¹ For illustrations of the argument, see below, n. 6. I here use the increasingly controversial word ‘homosexuality’ only by way of reproducing common scholarly assertions regarding ancient culture. Otherwise I avoid the abstract: apart from the essentialising implications of the term that I find problematic (see Section IV), the word has an unpleasant history of marginalising uses with resonances of clinical pathology.

² Cf. Kroll, Wilhelm, ‘Römische Erotik’, in Siems, Andreas Kärsten, Sexualität und Erotik in der Antike (Darmstadt, 1988), p. 93Google Scholar: ‘Bei einer Sitte, die sich so leicht von selbst einstellt, auf fremden Einfluϐ zu schlieϐen, ist miϐlich’ (the article was originally published in Zeitschrift für Sexualwissenschaft und Sexualpolitik 17 [1930], 145–78Google Scholar); Richlin (1992), p. 223: ‘While it cannot be denied that erotic epigram has a Hellenistic form, the Greek and Asiatic influence on Roman pederasty must be seen as an augmentation, not as the basis.’

³ Veyne (1978), p. 50.

⁴ Veyne (1985), pp. 28–9.1 quote from the English translation published in the 1985 collection Western Sexuality. The original article, ‘L'homosexualité à Rome’, appeared in Communications 35 (1982), 26–33CrossRefGoogle Scholar. Veyne oversimplifies the relationship between Plautus and Greek culture; see Gruen (1990), pp. 124–57, for a more detailed discussion.

⁵ See, e.g., Housman, A. E., ‘Praefanda’, Hermes 66 (1931), p. 408n. 1Google Scholar; Gonfroy (1978); Boswell (1980), pp. 74–6; Lilja (1983); Kay, N. M., Martial Book XI: A Commentary (London, 1985), p. 127Google Scholar; Wiseman, T. P., Catullus and his World: A Reappraisal (Cambridge, 1985), pp. 10–11Google Scholar; Richlin (1992, 1993); Cantarella (1992).

⁶ Wilkinson, L. P., Classical Attitudes to Modem Issues (London, 1978), p. 136Google Scholar: ‘In the early Republic the Romans' attitude to homosexuality was that of most non-Greeks; it was a Greek idiosyncrasy which they despised …’; Balsdon, J. P. V. D., Romans and Aliens (London, 1979), p. 225Google Scholar: Cicero, believed that ‘homosexuality … could not have originated spontaneously in Rome; it must have been caught like an infection from abroad, from Greece’; Verstraete (1980), p. 227Google Scholar: ‘The pervasive Hellenization of Roman society in the second and first centuries B.C. mitigated the traditional hostility towards homosexuality and homosexual relations’; MacMullen (1982), passimGoogle Scholar; Jocelyn, H. D., ‘Concerning an American View of Latin Sexual Humour’, Echos du monde classique/Classical Views 29 (n.s. 4) (1985), pp. 13–14Google Scholar: ‘For the average upper-class Roman the reality [of “pederasty”] was something Greek and rather distasteful, even when the love object was a slave or a foreigner’ (although Jocelyn uses the term ‘pederasty’, he clearly means male homosexual behaviour as a whole); Dalla (1987), p. 10: ‘Il mos Graecorum diviene cosi sinonimo di omoerotismo e pederastia …’; Hallett, J. P., ‘Roman Attitudes toward Sex’, in Grant, M. and Kitzinger, R. (eds.), Civilization of the Ancient Mediterranean (New York, 1988), II.1274Google Scholar: ‘… Roman society of the second century B.C. thought of male homosexuality as a Greek custom’; Hallett (1989), p. 223Google Scholar: ‘republican Roman references to male same-sex love as a Greek import’.

⁷ Cantarella (1992), pp. 97–8. I quote from the English translation of her book, published four years after the original appeared.

⁸ This is the sole referent of the term ‘pederasty’ in this article.

⁹ For example, Edwards (1993), p. 94, writes of ‘Roman claims that homosexuality was a Greek practice which was adopted in the later republic by some Romans’. In 1991 I delivered a paper that began by arguing in some detail the point made by Veyne and Cantarella, and the audience's responses suggested to me that the notion of ‘Greek love’ had not in fact been laid to rest.

¹⁰ Cantarella (1992), p. 97.

¹¹ See above, n. 5. While the terms ‘active’ and ‘passive’ do not necessarily reflect either the psychological or the physiological realities of penetrative acts, they do accurately reflect the usual ancient representations of those acts. See further Edwards (1993), pp. 72–3.

¹² Veyne (1985), p. 29; cf. above, n. 3.

¹³ Cato fr. 212, 213, 222 Malcovati; Scipio fr. 17 Malcovati; Gracchus fr. 26–8 Malcovati.

¹⁴ Definite allusions to male homoerotic desire or activity include (following Marx's numeration) fr. 23, 296–7, 895–6, 967, 1058, 1138–42, 1186; probable allusions include fr. 173, 276–7 (cf. Apul, . Apol. 10Google Scholar), 418–20, 866–7; and possible allusions include fr. 72, 74, 278, 1267.

¹⁵ Quintilian (Inst. 10.1.100) notes the following: ‘togatis excellit Afranius; utinam non inquinasset argumenta puerorum foedis amoribus.’ Afranius fr. 32 Ribbeck (‘praeterea nunc corpus meum pilare primum coepit’) is clearly in the voice of an adolescent youth, and in light of Quintilian's comment, which is not directly substantiated by any other of the surviving fragments of Afranius, the context of this fragment may well have been homoerotic. It deserves mention that the togatae were not based on Greek originals, instead bringing Roman and Italian life directly to the stage.

¹⁶ Novius fr. 19, 20–21 Ribbeck.

¹⁷ Pomponius fr. 75–6, 125–6 Ribbeck. It is noteworthy that Pomponius dedicated an entire Atellan farce (the Prostibulum) to a male prostitute who services male citizen customers (see especially fr. 148–9, 151, 153 Ribbeck).

¹⁸ See, e.g., Asinaria 703–5, Captivi 867–8, Casino 437–514, 963–5, Curculio 33–8, 473–82, Mostellaria 722–3, Pseudolus 767–89, 1169–237, Rudens 1073–75.

¹⁹ So too Verstraete (1980), p. 232; Veyne (1985), pp. 28–9 (cited above); and Cantarella (1992), pp. 99–100. MacMullen (1982), p. 488, argues for precisely the opposite view.

²⁰ E.g. Livy 8.28 (on the lex Poetelia de nexis, which he dates to 326 B.C.; cf. Val. Max. 6.1.9); Val. Max. 6.1.6, 6.1.10, 8.1.abs.12, 8.1.damn.8.

²¹ τοῖς παλαιοῖς οἰκετν μν ρν ὥραν οὐκ ἦν ἄδοξον οὐδ' αἰσχρν ώς ὡς ἔτι νÛν αί κωμῳδίαι μαρτυροσιν· λευθέρων δ παίδων ἰαίδων πείχοντο. (Plut. Rom. Ques. 101.288A).

²² Cf. Cantarella (1992), p. 104: ‘To sum up, homosexuality in itself was neither a crime nor a socially reproved form of behaviour. Carrying on with a slave (so long as he did not belong to somebody else) was accepted as normal behaviour, as was paying a male prostitute.’ This attitude entailed the related assumption that the distribution of physical role was aligned with the power-differential between master an d slave, the master playing the active role and the slave the passive role. Allusions to a reversal of roles include Sen, . Epist. 47.7Google Scholar (a characteristically scandalized description of a slave wh o is ‘in cubiculo vir, in convivio puer’) and Martial 3.71 (a characteristically teasing jab: ‘mentula cum doleat puero, tibi, Naevole, culus, / non sum divinus, sed scio quid facias’).

²³ It is worth noting that a common word meaning ‘prostitute’, scortum, is neuter and thus lacks specific reference to the gender of the person denoted; he or she is a thing to be used. (See Walde-Hofmann s.v. for the etymology of the noun, which originally denoted ‘pelt’ or ‘hide’.) The key to acceptability in making use of prostitutes was to avoid the appearance of excess: see Hor. Sat. 1.2.31–5 and ps.-Acro ad loc. for a well-known anecdote regarding Cato the Elder. Another pertinent tradition concerning Cato involves his expulsion of L. Quinctius Flamininus from the Senate after a scandal involving a prostitute whom Livy identifies as a Carthaginian named Philippus (Livy 39.42–3; cf. Plutarch, , Cato Maior 17Google Scholar, Titus Flamininus 18). But as Boswell (1980), pp. 68–9, Lilja (1983), p. 31 n. 71, and Cantarella (1992), pp. 101–102, have pointed out, Philippus' sex is an irrelevant detail. Indeed, Livy himself indicates that there is an alternative tradition according to which the prostitute was female (cf. Val. Max. 2.9.3, Sen. Contr. 9.2), and another testimonium to the incident fails to specify the prostitute's sex and simultaneously makes quite clear the issues most important to Cato: libido and voluptas (Cic, . Sen. 42Google Scholar). Thus it is my argument that Cicero's comments on the use of meretrices at Cael. 48 (‘quando enim hoc non factitatum est, quando reprehensum, quando non permissum, quando denique fuit ut quod licet non liceret?’) could just as well apply to the use of male prostitutes.

²⁴ Veyne (1985), p. 28. He had made the point in his 1978 article as well: ‘Ce quiest surprenant n'est pas la bisexualité, mais son interdiction’ (Veyne [1978], p. 51).

²⁵ By way of illustration of this assumption, it will suffice to cite a few items of evidence taken almost at random. A graffito from the Domus Aurea in Rome reads as follows: ‘quisquis amat pueros, etiam sin(e) fine puellas, / rationem saccli non h(a)bet ille sui’ (Solen, Heikki, ‘Un epigramma della Domus Aurea’, RF 109 [1981], pp. 268–71Google Scholar; cf. Lilja [1983], p. 101). The close parallel with CLE 2153 from Rigomagus (Remagen) in Germany suggests that the notion expressed was something of a commonplace, and graffiti like these (as well as those scratched on the walls of Pompeii, for which see the selected list given at Boswell [1980], p. 57 n. 44) provide a glimpse at elements of society that were least likely to be phil-hellenising. Two widely different but equally Roman poets will serve to illustrate the prevalence of such assumptions in the literary sources. In the course of his vigorous attempt to make Epicurean philosophy meaningful and appealing to Roman readers, Lucretius represents those who are most likely to arouse sexual passion among men as being either boys or women (Lucretius 4.1052–7, ‘sive puer … seu mulier’), and in his programmatic introductory poem, Ovid describes the proper subject matter of Roman love elegy as ‘aut puer aut longas compta puella comas’ (Amores 1.1.20). For extended discussions see Boswell (1980), pp. 61–87; Lilja (1983); Richlin (1992); and Cantarella (1992).

²⁶ So Edwards (1993), p. 94.

²⁷ Adams (1982), p. 228. Cf. ibid., p. 123, and similar arguments offered by MacMullen (1982), p. 486, Hallett (1989), p. 223, and Edwards (1993), p. 94.

²⁸ ‘simplicius multo est “da pedicare” Latine / dicere’ (Priap. 3.9–10).

²⁹ For the phrase Latine dicere or Latine loqui see OLD s.v Latine (2.b). (Its application is not limited to sexual language: see, for example, Cic. Phil. 7.17.7–8.)

³⁰ MacMullen (1982), p. 486, claims that such borrowings as pedico, pathicus, catamitus, and cinaedus ‘had to be explained to contemporary readers as novelties’, but gives no evidence on these words, merely listing in a note further Greek words that in fact fail to support his argument. A concept that evidently did require explanation was philosophia: that the word itself needed to be glossed may be inferred from Cic, . Tusc. 1.1Google Scholar (‘studio sapientiae, quae philosophia dicitur’) and Sen, . Epist. 89.4Google Scholar (‘philosophia sapientiae amor est et adfectatio’), and the entire enterprise of philosophia is often associated with Greece (see Petrochilos [1974], pp. 186–96).

³¹ See further Boswell (1980), p. 28 n. 52, where he aptly compares fiancé.

³² It is worth noting that pedicare is not even restricted to homosexual anal intercourse. Martial three times applies th e verb pedicare to the anal penetration of a woman (11.78.5–6, 11.99.1–2, 11.104.17), as do a frankly detailed graffito found at Capua (CIL 10.4483 = Diehl 508) and (albeit obliquely) Priap. 3.7–8. For irrumare and fellare, also used in both heterosexual and homosexual contexts, see Krenkel, Werner, ‘Fellatio and irrumatio’, WZR 29 (1980), pp. 77–88Google Scholar; Richlin, Amy, ‘The Meaning of irrumare in Catullus and Martial’, CPh 76 (1981), pp. 40–46Google Scholar; and Adams (1982), pp. 125–34.

³³ One might have expected to find cevere being used to describe the motions of an anally receptive female or male equally, since the verb's original meaning is simply ‘to move the buttocks’ (cf. the grammaria n Probus' testimony: ‘cevere est clunes movere, ut in canibus videre est, qui clunes agitando blandiuntur’, GLK 4.37.8). Yet cevere is in fact only attested of males (see Adams [1982], pp. 136–7), but this may well be a result of the incomplete nature of our sources. One intriguing graffito from Pompeii reads ‘Quintio(s) hie futuit ceventes’ (CIL 4.4977). Commentator s assume that the ceventes are male and that futuere stands for pedicare (cf. Adams [1982], p. 119), but I see no reason to reject the possibility that the ceventes are women being penetrated a tergo, either vaginally or anally.

³⁴ See Ernout-Meillet and Walde-Hofmann s.v. There is in fact no Greek verb comparable in meaning to cevere.

³⁵ Adams (1982), pp. 145–9. For further examples see the discussion at Richlin (1992), p. 288.

³⁶ See the sources cited at OLD s.v. pullus. This colloquial usage is remarkably close to the modern gay slang ‘chicken’, and a man whom Ausonius calls felespullaria (Epigr. 77, pp. 340–41 Peiper) could today be called a ‘chicken hawk’. (For a possible parallel, cf. pullarius at Petr, . Sat. 43.8Google Scholar, although the word is a conjectural emendation of the MS reading puellarius).

³⁷ ‘pullus Iovis dicebatur Q. Fabius, cui Eburno cognomen erat propter candorem, quod eius natis fulmine icta erat. antiqui autem puerum, quem quis amabat, pullum eius dicebant’ (Festus 245.23–7). For details concerning this Fabius, see RE 111. Richlin (1992), p. 289, suggests that he was not literally struck by lightning but possessed some sort of birthmark on his buttocks.

³⁸ As Cantarella (1992), p. 102, observes, passages such as Polybius 31.25.2–5 bear on attitudes toward extravagance and luxury rather than ‘homosexuality’ (pace Edwards, [1993], p. 95 n. 104Google Scholar, although at pp. 176–8 she casts her discussion in terms of ‘luxury’). Other ancient sources describing the incident (for which see also Diod. 31.24 and Plut, . Quaest. conv. 4.4Google Scholar) clearly specify the object of Cato's displeasure: cf. Plut, . Cato Maior 8.2Google Scholar, κατηγορν δ τς πολυτελας; Diodorus 37.3.6, κατηγορν τς πιπολαζοὐσης ν τῇ Ῥώμῃ τρυϕς; Athenaeus 6.109 (274F–275A), δυσχραινε κα κεκράγει, ὅτι τινς τς ξενικς τρυϕς εἰσήγαγον εἰς τν. To use Athenaeus' terms, what Cato saw as foreign (ξεική) was not homosexuality but precisely the luxury (τρυϕή) exemplified by indulgence in expensive caviar, boy prostitutes, female prostitutes, etc. (Diodorus 37.3.5 adds costly wine and skilled chefs to the list.)

³⁹ See especially Dover (1978) and Halperin (1990) for the traditions and conventions of paiderastia.

⁴⁰ Valkenaer's, emendation of the MS reading Graecia to Creta has been adopted by Nipperdey (Berlin, 1879)Google Scholar, Malcovati (Turin, 1944), and Marshall (Leipzig, 1977), but it is a needless conjecture. Nepos' statement is applicable not just to Crete but to classical Athens, Sparta, Thebes, and other Greek city-states. The Loeb translator, J. C. Rolfe, prints Creta and offers a translation that obscures Nepos' point: ‘In Crete it is thought praiseworthy for young men to have had the greatest possible number of love affairs.’ On Cretan pederasty, see Dover (1978), pp. 185–96; Bremmer, Jan, ‘An Enigmatic Indo-European Rite: Paederasty’, Arethusa 13 (1980), p. 287Google Scholar(where Nepos is said to make an observation on the custom in ‘Crete’, with no indication that Creta is a conjectural reading found in no manuscript); Sergent, Bernard, Homosexuality in Greek Myth (translated by Goldhammer, Arthur, Boston 1986), pp. 7–39Google Scholar; and Koehl, R. B., ‘The Chieftain Cup and a Minoan Rite of Passage’, JHS 106 (1986), pp. 99–110CrossRefGoogle Scholar.

⁴¹ The tone with which Cicero writes of Dionysius and his love affairs is important. He takes this kind of sexual behaviour for granted, indeed appropriating the ‘Greek custom’ for his own purposes with no hint of disapproval. The fact that Dionysius was involved in erotic affairs with these young men functions merely as a further support for Cicero's argument: even though he loved them, he did not trust them. A telling parallel is to be found at Off. 2.25. Here he illustrates Dionysius' suspicious nature by means of a different example (due to his fear of knives, he had his hair singed rather than cut), but he adds the further instance of Alexander of Pheres, who did not trust his wife. Cicero treats Dionysius' love for his adulescentes no differently from Alexander's love for his wife, apart from the observation that the former is in conformance with mos Graeciae.

⁴² The line is preserved by Servius ad Aen. 10.325. Servius understands the words to apply to Greece in general, but the Loeb editor (Clinton Walker Keyes, 1928) suggests that this pertains specifically to Sparta, given what follows. On Spartan ‘licence’ in this regard, see also Martial 4.55.6–7 (‘aut libidinosae / Ledaeas Lacedaemonos palaestras’).

⁴³ ‘iuventutis vero exercitatio quam absurda in gymnasiis! quam levis epheborum ilia militia! quam contrectationes et amores soluti et liberi! mitto Eleos et Thebanos, apud quos in amore ingenuorum libido etiam permissam habet et solutam licentiam’ (Republic 4.4).

⁴⁴ Amor amicitiae refers to male friendships with philosophical airs but clearly they can include a sexual element, for Cicero later drily asks why no one ever loves an ugly youth (deformem adulescentem) or a beautiful old man, and points out that everyone knows what the poets mean when they speak of Jupiter and Ganymede or of Laius and Chrysippus (4.71).

⁴⁵ ‘quae omnia apud nos partim infamia, partim humilia atque ab honestate remota ponuntur’ (pr. 5).

⁴⁶ ‘ceterum abolitos paulatim patrios mores funditus everti per accitam lasciviam, ut quod usquam corrumpi et corrumpere queat in urbe visatur, degeneretque studiis externis iuventus, gymnasia et otia et turpes amores exercendo …’ (Tac, . Ann. 14.20Google Scholar). For a similar prejudicial view of Greek gymnasia as being associated with pederastic pursuits, cf. Plut, . Rom. Ques. 40.274DGoogle Scholar, quoted below, n. 71; and probably Sil. Ital, . Pun. 14.134–8Google Scholar: ‘ite, gregem metite imbellem ac succidite ferro … / pigro luctandi studio certamen in umbra / molle pati docta et gaudens splendescere olivo / stat, mediocre decus vincentum, ignava iuventus.’

⁴⁷ It is worth noting that the collocution turpis amor is elsewhere used of a relationship with a female prostitute (Hor. Sat. 1.4.111–112) and of the passion that might induce a woman to poison a man (Rhet. Herenn. 4.23).

⁴⁸ On the Roman resistance to nudity amongst male citizens, see also Cic, . Off. 1.129Google Scholar, De Orat. 2.223–224; Plut, . Cat. Maior 20Google Scholar, Rom. Ques. 40.274A–B.

⁴⁹ Aeschines (1.138–9) attributes to Solon Athenian traditions that prohibited slaves from being the lovers of free boys. Cf. Plutarch, , Amatorius 751BGoogle Scholar.

⁵⁰ For example, Lilja (1983), p. 112, at first appears to make the point argued here, but then inconsistently writes that Cicero's discussion in the Tusculans represents his ‘opinions on the origin of and different traits in homosexuality’ (p. 123; cf. also pp. 50, 107). So too Edwards (1993), p. 94 n. 99, citing Lilja, asserts that Cic, . Tusc. 4.70–1 is ‘on homosexual relations as a peculiarly Greek practice’Google Scholar. Cf. ibid., p. 96: ‘A taste for homosexual relations might be associated with the “Greek” literary life.’ I would argue that only a taste for pederastic relations might be so associated.

⁵¹ I quote from the published translation (Veyne [1985], p. 29)Google Scholar. Veyne himself wrote ‘quefcs Latins avaient en horreur’ (Veyne [1982], p. 28)Google Scholar.

⁵² Cantarella (1992), p. 97.

⁵³ Cf. Dalla (1987), p. 10, cited in n. 6 above.

⁵⁴ Verr. 2.1.66, ‘fit sermo inter eos et invitatio ut Graeco more biberetur’ (this refers to the Greek practice of προπνειν); Verr. 2.2.158, ‘… apud omnis Graecos hie mos est, ut honorem hominibus habitum in monumentis eius modi nonnulla religione deorum consecrari arbitrentur.’

⁵⁵ Balb. 12, ‘cum quida m apu d eos [sc., Athenienses] qui sancte graviterque vixisset testimonium publice dixisset et, ut mos Graecorum est, iurandi causa ad aras accederet …’

⁵⁶ De orat. 2.341, ‘ornandi caus a Graecorum more’; Tusc. 1.7, ‘… ut iam etiam scholas Graecorum more habere auderemus’; Tusc. 5.103, ‘aquam ferentis mulierculae, ut mos in Graecia est’; N.D. 3.84, in quibus quod more veteris Graeciae inscriptum esset “bonorum deorum” (cf. Val. Max. l.l.ext.3, ‘in his more Graeciae scriptum erat bonorum deorum eas esse’); Inv. 2.69, ‘cum Thebani Lacedaemonios bello superavissent et fere mos esset Graiis, cum inter se bellum gessissent, ut ii qui vicissent tropaeum aliquod in finibus statuerent victoriae modo in praesentiam declarandae causa, non ut in perpetuum belli memoria maneret, aeneum statuerunt tropaeum’; Verr. 2.1.66, ‘negavit moris esse Graecorum ut in convivio virorum accumberent mulieres’ (cf. Nepos, pr. 6—7).

⁵⁷ Livy 32.20.1, ‘sicut Graecis mos est’; 10.47.3, ‘translato e Graeco more’; 29.16.6, ‘ut Graecis mos est’ (cf. Pliny, , N.H. 11.251Google Scholar, ‘antiquis Graeciae in supphcando mentum attingere mos erat’); cf. 36.28.4–5, also on suppliants: ‘quae moris Graecorum non sint’.

⁵⁸ ‘sed eodem illo tempore Graeciae morem imitati verberibus animadvortebant in civis, de condemnatis summum supplicium sumebant’ (Sail, . Cat. 51.39Google Scholar).

⁵⁹ Front. Strat. 3.2.6, ‘in theatro, ubi ex more Graecorum locus consultationi praebebatur’ (cf. Tac, . Hist. 2.80Google Scholar); Festus 262.57, ‘rutrum tenentis iuvenis est effigies in Capitolio ephebi more Graecorum harenam ruentis exercitationis gratia’; Val. Max. 1.1.1, ‘Cererique, quam more Graeco venerari instituerent, sacerdotem a Velia…peterent’ (cf. Festus 237.11); Ov. Am. 3.13.27–28, ‘more patrum Graio velatae vestibus albis / tradita supposito vertice sacra ferunt’; Petr, . Sat. 111.2Google Scholar, ‘positum in hypogaeo Graeco more corpus’; Apul, . Met. 10.10Google Scholar, ‘nec rota vel eculeus more Graecorum tormentis eius apparata iam deerant’; Suet, . Nero 12.3Google Scholar, ‘instituit et quinquennale certamen primus omnium Romae more Graeco triplex, musicum gymnicum equestre, quod appellavit Neronia’ (cf. Tac, . Ann. 14.20Google Scholar, ‘quinquennale ludicrum Romae institutum est ad morem Graeci certaminis’ and Stat, . Theb. 6.5–7Google Scholar, ‘Graium ex more decus: primus Pisaea per arva / hunc pius Alcides Pelopi certavit honorem / pulvereumque fera crinem detersit oliva’); Serv, . ad Aen. 2.247Google Scholar, ‘licet Terentius Graeco more dixerit “agrum in his regionibus meliorem neque pretii maioris nemo habet”’; 3.691, ‘epitheton ad inplendum versum positum more Graeco’; 11.213, ‘more Graeco epitheton incongruum loco posuit’; Festus 293.33, ‘quam consuetudinem Ennius mutavisse fertur, utpote Graecus Graeco more usus, quod illi aeque scribentes ac legentes duplicabant mutas …’

⁶⁰ The gesture is thus intrinsically no different from Valerius Maximus' reference to προσκὐνησις as mos Persarum (4.7.ext.2) or from Sallust's allusion to Numidian customs in physical exercise as mos gentis illius (Jug. 6.1).

⁶¹ The reaction reported by Tacitus (above, n. 46) does not exactly constitute an exception, for it is directed at the importation of practices that are distanced as foreign (cf. accitam lasciviam and studiis extemis).

⁶² It is well known that Roman attitudes toward Greece were complex, ranging from lively distrust to outright scorn but generally coupled with a grudging admission of the great cultural achievements of the Greece of old. See, e.g., Petrochilos (1974), Gruen (1990), Gruen (1992), and Edwards (1993), pp. 92–7.

⁶³ ‘… quae tanta umquam in ullo iuventutis inlecebra fuit quanta in illo? qui alios ipse amabat turpissime, aliorum amori flagitiosissime serviebat, aliis fructum libidinum, aliis mortem parentum non modo impellendo verum etiam adiuvando pollicebatur’ (Cic, . Cat. 2.8Google Scholar).

⁶⁴ The consul Postumius opens his address to the assembled citizenry on a note of fearful warning with regard to foreign gods (39.15.2–3; he returns to the theme at 39.16.8). For other indications of Roman feeling against foreign religious practices, consider the incidents summarized in Val. Max. 1.3 (‘De Superstitionibus’) as well as one of the laws proposed by Cicero in his De Legibus: ‘separatim nemo habessit deos neve novos neve advenas nisi publice adscitos’ (2.19; cf. 2.25–6).

⁶⁵ Consider the phrase ‘stupra promiscua ingenuorum feminarumque’ (39.8.7; cf. 39.8.5–6, 39.13.10, 39.15.12). It is telling that Livy writes ‘ingenuorum feminarumque’ rather than, for example, ‘adulescentium feminarumque’. This underscores what is so disgraceful in the entire affair: not the initiates' gender but their free-born status. See Section III for further discussion of stuprum.

⁶⁶ See below, n. 84, and cf. the texts quoted in nn. 85–6.

⁶⁷ There is no reason to doubt that Juventius is free-born: cf. Neudling, Chester Louis, A Prosopography to Catullus (Oxford, 1955), pp. 94–6Google Scholar. In contrast with Catullus' liberties stands the fate of a certain Valerius Valentinus, who had been condemned by public opinion for having written a poem on his relationships with a free-born boy (puer praetextatus) and a free-born girl (ingenua virgo) (Val. Max. 8.1.abs.8; cf. RE 372, where the incident is dated to ca. 111 B.C.). Valerius Maximus offers no hint to the effect that the relationship with the puer was thought to be more ‘Greek’ than the relationship with the virgo.

⁶⁸ Indeed, it could be argued that of the two relationships it is the heterosexual one that is endowed with more of a Greek colouring: not only does he give Clodia, and not Juventius, a Greek pseudonym, but he appropriates a Greek lyric poem to represent what appears to be his first encounter with her (Catullus 51).

⁶⁹ Plaut, . True. 149–57Google Scholar; Prop. 2.4.17–22; Ovid, Ars 2.683–4; Juvenal 6.33–7.

⁷⁰ When Cantarella writes ‘the Greek vice’ (‘il vizio greco’) in inverted commas, one receives the impression that she is translating a Latin phrase. But a search of the PHI disk for viti- in conjunction with Graec- turns up no such phrase. And when vitium is indirectly attributed to Greeks, we read not of pederasty but of stylistic flaws (Cic, . De Orat. 2.18Google Scholar, ‘hoc vitio cumulata est eruditissima ilia Graecorum natio’; Sen. Rhet. Contr. 1.2.22, ‘hoc autem vitium aiebat Scaurus a Graecis declamatoribus tractum, qui nihil non et permiserint sibi et impetraverint’) or of the unnecessary and luxurious uses to which that decadent nation has put the gifts of nature (Plin, . N.H. 15.19 on olive oilGoogle Scholar: ‘usum eius ad luxuriatn vertere Graeci, vitiorum omnium genitores, in gymnasiis publicando’).

⁷¹ One exception has already been cited, Tac, . Ann. 14.20Google Scholar: ‘degeneretque studiis externis iuventus, gymnasia et otia et turpes amores exercendo.’ In the same vein, according to Plutarch (Rom. Ques. 40.274D). the Romans thought the gymnasia and palaestrae to be the most important factor in the ‘enslavement and softness’ (δουλεα and μαλακα) of the Greeks, since they gave rise to ‘much ennui and idleness, mischief, pederasty, and the ruin of young men's bodies by sleep, strolling about, rhythmic exercises, and strict diets’ (πολὐν ἂλυν κα σχολν ντεκοὐσας ταῖς πλεσι κα κακοσχολαν κα τ παιδεραστεῖν κα τ διαϕθερειν τ σώματα τν νέων ὒπνοις κα περιπάτοις κα κινήσεσιν εὐθμοις κα διαταις κριβέσιν). But it is worth noting that in both texts the primary emphasis is not so much on sexual issues per se as on questions of discipline (note Plutarch's μαλακα and Tacitus' otia; and Plutarch adds that the pursuits of the gymnasia and palaestrae were thought to induce young men to turn from the life of military discipline: ὐϕ' ὧν ἓλαθον κρυέντες τν ὃπλων κα καλο γαπήσαντες νθ' πλιτν κα ἱπέων γαθν εὐτρπελοι κα παλαιστρῖται κα καλο λγεσθαι.) Also significant is the unemphatic position of the reference to pederasty in each passage; it is represented as one among several associations of the gymnasium and palaestra.

⁷² Plaut, . Bacch. 742–3Google Scholar (‘atque id pollicetur se daturum aurum mihi / quod dem scortis quodque in lustris comedim, congraecem’), 812–13 (‘propterea hoc facio, ut suadeas gnato meo ut pergraecetur tecum, tervenefice’), Most. 22–4 (‘dies noctesque bibite, pergraecamini, / arnicas emite liberate, pascite / parasitos, obsonate pollucibiliter’), 64–5 (‘bibite, pergraecamini, / este, ecfercite vos, saginam caedite’), 959–61 (‘triduom unum est haud intermissum hie esse et bibi, / scorta duci, pergraecari, fidicinas tibicinas / ducere’), Poen. 602–603 (‘liberum ut commostraremus tibi locum et voluptarium, / ubi ames, potes, pergraecere’), Truc. 86–7 (‘peperisse simulat sese, ut me extrudat foras / atque ut cum solo pergraecetur milite’); Hor. Sat. 2.2.10–11 (‘si Romana fatigat / militia adsuetum graecari’) with Porphyrio ad loc. (‘aut luxuriari aut Graeco more ludere’).

⁷³ For levitas see, e.g., the sources cited in Petrochilos (1974), pp. 40–45 (a striking example is Cicero's defence of Flaccus, throughout which the orator draws on the prejudicial sentiment that Greeks are characterized by levitas). For Greek frivolity, hedonism', and self-indulgence see also Cic, . Pis. 22, 42, 67Google Scholar, Prov. Cons. 14, Verr. 2.2.7; Plut, , Cic. 5.2Google Scholar. For luxury see, e.g., Hor, . Epist. 2.1.93–102Google Scholar, and Edwards (1993), pp. 92–7.

⁷⁴ For more or less disparaging references to Graeculi, none of which contain any allusion to pederasty, see (for example) Cic. Verr. 2.2.72, 2.4.127, Red. Sen. 14, Sest. 110, Pis. 70, Scaur. 4, Mil. 55, Phil. 5.14, 13.33, De Orat. 1.47, 1.102, 1.221, Tusc. 1.86; Petr. Sat 88.10; Plin. Epist. 10.40.2, Pan. 13.5; Sen. Apocol. 5.4; Sen. Rhet. Suas. 1.6.16; Suet, , Tib. 11.1, 56.1Google Scholar, Claud. 15.4; Tac, . Dial. 3.4, 29.1Google Scholar.

⁷⁵ E.g., Livy 39.6.7–9, Pliny, , N.H. 34.14Google Scholar.

⁷⁶ Sail, . Cat. 13.3Google Scholar, ‘sed lubido stupri, ganeae ceterique cultus non minor incesserat; viri muliebna pati, mulieres pudicitiam in propatulo habere …’

⁷⁷ Juv. 3.109—12, ‘praeterea sanctum nihil †aut† ab inguine tutum, / non matrona laris, non filia virgo, nee ipse / sponsus levis adhuc, non filius ante pudicus. / horum si nihil est, aviam resupinat amici.’

⁷⁸ Paus. 4.1, ‘Argilius quidam adulescentulus, quem puerum Pausanias amore venerio dilexerat’; Ale. 2.2, ‘ineunte adulescentia amatus est a multis more Graecorum’.

⁷⁹ Pr. 1–3, ‘Non dubito fore plerosque, Attice, qui hoc genus scripturae leve et non satis dignum summorum virorum personis iudicent, cum relatum legent, quis musicam docuerit Epaminondam, aut in eius virtutibus commemorari, saltasse eum commode scienterque tibiis cantasse… non admirabuntur nos in Graiorum virtutibus exponendis mores eorum secutos.’

⁸⁰ Epam. 1.1, ‘de hoc priusquam scribimus, haec praecipienda videntur lectoribus, ne alienos mores ad suos referant, neve ea, quae ipsis leviora sunt, pari modo apud ceteros fuisse arbitrentur.’

⁸¹ See the Appendix for discussion of a manuscript variant (amore Graecorum) that would constitute an exception if it were the correct reading; I argue that it is not. One manuscript (Guelferbytanus 294) containing Cicero's discussion of Dionysius cited above (‘haberet etiam more Graeciae quosdam adulescentes amore conhmctos’, Tusc. 5.58) reads amore Graeciae instead of more Graeciae. But this is obviously a scribal error; the a was struck by the hand of a later corrector of the manuscript, and every other manuscript reads more, which is universally adopted by modern editors. Nonetheless, while having no ancient models, the phrase ‘Greek love’ is alive and well in modern scholarship. Consider, for example, the following titles: Meier, M. H. E. and de Pogey-Castries, L. R., Histoire de l'amour grec dans l'antiquité (Paris, 1930)Google Scholar; Crompton, Louis, Byron and Greek Love: Homophobia in 19th-century England (Berkeley, Calif., 1985)Google Scholar: MacMullen, Ramsay, ‘Roman Attitudes to Greek Love’, Historia 31 (1982): 484–502Google Scholar.

⁸² The word originally denoted any shameful behaviour, without specific reference to sexual behaviour (Festus 418.8—18), but the term came to narrow its focus, designating the specific disgrace of violating the sexual integrity (pudicitia) of free-born Romans. The notion thus reflects and reconfirms the ideal of the inviolability of the free Roman citizen (whether male or female) that sets him or her apart from non-citizens and from slaves. Recent discussions of stuprum include Dalla (1987), pp. 71–99; Fantham (1991); and Richlin (1992), pp. 224–5.

⁸³ Fantham, , by contrast, argues that ‘for our purposes it is useful to separate his heterosexual episodes from the homosexual, and to take them out of order so as to illustrate common elements in the handling of both categories of sexuality’ (Fantham [1991], p. 277)Google Scholar.

⁸⁴ See, e.g., Cic, . Phil. 3.31Google Scholar, ‘matres familiae, virgines, pueri ingenui abripiuntur, militibus traduntur’; Verr. 2.4.116, ‘adhibitam vim ingenuis, matres familias violatas’ (cf. Verr. 2.1.62 [stupra]); Fam. 5.10a.l, ‘qui tot ingenuos, matresfamilias, cives Romanos occidit, arripuit, disperdidit’; Livy 26.13.15, ‘nee dirui incendique patriam videbo, nee rapi ad stuprum matres Campanas virginesque et ingenuos pueros’; Rhet. Herenn. 4.12, ‘matribusfamilias et ingenuis sub hostilem libidinem subiectis’; Sail, . Cat. 51.9Google Scholar, ‘rapi virgines, pueros … matres familiarum pati quae victoribus conlubuissent’. Also relevant are expressions of alarm at the prostitution of the free-born that make no distinction between the two sexes. See, e.g. Cic, . Phil. 2.105Google Scholar, ‘ingenui pueri cum meritoriis, scorta inter matresfamilias versabantur’; Val. Max. 9.1.8, ‘lupanari enim domi suae instituto Muniam et Flaviam, cum a patre turn a viro utramque inclitam, et nobilem puerum Saturninum in eo prostituit.’

⁸⁶ Cic. Alt. 1.16.5, ‘iam vero (o di boni, rem perditam!) etiam noctes certarum mulierum atque adulescentulorum nobilium introductiones nonnullis iudicibus pro mercedis cumulo fuerunt.’ On this trial, see also Cic. Att. 1.18.3, Mil. 87; Val. Max. 9.1.7 (‘noctes matronarum et adulescentium nobilium magna summa emptae mercedis loco iudicibus erogatae sunt’); Sen, . Epist. 97.2Google Scholar (‘atqui dati iudicibus nummi sunt et, quod hac etiamnunc pactione turpius est, stupra insuper matronarum et adulescentulorum nobilium exacta sunt’); Dio 37.45–6.

⁸⁶ See, for example, Cic, . Cat. 3.1, 3.23, 4.2, 4.3, 4.24Google Scholar, Leg. Man. 66, Mil. 76, Phil. 14.9, 14.10, Verr. 1.14; Horace, , Odes 4.9. 23–4Google Scholar; Livy 39.15.14. The phrase makes an appeal not only to the ideally inviolable category ‘free-born’ but also to the marker ‘dependent’, ultimately bolstering the proprietary claims of the paterfamilias. It is not coincidental that this phrase makes reference to liberi, a word whose origin reveals the division of the familia into the slave and free dependants of the pater, the famuli and liberi respectively. See also Treggiari (1991), pp. 309–11.

⁸⁷ It is significant that many of the texts referring to stuprum with males emphasise the free-born status, and not the gender, of the passive partner: note the frequency of such qualifiers as ingenuus or nobilis in the texts cited in nn. 84–5 above. Treggiari (1991), p. 264, accurately notes that ‘the Augustan law defined all sexual intercourse with people of either sex who fell under the law [i.e., the free-born] as stuprum’, but proceeds to make the imprecise claim that the word was ‘in general use for any irregular or promiscuous sexual acts, especially rape or homosexuality’. Similarly the RE article on stuprum (RE 2.7 [1931], p. 423) imprecisely claims that the term denotes relations with free women on the one hand and all male homosexual acts on the other: ‘S[tuprum] ist demnach der unziichtige Geschlechtsverkehr mit einer freien, anständigen unverheirateten Frau oder von Personen männlichen Geschlechts untereinander …’

⁸⁸ I do not mean to overlook an obvious and crucial difference between stuprum with maidens and stuprum with young men: the latter infringed upon gender definitions in that it placed those who would one day be Roman citizens in the passive role associated with women. This kind of concern surfaces, for example, in the comments of Postumius, the consul of 186 B.C., as reported by Livy: ‘si quibus aetatibus initientur mares sciatis, non misereat vos eorutn solum, sed etiam pudet. hoc sacramento initiatos iuvenes milites faciendos censetis, Quirites? his ex obsceno sacrario eductis arma committenda?’ (39.15.13). So too Quintilian observes that stuprum is especially embarrassing to speak of in the courtroom when it is committed against males (Inst. 11.1.84, ‘illic maior aestus ubi quis pudenda queritur, ut stuprum, praecipue in maribus, aut os profanatum’). What is noteworthy, though, is that such expressions of concern over the forced feminisation of Roman men by means of stuprum occur so infrequently. Much more common is the representation—as in Cicero's letter to Atticus—of all stuprum, whether with maidens, boys, or women, as an undifferentiated offence.

⁸⁹ See, e.g., Hor. C. 3.6.17–32, Petr. Sat. 55.6.10–11, Juv. 3.45–6, Tac, . Hist. 1.2Google Scholar. Edwards (1993), pp. 34–62, provides a useful discussion.

⁹⁰ E.g., Fin. 2.27, Leg. 1.43, Off. 1.128.

⁹¹ Valerius Maximus begins his catalogue with Lucretia, bestowing on her the not insignificant praise ‘dux Romanae pudicitiae’ (6.1.1). In Hyginus' list of ‘quae castissimae fuerunt’ (256), Penelope is the first of several Greek examples, Lucretia the one and only Roman example. Martial uses Lucretia as a figure for the extreme of sexual purity on three occasions (1.90, 11.16, 11.104). And, quite strikingly, the author of the Declamationes Maiores attributed to Quintilian has the famous miles Marianus (who killed his superior after being propositioned by him) defend his action by citing Lucretia rather than, say, the young Marcellus who was propositioned by Scantinius (Val. Max. 6.1.7), as an example of the value placed by his Roman ancestors on chastity ([Quint.] Dec!. Maiores 3.11; cf. Calp. Flacc. 3). Petronius provides a boisterously humorous invocation of Lucretia that is likewise in an exclusively male context. When Ascyltus comes at night to make a sexual advance on Giton, Encolpius' boyfriend, he draws his sword with the words, ‘If you are Lucretia, you've found your Tarquinius!’ (‘si Lucretia es, Tarquinium invenisti’. Sat. 9.5).

⁹² Aen. 6.608–14, ‘hie, quibus invisi fratres, dum vita manebat, / pulsatusve parens et fraus innexa clienti, / aut qui divitiis soli incubuere repertis / nee partem posuere suis (quae maxima turba est), / quique ob adulterium caesi, quique arma secuti / impia nee veriti dominorum fallere dextras, / inclusi poenam exspectant’.

⁹³ It is clear from Cicero's defence of Caelius that the prosecution capitalised on similar feelings concerning Caelius' affair with the widowed Clodia. On univira, see Williams, Gordon, ‘Some Aspects of Roman Marriage Ceremonies and Ideals’, JRS 48 (1958), pp. 16–29Google Scholar.

⁹⁴ Aen. 10.324–7, ‘tu quoque, flaventem prima lanugine malas / dum sequeris Clytium infelix, nova gaudia, Cydon, / Dardania stratus dextra, securus amorum / qui iuvenum tibi semper erant, miserande iaceres …;’

⁹⁵ We might recall Cicero's similar treatment of Dionysius' love affairs, more Graeciae, with adulescentes (Tusc. 5.58; see above, n. 41).

⁹⁶ Makowski, John F., ‘Nisus and Euryalus: A Platonic Relationship’, CJ 85 (1989), pp. 1–15Google Scholar, stresses the representation of the pair as erastês and erômenos on the Greek model. I would add that the valorising distinctions inherent in the usual Athenian construct of paiderastia fade slightly with the Roman poet's remark that the two rushed into war side by side (‘panterque in bella ruebant’, 9.182), and practically disappear when the aged and dignified Aletes addresses the two as viri (9.252). In other words, although Euryalus is the junior partner in this relationship, not yet endowed with a full beard and capable of being labelled the puer (9.181), his actions prove him to be, in the end, as much of a vir—as capable of displaying virtus—as is his older lover Nisus.

⁹⁷ See Williams, Gordon, Technique and Ideas in the Aeneid (New Haven, 1983), pp. 205–7, 226–31Google Scholar.

⁹⁸ Aen. 9.446–449, ‘fortunati ambo! si quid mea carmina possunt, / nulla dies umquam memori vos eximet aevo, / dum domus Aeneae Capitoli immobile saxum / accolet imperiumque pater Romanus habebit.’

⁹⁹ Cases involving adultery: Sen, . Contr. 1.4, 1.7, 2.7, 6.6, 7.5, 8.3, 9.1Google Scholar; Calp. Flacc. 2, 31, 40, 48, 49; [Quint.] Declam. Minor. 244, 249, 273, 275, 277, 279, 284, 286, 291, 300, 310, 319, 330, 335, 347, 355, 357. Cases involving stuprum with young men: Sen, . Contr. 3.8, 5.6Google Scholar; Calp. Flacc. 3, 20; [Quint.] Declam. Minor. 279, 292.

¹⁰⁰ See especially Richlin, Amy, ‘Approaches to the Sources on Adultery at Rome’, in Foley, Helene B. (ed.), Reflections of Women in Antiquity (New York, 1981), pp. 379–404Google Scholar; Gardner, Jane F., Women in Roman Law and Society (Bloomington and Indianapolis, 1986), pp. 127–32CrossRefGoogle Scholar; Richlin (1992), pp. 215–19; Cantarella (1992), pp. 142–5; Edwards (1993), pp. 37–42.

¹⁰¹ On this notoriously shadowy law, see Boswell (1980), pp. 65–9; Lilja (1983), pp. 112–21; Fantham (1991), pp. 285–7; Cantarella (1992), pp. 106–19; and Richlin (1993), pp. 569–71. It is possible that the Augustan lex Julia applied not only to adultery but to stuprum as a whole and thus included pederasty in its purview, thereby reinforcing the lex Scantinia (so, for example, Richlin [1992], p. 224; cf. Digest 48.5.13, ‘ne quis posthac stuprum adulteriumve facito sciens dolo malo’ and Digest 48.5.35, ‘adulterium in nupta admittitur; stuprum in vidua vel virgine vel puero committitur’). But, although those later sources that do mention the lex Scantinia are maddeningly unclear as to its specific bearing, they obviously assume that it and the lex Julia, while penalising related offences, are distinct and that the lex Scantinia is still alive, albeit dormant (see, e.g., Suet, . Domit. 8Google Scholar, Juvenal 2.36–44).

¹⁰² Forster, E. M., Maurice (New York: W. W. Norton and Company, 1971), p. 51Google Scholar. Vivid confirmation of a similar association in French popular belief is provided by colloquial insults of the type va tefaire voir chez les Grecs, wishing anal penetration on their recipients.

¹⁰³ On this immense question see, for example, Richlin (1992) passim and Edwards (1993), pp. 47–58.

¹⁰⁴ This complicated issue is receiving an increasing amount of attention in studies of sexual experience in classical antiquity. Important contributions to the debate include Foucault, Michel, The History of Sexuality (especially Volume 2, The Use of Pleasure [New York, 1985] and Volume 3, The Care of the Self [New York, 1986])Google Scholar, Halperin (1990), Boswell (1990), and Richlin (1993). Here I offer a summary of my views; I will discuss the question in greater detail in a revised form of my dissertation, which I am currently preparing for publication. It should be noted that scholars have detected a gradual, deeply-rooted transformation in the realm of sexual morality over the course of the 1st and 2nd centuries A.D.: see Veyne (1978); Boswell (1980), pp. 119–36; Rousselle, Aline, Porneia: On Desire and the Body in Antiquity (trans. Pheasant, Felicia, Oxford 1988)Google Scholar; Foucault, Michel, The Care of the Self (trans. Hurley, Robert, New York 1986)Google Scholar; Boswell (1990), pp. 73–5; Cantarella (1992), pp. 187–210. This new movement, that ultimately meshed with Judaeo-Christian holiness codes to produce a radically different sexual ethic that came to dominate western Europe, emphasized an ascetic approach to the body and thus to sexual behaviour, and reached its logical conclusion in the problematisation of all sexual activity that did not lead to procreation and concomitantly the organisation of homosexual behaviour as a discrete and disapproved realm within the larger area of sexual experience. But this mode of conceptualising sexual experience stands in marked and illustrative contrast with the traditional ethic that was inherited by the audiences and readerships of Plautus, Catullus, and Cicero, with which this essay is concerned.

¹⁰⁵ Hence, as Richlin has noted (Richlin [1992], pp. 91–3, 139; Richlin [1993], pp. 533, 549 with n. 61), we encounter a number of allusions in ancient sources to the possibility that cinaedi might also be womanizers or adulterers. It seems to me (pace Richlin) that the tone of these references is casual and provides no hint that a womanizing cinaedus was thought to be an anomaly. See further Edwards (1993), pp. 63–97.

¹⁰⁶ Cf. Veyne (1985), p. 30, for a similar view. Richlin (1993) argues to the contrary: countering the claim that ‘homosexuality’ is ‘so much a modern production that nothing like it can be found in classical antiquity’, she sets out to demonstrate that ‘”homosexual” in fact describes, in Roman terms, the male penetrated by choice’ (Richlin [1993], p. 526Google Scholar; cf. ibid., p. 530: ‘I aim to show, first, that men identified as homosexuals really existed at Rome and, second, that their existence was marked both by homophobia within the culture and by social and civil restrictions.’) As an initial response, I would register the logical difficulty I find in attributing ‘homophobia’ to a cultural tradition that (as Richlin herself emphasizes throughout her work) simply assumed as part of human nature the fact that men desire and engage in penetrative acts with boys. To be sure, the ancient sources bespeak an antipathy against cinaedi, who are represented as desiring to be penetrated by men; but to describe this as ‘homophobic’ or as a problematisation of ‘homosexuality’ is as unhelpful as it would be to describe ancient criticism of female prostitutes or adulteresses as ‘heterophobic’ or in some way problematising ‘heterosexuality’.

¹⁰⁷ For Roman representations of female homoerotic experience, see Hallett (1989). Occasionally an ancient author draws a parallel between male-male and female-female sexual behaviour: at Juvenal 2.49–50, for example, Laronia notes that no woman performs cunnilinctus on another, whereas countless men fellate and are penetrated by other men. But I would note the typical Roman emphasis on activities in this passage and the concomitant problematisation of passive males only, and not their active partners. Laronia's statement, I would argue, is significantly different from a twentieth-century claim that ‘There are many gay men, but no gay women’. Furthermore, I find it telling that Ovid's Metamorphoses presents its Roman readers with a tortured rejection of female homoerotic desire in the course of the tale of Iphis (9.731–4) yet simply assumes the normalcy of male homoerotic desires (e.g., in the tales of Apollo and Hyacinth [10.162–219] and Echo and Narcissus [3.339–510]). The same poet elsewhere observes in passing that he personally prefers women on practical and admittedly altruistic grounds, but his phrasing is significant: ‘odi concubitus qui non utrumque resolvunt; / hoc est cur pueri tangar amore minus’ (Ars 2.683—4). He does not feel it necessary to disclaim all interest in boys (‘minus’) and certainly does not attempt any general problematisation of homosexual eros.

¹⁰⁸ Cf. Suet, . Claud. 33Google Scholar, Mart. 11.87.

¹⁰⁹ Guillemin (Paris, 1923), Malcovati (Turin, 1944) and Marshall (Leipzig, 1977) print more; Nipperdey (Berlin, 1879) and Winstedt (OCT, 1904) print amore. Rolfe, John C. (The Lives of Cornelius Nepos, 3rd ed., Boston, 1894) simply omits this and the following sentence. Petrochilos (1974), p. 181Google Scholar, quotes Nepos' text with amore, but gives no indication of the textual problem; and he imprecisely frames the question in terms of ‘homosexual love’. Likewise the OLD s.v. Graecus simply quotes this passage with amore. For scribal confusion of more and amore we might compare PI. Mil. 1377 (‘<a> mores meos’) and Cic. An. 6.1.12 (‘in <a> moribus est’), although in these cases the more likely reading is amor.

¹¹⁰ Marshall in the apparatus to his 1977 Teubner edition adduces Cotton 3.2, ‘ex more Persarum’, as a parallel from within Nepos' text.

¹¹¹ For their helpful comments and suggestions at various stages of this article's development I wish to thank Ralph Hexter, Gordon Williams, Edward Harris, Panayotes Dakouras, and the editors of CQ as well as the anonymous referee.