept (but not catalogued) in the Museo Archeologico of Naples; a partial picture of it is now available in Varone 2012, 2.437. Its size, in my opinion, also makes it impossible to think that it was outside the door of this house: on the right wall of the entryway there is a hole, which evidently accommodated the door crossbar, and it 50 Copley 1939, 334 n. 3, Courtney 1995, 306 and Goold 1998, 18. Copley 1939, 333 n. 3. 52 CIL IV 2413h: Cestilia regina pompeianoru(m) anima dulcis uale. 53 Sogliano 1888, 519; cf. Mau 1889, 122, «sul muro d. dell’ingresso». 51 - 16 - OVIDIAN GRAFFITI is only 40 cm. from the outside end of the entryway. Now, since the exclusus amator is by definition closed outside his beloved’s house, the position of our graffito inside a house can hardly add to its characterization as a paraklausithyron; if we consider that this poem was inscribed in a private and not public space, the typical ‘real-life’ reconstructions offered by several scholars do not make much sense either54. Our graffito is therefore one of those inscriptions frequently found inside private houses: greetings, names, calculations and number games, simple notes taken for practical purposes, but also more ambitious texts such as short love messages, literary quotes and brief poems, written both by men and women, house owners and visitors, sometimes in dialogue with each other. Contrary to what we might suppose, these graffiti, or at least many of them, were not considered a defacement of the house’s decoration; in fact, they are frequently found in visible and open spaces, where people would gather or pass by55. Many different situations might have prompted somebody to inscribe this short poem on the wall of a small Pompeian house. A troubled love affair between the writer and a resident of that house is only one of the possible reconstructions, however tantalizing. For example, we cannot exclude the opposite, a message left by a resident for a visitor; or that our verses were simply written without any immediate sentimental reason, just for showmanship, or to record a badly remembered but appreciated poem. An entranceway is a place where people pass through but also wait, and someone might have decided to simply kill some time by writing a poem on the wall – and left it unfinished when his wait was over56. At least in the context of a literary approach, after all, the most notable feature of this poem is not the reason why it has been inscribed on a wall, which is ultimately impossible to reconstruct, but the very peculiar fact that it records the poetic voice of a woman talking to another woman. Indeed, it is so peculiar that it has been sometimes ignored or 54 Goold 1998, 28 imagined a group of youths scribbling these verses on the wall of a house «where there lived a girl who had set their hearts a-flutter»; other reconstructions in Della Valle 1937 and Copley 1939. Courtney 1995, 306 more cautiously only alludes to the paraclausithyron as a generic designation, with no references to real-life situations. 55 On graffiti inside private houses in Pompeii see Benefiel 2010 and 2010b. 56 As we have seen (see above, n. 25), CIL IV records several other short graffiti located on the same wall, below our inscription; they are all fragments of a dialogue between two or more persons. Two of them are greetings (5297 Crocine uale | Ismare uale; 5298 Ismarus Crocineni suae salutem; v. Benefiel 2010b, 27); one is perhaps a joke on the name of one of the persons mentioned in the two previous inscriptions (5300 duodsmaro; the CIL suggests «Duodsmaro pro Ismaro… ut pro I, i.e. uno, Duod, i.e. duo, poneretur?»); the last one (5299 deuronym) is an erotic invitation in Greek, on which see Barchiesi 2012. All these graffiti are visible in the picture provided by Varone 2012, 2.437, except 5300 (5298 only partially). - 17 - LUCA GRAVERINI explained away: as we have seen, implausible readings have been suggested where either the poem speaker or his addressee is a man, thus restoring the ‘correct’ gender relationship in elegiac poetry. While this ‘gender normalization’ can only stand on a forced interpretation of the text, one should also be careful not to jump directly to the opposite conclusion, and claim that this poem is an unmistakable reference to lesbian love. This may be inferred from the hugs and kisses described in vv. 1-2, but the rest of the poem obliquely refers to man-woman relationships (v. 4 leuis est natura uirorum, voiced by a female speaker; v. 6 multos includes both men and women; v. 8 amantum is not explicitly gendered, but the original Lucretian context is clearly heterosexual). As for vv. 1-2, it is difficult to decide if they describe an affectionate or passionate, friendly or amorous relationship without making unwarranted assumptions, biased by contemporary habits and socio-cultural expectations. Should we compare the kisses and embrace of these verses to Propertius’ passionate lovemaking with his Cynthia (II 15,9 quam uario amplexu mutamus bracchia! quantum / oscula sunt labris nostra morata tuis!) or to Lucius’ affectionate but chaste relationship with a priest in Apuleius’ novel (met. XI 25,7 complexus Mithram sacerdotem et meum iam parentem colloque eius multis osculis inhaerens...)? Indeed, similar words can describe very different situations57. In our case, since we lack a proper context for this anonymous and fragmentary poem, written for unknown reasons on a wall of a Pompeian house, we will need to keep an open mind, and allow for different interpretations. The poem’s speaker might be urging a friend of hers to choose her company and her love over that of men, who are notoriously unreliable partners; on the other hand, she might simply want to offer warm gestures of consolation to a lovelorn friend and tell her about her own experience of love sufferings58. 57 Ambiguities and interpretive difficulties are common in this field. It is well known that the ancient Greeks and Romans themselves interpreted the philia between Achilles and Patroclus in very different ways (see e.g. Fantuzzi 2012, 187-265); contemporary scholars are beginning to read between the lines in the sometimes very affectionate correspondence between Marcus Aurelius and Fronto (see Richlin 2006). On female homosexuality in ancient Greece and Rome see Boehringer 2007. 58 We might be tempted to think that the decision between the two possible interpretations depends on the case of teneris... labellis in v. 2 (see above, section a), which implies a kiss on the mouth if it is dative. Nevertheless, again we would be making unwarranted assumptions on the sexual or simply affectionate and friendly value of such a kiss. For example, it is useful to remember that the Roman ius osculi (according to which men could, and indeed should, kiss their female relatives within the sixth degree on the mouth) was perplexing to Plutarch (quaest. Rom. 265 b-e) no less than to us: see Bettini 1988, 79 and 2009, 27-28. For a recent overview on kisses in antiquity (which however does not provide support for either interpretation of this - 18 - OVIDIAN GRAFFITI As regards the person who inscribed the poem on the wall, the only thing we know for sure is that she59 did not carry her work through – perhaps she got tired, or was interrupted by someone, or her recollection of the poem she was writing, already faltering at vv. 6-7, simply became too blurred to continue. The extant ending of the poem are the words diuidit lux et se: the last line of the graffito, paries quid ama, appears to have been written by a different hand60 and, as I am going to show now, is not part of it. e) The last line of the graffito This line was added later by another visitor or resident of this house who noticed our graffito and deemed it worthy of comment61; in his62 own way, he contributed to the lively dialogue we can now appreciate on that wall63. His handwriting is uncertain, but he does not lack literary education. He has probably recognized the inscribed poem’s elegiac and markedly Ovidian inspiration, and the final mention of the lovers’ separation has led him to think of one of the most famous myths about separated lovers, that of Pyramus and Thisbe. Ovid’s story of the two Babylonian lovers was popular in Pompeii, as attested by several wall paintings clearly dependent on the Metamorphoses64; it is therefore not unexpected at all that a casual reader of our graffito, even a not particularly well-read one, could remember the two lovers’ desperate apostrophe to the wall (Ou. met. 4,74) «invide – dicebant - paries, quid amantibus obstas?» graffito) see e.g. Hawley 2007. Useful material also in Boehringer 2007, esp. 314-321 on the kiss of Fortunata and Scintilla in Petronius’ Satyricon. 59 I am assuming the writer was a woman, since a man would hardly choose a female-voiced poem to express his feelings for a woman. However, as we have seen, we must also consider the possibility that the inscription was not directly motivated by a sentimental reason. 60 See above, section a on v. 10. 61 Already Copley 1939, 348 (followed by Cugusi 1996, 377) thought that «the words look like the beginning of some comment on the foregoing verses». 62 I am only conventionally assuming that this writer is a man. His sexual identity, in fact, is even more uncertain than that of who wrote the previous lines of the inscription (see above, n. 59). 63 See above, n. 56. 64 See e.g. Linant de Bellefonds in LIMC 7,1994; Sauron 2004; Rosati in Barchiesi - Rosati 2007, 256. Of course, the wall separating the two lovers is an important figurative element in some of these paintings. The name PIRAMVS is also painted in red in the atrium of another house (I 16; cf. Giordano - Casale 1991, 277). It also occurs in two obscene inscriptions (CIL IV 10041d Piramo cottidie linguo; 10090 Petroni, Pyramo pedicas), where it is clearly used for a slave or a puer delicatus. - 19 - LUCA GRAVERINI and use the second half of that verse as a poetic comment to the lines he had just read: this second writer is demonstrating that he can add his two cents to the unfinished inscription, keeping its distinct Ovidian flavour. Needless to say, the word paries in Ovid’s line makes it particularly appropriate for a graffito, and might have functioned as a sort of ‘mnemonic magnet’ for our unknown commentator. Pompeian walls were often used as places for poetic dialogues, sometimes involving satire and jokes; it was a living dialogue, often impromptu, not rarely left interrupted at mid-sentence or even mid-word. A short survey might start with a graffito connected to our poem both by a similar Ovidian inspiration and by geographical proximity. A few meters to the north, between the door behind which our graffito is located and the door to the next house, three words were scratched that might form the beginning of a hexameter (CIL IV 5303 = CLE 359): Aureus est Danae Mau 1889, 122 suggested that these words are a blurred memory of the beginning of an Ovidian verse that celebrates Zeus’ amorous conquests: met. VI 113 aureus ut Danaen, Aesopida luserit ignis65. In his CIL entry he changed his mind, and compared this graffito to CIL IV 4207 (also collected by Bücheler, who suggested the integrations to the second verse, as CLE 359): esurit Danne e]co sum a[ureus In view of this comparison, he discarded the reference to Ovid and suggested that both inscriptions quote from a poem that took an ironic or moralistic view of Danae’s myth, in the same vein as at the beginning of Horace carm. III 1666. Nevertheless, even granting the existence of this hypothetical comic poem about Danae, it is difficult to deny the Ovidian inspiration of the first inscription. In particular, the small difference between the graffito and Ovid’s verse cannot be ascribed to the writer’s faulty memory: even an absent-minded writer could certainly not make aureus agree with Danae, both because of gender and mythology. Therefore, the sentence has 65 Sogliano 1888 p. 519 n. 27, the first editor, only noted the dactylic pattern of the words. Cf. vv. 7-8 fore enim tutum iter et patens / conuerso in pretium deo. This view is of course very common: one might also mention Ovid, am. III 8,29-30 Iuppiter, admonitus nihil esse potentius auro, / corruptae pretium uirginis ipse fuit; Martial XIV 175 Cur a te pretium Danae, regnator Olympi, / accepit, gratis si tibi Leda dedit?; Petronius 137,9 quisquis habet nummos, secura nauigat aura / fortunamque suo temperat arbitrio. / uxorem ducat Danaen ipsumque licebit / Acrisium iubeat credere quod Danaen. 66 - 20 - OVIDIAN GRAFFITI obviously been left unfinished, just like our poem and the short comment below it67. The second inscription (possibly itself a parody of Ovid’s already quoted line: esurit Danae is also very close, at least phonetically, to aureus ut Danae) is indeed more comic, and seems to suggest that hunger or greed played an important role in Danae’s submission to Zeus. In the context of this paper, this graffito is especially interesting in showing that even a writer learned enough to play with myths could have problems with orthography (Danne instead of Danae)68. If Bücheler’s integration is correct, the second line, written in slightly smaller characters, could be a mocking remark by a different writer, who declares himself willing to satisfy Danae’s hunger in the place of Zeus – after all, one can simply have some money in his purse instead of transforming himself into a golden shower...69. Although the reconstruction of this line is only speculative, mocking or praising remarks or simple comments are not rare at all in Pompeian poetic inscriptions: our paries, quid ama(ntibus obstas) would certainly not be an isolated case. An especially interesting analogy is provided by CIL 4,1837 (= CLE 949, Courtney 91): Si potes et non uis, cur gaudia | differs spemque foues et | cras usque redire iubes? | er]go coge mori quem | sine te uiuere cogis | munus erit, certe non | cruciasse boni. quod spes | eripuit spes certe redd[i]t amanti 5 If you can but are unwilling, why do you postpone pleasure and foster hope and continually tell me to come back tomorrow? So force me, whom you force to live without you, to die. At any rate to have refrained from torturing will be the gift of a good man. Assuredly hope restores to a lover what (lack of ) hope has taken away (trans. Courtney). Like the inscription that is the main subject of this study, this graffito presents some metrical problems that are probably the consequence of memory slips by a careless writer 67 We can only speculate on what has been left out; exempli gratia, and assuming we have the beginning of a hexameter here, I would suggest something like aureus est Danaen complexus Iuppiter olim. 68 Or was it a non-mythological name? The only parallel I could find is in an inscription from Africa Proconsularis that mentions a Dannea Donata (AE 2004, 01721). 69 The CIL editor calls into question Bücheler’s conjecture and points out that «nihil periit». I adopt Bücheler’s text exempli gratia and for the sake of my argument; alternatively, one could think of a form of the verb consumo used in an erotic context as in Statius, sil. III 4,42 quem sterilis consumpsit amor, or accept the reading cum suma[t] hesitantly suggested by Varone 2012, 1.228 (who also provides a picture of the inscription, very hard to read). - 21 - LUCA GRAVERINI rather than the product of an unskilled author. So, at line 1 a foot is lacking: Bücheler suggests that fallax might be added after vis, or nunc dare at the beginning; a further proposal, accepted by Courtney, is to add mutua before gaudia, stressing even more the imitation of Ovid, am. 3,6,87-88 quid mutua differs / gaudia. In my opinion, this last suggestion is only slightly off the mark: it would be better to correct cur to quo nunc mea, obtaining a quotation of Ovid, am. II 5,29 quo nunc mea gaudia differs and keeping the poem focused on its speaking-I (he does not seem to be interested in his partner’s joy or pleasure, as mutua would imply; moreover, the corruption of quo nunc mea into cur is in my opinion an easier mistake than the loss of a rather relevant adjective like mutua). Verse 3 is also ametrical; the hexameter can be easily restored, as Bücheler suggests, by adopting a different word order (te sine or sine te quem). Be that as it may, the model is again Ovid: cf. her. 3,140 quam sine te cogis uiuere, coge mori. Verse 4 also begins with a possible Ovidian reminiscence, cf. trist. I 2,52 mors mihi munus erit. What follows is a bit convoluted and difficult to understand: Courtney considers boni as a masculine governed by munus (and this would imply that the speaking voice is that of a woman); Gigante 1979, 212 considers it a neuter again with munus («il dono di un’azione buona»); Bücheler (best of all, in my opinion) interprets «mors mihi munus erit, certe boni erit non cruciasse». This inscription does not only provide us with a mix of sophisticated literary texture, metrical clumsiness and occasional obscurity that is similar to CIL IV 5296; it is also followed by several remarks, similar to our paries quid ama. Verse 5 might already be a comment by somebody who wanted to balance the hopeless pessimism of the first two couplets with an isolated hexameter containing a comforting thought70; but the poem is followed by other lines that are beyond doubt comments left by casual readers71: qui hoc leget nuncquam posteac aled72 legat nunquam sit saluos qui supra scripsit 70 So e.g. Gigante 1950, 136 e 1979, 211, following Wick 1907, 28 (which I could not access directly). This hypothesis would explain very well the sudden change of topic of v. 5, but it does not find a clear palaeographical support: Zangemeister in the CIL informs us that the whole poem was written by the same hand, while Mommsen attributed vv. 4 and 5 to a different hand. It might also be added that it would be strange for a comment to span across two lines of the inscription, with the first two words immediately following the ending of v. 4. Monda 1993, 243 n. 35 suggests that v. 5 could be a consideration on spemque foues of v. 2. 71 For the attribution of the comments to different hands see the addendum to CIL IV1837 at p. 212. 72 For this form see Todd 1939,9, who attributes it to vulgar language. - 22 - OVIDIAN GRAFFITI uere dicis Hedysto f[e]licit[e]r The first reader is clearly impressed with the poem and praises it as the ‘definitive’ love poem: after reading this one, there is no reason to read anything else73. The second one apparently does not agree with his predecessor’s statement; he scolds him with a scathing remark, and is supported by a third ‘commentator’. A fourth person puts an end to the literary dispute, and dedicates the poem to some puer delicatus. This is just an impressive instance of a rather common practice: as Paolo Cugusi notes, inscribed erotic poems often drew the attention of passers-by, who expanded the poetic message with their personal observations74. Conclusions Although most of them were clearly not great poets, many of the unknown authors of Pompeian verse inscriptions were evidently provided with some learning75. together with Virgil, Ovid featured prominently in their literary education76. New 73 Unless there is some heavy irony in his words: a passer-by might not be able to read anything else also because he became blind (or even died) after reading such a bad poem. 74 Cugusi 1996, 377; cf. Monda 1993, 243-244. Comments might also assume a more sophisticated form. See e.g. Milnor 2009, 299-301 on CIL IV 9131 fullones ululamque cano, non arma uirumque, clearly a humorous reworking of the first verse of the Aeneid, which might be intended as a witty response to a wall painting and electoral programme painted on the nearby door to the so-called house of Fabius Vlulitremulus. On dialogues between graffiti also see Benefiel 2010, who states that «clustering shows that graffiti were not the domain of the individual but were social in nature, often the work of someone contributing to a conversation» (87). 75 Gigante 1979 and Milnor 2009 provide complementary pictures, the former showing the sheer amount of literary culture that is detectable in Pompeian inscriptions, the latter suggesting that some poetic quotations might simply be used to fill in blank spaces and give more prominence to ‘utilitarian’ inscriptions. 76 On the prominence of Ovid in Pompeian inscriptions see e.g. Gigante 1979, 185-194; Munzi 1996; Sauron 2004; for a specific case (CIL IV 10241, mentioned above at the end of section b), Lebek 1976; in the CLE in general, e.g. Cugusi 1982, 89-98; 1996, 184-186 and 357-358; Rodríguez-Pantoja 1994. Among the most notable Ovidian quotations in Pompei are CIL IV 1520 (= CLE 354) Candida me docuit nigras odisse puellas. / Odero si potero, si non, inuitus amabo (cf. Propertius I 1,5 donec me docuit castas odisse puellas; Ovid, am. III 11,35 odero, si potero; si non, inuitus amabo); CIL IV 1895 (= CLE 936) Quid pote tam durum saxo aut quid mollius unda? / dura tamen molli saxa cauantur aqua (cf. ars I 475-476 quid magis - 23 - LUCA GRAVERINI anonymous poems were the spontaneous result of this culture. Enthusiastic students and adults were eager to recycle and adapt in their own verses the topoi and phrases used by the famous poets they read in school or in their leisure time, and on some occasions those verses found their way to the walls of Pompeian houses; others gave proof of their cultural prowess (or, indeed, just of their literacy) by adding their comments to inscriptions they happened to read. These activities do not necessarily give evidence of a highly cultured society, but at least to some extent they are proof of the vitality and diffusion of sophisticated literary products that were by no means confined to the bookshelves of a narrow elite. CIL IV 5296 is in all likelihood the result of these mid-level cultural activities. It was probably composed by a learned young woman, and this possibility makes it extremely important to us as one of the very few instances of preservation of an ancient female poetic voice – perhaps, also one of the even fewer literary testimonies of lesbian love. Her poem enjoyed some diffusion, until for some reason somebody decided to write it on the wall of a house. She or he did a rather clumsy job and left it unfinished; nevertheless, the inscription drew the attention of somebody else, who grasped its Ovidian inspiration and decided to leave a comment in the same vein, borrowing a line from Ovid’s account of the myth of Pyramus and Thisbe – and for some reason, leaving it also unfinished77. est saxo durum, quid mollius unda? / dura tamen molli saxa cauantur aqua); CIL IV 1069a (= CLE 350) Barbarus aere cauo tubicen d[ed]it [horrida si]gna, cf. met. III 704 cum bellicus aere canoro signa dedit tubicen. 77 I am grateful to the organizers and audience of the Incontri di Filologia Classica in Trieste, in particular Lucio Cristante, Marco Fernandelli, and Marco Fucecchi: the seminar was an invaluable occasion to meet learned friends and colleagues and to receive useful advice. My gratitude also goes to Sandro Barchiesi, who first drew my attention to CIL IV 5296 and discussed several important points with me; to Lara Nicolini, who read an early draft of this paper and was of great help with her suggestions; to the anonymous referees of this journal, for their helpful remarks; and to the Soprintendenza per i Beni Archeologici di Napoli e Pompei, for allowing me to visit the original location of the graffito in Pompeii. William Short has kindly agreed to revise my English. - 24 - OVIDIAN GRAFFITI WORKS CITED Baldi 1982 A.Baldi, Iscrizioni pompeiane, Cava dei Tirreni 1982. Barchiesi – Rosati 2007 Ovidio, Metamorfosi, vol. II: Libri III-IV, a cura di A.Barchiesi e G.Rosati, Milano 2007. Barchiesi 2012 A.Barchiesi, Pompei, i graffiti, e il linguaggio amoroso dei Romani: una piccola frase, in M.Baioni – P.Gabrielli (ed.), Non solo storia. Saggi per Camillo Brezzi, Cesena 2012, 37-48. Benefiel 2010 R.R.Benefiel, Dialogues of Ancient Graffiti in the House of Maius Castricius in Pompeii, «AJA» CXIV (2010), 59-101. Benefiel 2010b R.R.Benefiel, Dialogues of Graffiti in the House of the Four Styles at Pompeii (Casa Dei Quattro Stili, 1.8.17, 11), in J.A.Baird – C.Taylor (ed.), Ancient Graffiti in Context, London 2010, 20-48. Bettini 1988 M.Bettini, Il divieto fino al sesto grado incluso nel matrimonio romano, «Athenaeum» LXVI (1988), 69-98 (then in J.Andreau – H.Bruhns, ed., Parenté et stratégies familiales dans l’Antiquité romaine. «Actes de la table ronde des 2-4 octobre 1986, Paris», Paris 1990, 27-52). Bettini 2009 M.Bettini, Comparare i Romani: per una antropologia del mondo antico, in A.Barchiesi – G.Guidorizzi (ed.), La stella sta compiendo il suo giro. «Atti del Convegno internazionale di Siracusa (21-23 maggio 2007)», Firenze 2009, 1-48 [suppl. SIFC]. Boehringer 2007 S.Boehringer, L’homosexualité féminine dans l’antiquité grecque et romaine, Paris 2007. Buonopane 2009 A.Buonopane, Una voce di chi non aveva voce: i graffiti delle donne, in M.G.Angeli Bertinelli – A.Donati (ed.), Opinione pubblica e forme di comunicazione a Roma: il linguaggio dell’epigrafia. «Atti del colloquio AIEGL-Borghesi 2007», Faenza 2009, 231-245 Callebat 1968 L.Callebat, Sermo Cotidianus dans les Métamorphoses d’Apulée, Caen 1968. Churchill – Brown – Jeffrey 2002 L.J.Churchill – P.R.Brown – J.E.Jeffrey, Women Writing Latin. From Roman Antiquity to Early Modern Europe. Vol. 1: Women Writin Latin in Roman Antiquity, Late Antiquity, and the Early Christian Era, New York-London 2002. - 25 - LUCA GRAVERINI Copley 1939 F.O.Copley, A Paraclausithyron from Pompeii: A Study of C.I.L., IV, Suppl. 5296, «AJPh» LX (1939), 333-349. Copley 1956 F.O.Copley, Exclusus amator. A Study in Latin Love Poetry, Ann Arbor 1956. Courtney 1995 E.Courtney, Musa Lapidaria. A Selection of Latin Verse Inscriptions, Atlanta 1995. Cugusi 1996 P.Cugusi, Aspetti letterari dei Carmina Latina Epigraphica, Bologna 1996 (19851). Della Corte 1976 M.Della Corte, Amori e amanti in Pompei antica. Antologia erotica, Cava dei Tirreni 1976 (19601). Della Valle 1937 G.Della Valle, L’amore in Pompei e nel poema di Lucrezio, «A&R» XV (1937), 139-175. Dominicy 1974 M.Dominicy, Les premières attestations de modo au sens de nunc, «AC» XLIII, 267-303. Fantuzzi 2012 M.Fantuzzi, Achilles in Love. Intertextual Studies, Oxford 2012. Gaisser 2009 J.H.Gaisser, Catullus, Chichester/Malden (MA) 2009. Gigante 1950 M.Gigante, La cultura letteraria a Pompei, in Pompeiana. Raccolta di studi per il secondo centenario degli scavi di Pompei, Napoli 1950, 111-143. Gigante 1979 M.Gigante, Civiltà delle forme letterarie nell’antica Pompei, Napoli 1979. Giordano – Casale 1991 C.Giordano – A.Casale, Iscrizioni pompeiane inedite scoperte tra gli anni 1954-1978, «AAP» XXXIX (1991), 273-378. Goold 1988 G.P.Goold, A Paraclausithyron from Pompeii, in P.Knox – C.Foss (ed.), Style and Tradition. Studies in Honor of Wendell Clausen, Stuttgart-Leipzig 1988, 16-29. Harris 1989 W.V.Harris, Ancient Literacy, Cambridge-London: Harvard University Press 1989. Harrison 1991 S.J.Harrison, Vergil: Aeneid 10. With Introduction, Translation, and Commentary, Oxford 1991. Hawley 2007 R.Hawley, ‘Give me a thousand kisses’: the kiss, identity, and power in Greek and Roman antiquity, «LICS» VI.5 (2007), 1-15. - 26 - OVIDIAN GRAFFITI Hemelrijk 2004 E.A.Hemelrijk, Matrona Docta. Educated women in the Roman élite from Cornelia to Julia Domna, London-New York 2004. Keith 2006 A.Keith, Critical Trends in Interpreting Sulpicia, «CW» C (2006), 3-10. Lebek 1976 W.D.Lebek, Ein lateinisches Epigramm aus Pompei (vellem essem gemma eqs.) und Ovids Gedicht vom Siegelring (Am. 2,15), «ZPE» XXI (1976), 21-40. Linant de Bellefonds1994 P.Linant de Bellefonds, Pyramos et Thisbe, in Lexicon Iconographicum Mythologiae Classicae, VII (1994), 605-607. Lorenz 2007 S.Lorenz, Catullus and Martial, in M.B.Skinner (ed.), A Companion to Catullus, Malden-Oxford-Victoria 2007, 418-438. Luck 1977 P. Ovidius Naso, Tristia. Herausgegeben, übersetzt und erklärt von G.Luck, Band II: Kommentar, Heidelberg 1977. Mau 1889 A.Mau, Scavi di Pompei 1886-88. Insula IX, 7, «Mitt. des Kais. Deutschen Arch. Inst. Roem.» IV (1889), 3-31 + 101-125. Milnor 2009 K.Milnor, Literary Literacy in Roman Pompeii. The Case of Vergil’s Aeneid, in W.A.Johnson – H.N.Parkes (eds.), Ancient Literacies. The Culture of Reading in Greece and Rome, Oxford 2009. Monda 1993 S.Monda, Procedimento allusivo: Ovidio e CLE 2054, «InvLuc» XV-XVI (1993), 231-251. Munzi 1996 L.Munzi, Da Properzio a Ovidio. Un itinerario letterario nel pastiche di un anonimo pompeiano, «AION (filol)» XVIII (1996), 93-107. Nisbet – Hubbard 1970 R.G.M.Nisbet – M.Hubbard, A Commentary on Horace, Odes. Book I, Oxford 1970. Pepe 1974 L.Pepe, Un distico pompeiano e Ovidio, in Poesia latina in frammenti. Miscellanea filologica, Genova 1974, 223-234. Richlin 2006 A.Richlin, Marcus Aurelius in Love. Marcus Aurelius & Marcus Cornelius Fronto, Chicago-London 2006. - 27 - LUCA GRAVERINI Rodríguez-Pantoja 1994 M. Rodríguez-Pantoja, Notas sobre tipología literaria de lo “Carmina Latina Epigraphica”, «Helmantica» XLV (1994), 107-123. Sauron 2004 G.Sauron, La peinture pompéienne et la poésie augustéenne, «REL» LXXXII (2004), 144-166. Shackleton Bailey 1956 D.R.Shackleton Bailey, Propertiana, Cambridge 1956. Sogliano 1888 A.Sogliano, VII. POMPEI – Degli edificii recentemente scoperti, e degli oggetti raccolti negli scavi dal dicembre 1887 al giugno 1888, «Notizie degli Scavi di Antichità comunicate alla R. Accademia dei Lincei» VIII (1888), 509-530. Stevenson 2002 J.Stevenson, An Introduction to Epigraphic Poetry, in Churchill – Brown – Jeffrey 2002, 25-43. Stevenson 2005 J.Stevenson, Women Latin Poets. Language, Gender & Authority from Antiquity to the Eighteenth Century, Oxford 2005. Todd 1939 F.A.Todd, Three Pompeian Wall-Inscriptions, and Petronius, «CR» LIII (1939), 5-9. Varone 1994 A.Varone, Erotica pompeiana. Iscrizioni d’amore sui muri di Pompei, Roma 1994. Varone 2012 A.Varone, Titulorum graphio exaratorum qui in C.I.L. vol. IV collecti sunt imagines, 2 vol., Roma 2012. Wick 1907 F.C.Wick, Vindiciae carminum Pompeianorum, Napoli 1907. Wiseman 1985 T.P.Wiseman, Catullus and His World. A Reappraisal, Cambridge 1985. Woeckner 2002 E.Woeckner, Women’s Graffiti from Pompeii, in Churchill – Brown – Jeffrey 2002, 67-84. - 28 -