1. For the distinction between the disclaimer of Martial (1.4.8), which in the context of the more repressive atmosphere of the empire certainly did involve a distinction between life and art, and this passage of Catullus, see Citroni 1975, 32-33. The distinction here is between what is appropriate to the bard or epic poet (pium poetam ipsum, 5-6) and what is appropriate to the writer of versiculi . Richlin 1992a, 12-13, 145-47, has excellent discussions of this poem, pointing out that the separation of life from work is not the issue at hand and stressing the importance of Catullus' ambivalent sexual pose in the poem.

2. Richlin 1992a and Selden 1992 both make excellent cases for a sexual model of the relationship between reader and poet in Catullus. Richlin's book describes the priapic model in Roman humorous poetry (mainly invective and satire), and deals with Catullus specifically on pages 144-56, and Selden discusses Catullus' game with the reader in a framework that is quite similar to mine, though in his version the game is played on the field of interpretation.

3. A selection of English imitations of these poems can be found in Duckett 1925, 10-25.

4. The poems of Catullus are quoted according to the text of Mynors 1967.

5. On Catullus' use of this word, see Seager 1974. The combination of erotic and aesthetic senses in this word is best illustrated by c.35.13-18.

6. The OLD covers all of these meanings with appropriate citations, but I will cite a few passages mainly to illustrate the negative or questionable connotations of this word. The basic notion of "inessential/luxury" is supplied by Lucretius' lines: "Navigia atque agri culturas moenia leges / arma vias vestis [et] cetera de genere horum, / praemia, delicias quoque vitae funditus omnis, / carmina picturas, et daedala signa polire" (5.1448-51). Compare Seneca, Benef. 4.5.2: "Neque enim necessitatibus nostris provisum est, usque in delicias amamur." Cicero tells Atticus that he won't be attending the Games because he wishes to avoid "omnem deliciarum suspicionem'' ( ad Att. 2.10). Martial complains to a sluggish reader, ''Quid prodest mihi tam macer libellus, / nullo crassior ut sit umbilico, / si totus tibi triduo legatur? / numquam deliciae supiniores" (2.6.10-13). Quintilian distinguishes those readers who only consider the old poets, with their natural eloquence and manly strength, to be worth reading from those who are delighted by "recens haec lascivia deliciaeque" ( Inst. 10.1.43). For the meaning "airs" or "caprices," there is Cicero's "ecce aliae deliciae equitum vix ferendae!" ( ad Att. 1.17.9). Finally, for a usage quite close to that of Catullus 2, compare Martial's reference to the "delicias lususque iocosque leonum" (1.14.1), the miraculous playing of the lions with the hares at the Games, where the lions tease the hares by catching them in their mouths and then letting them go. Bayet 1953, 20, has described this word very well in a passage that is worth quoting in full:

Another, equally notable, aspect of the Catullan position is expressed by the word deliciae , which covers all the games of love, psychic and physical. It has a nearly unbelievable depth of suggestion, grouping meanings as different as "kindnesses" and "amusements," "caprices" and "enjoyments"; but with the etymological implication of a "temptation" and a pleasurable "seduction." So that the word, titillating in itself, represented both the opinion of commmon morality and a provocative attitude with respect to it.

7. Delicatus may not be etymologically related to deliciae ; the OLD gives lacio (entice) as the root of deliciae but gives the etymology of delicatus as dubious. To the Roman mind, though, these words were clearly related; Cicero, for instance, says that, because he wishes to avoid "omnium deliciarum suspicionum," it is inappropriate for him to appear "non solum delicate sed etiam inepte peregrinantem" ( ad Att. 2.10).

8. When Catullus spends the day playing poetic games with his friend Calvus, it is because they had agreed to be delicati (c. 50.3), and when Catullus explains to Mallius that his brother's death has chased from his mind all interest in love poetry, the expression he uses is "all toys ( delicias ) of the mind" (c.68.16).

9. See Newman 1990, 30-32, on nugae and mime.

10. Chapter 2 of Edwards 1993 is an excellent discussion of the range of associations and moral issues engaged by mollitia (softness, effeminacy). The identification of sexual effeminacy with homosexuality in modern times has led several scholars to see the versiculi criticized by Furius and Aurelius in c.48, where Catullus tells Juventius of his insatiable desire to kiss him, rather than in cc.5 and 7. Quinn 1970, 145, comments: "Can anyone doubt, after reading Poems 5 and 7, that Catullus is a man?" But for the Roman there was no incompatibility between effeminacy and heterosexual activity. Seneca's diatribe against Maecenas ( Ep. 114.4, 6), in which he accuses him of effeminacy of literary style as of lifestyle, refers, in the midst of a host of other signs of effeminacy, to the fact that Maecenas was a man "qui uxorem miliens duxit, cum unam habuerit," and this immediately after citing the fact that he was accompanied by two eunuchs, ''magis . . . viri quam ipse"!

11. Probably the most important of the polymeric poems in this connection are cc. 6, 10, 12, 17, 21, 25, 37, 42, 47, and 56.

12. As Veyne (1978, 55) points out, the old Roman sexual morality was a puritanism just as much as was the Christian, only the Roman was a puritanism of virility. The immensely popular pantomime was looked at askance by the moralists because, as Veyne puts it, "the charm of the dance and of song made these men of state and these thinkers tremble, so that they opposed to it the spectacles of the gladiators, more virile, more educational for the citizen" (54).

13. Much of my discussion of this poem focuses on passages and issues that have received attention from previous scholarship, but I treat them in different contexts and from different angles. For an excellent presentation of the main stream of scholarship on this poem, see Latta 1974.

14. Publication, like any other major decision in the life of an upper-class Roman, was undertaken after a consultation with a friend or friends. Several of the letters of Cicero and Pliny reflect this practice; see, for instance, Cicero, ad Att. 4.8a.3, and Pliny, Ep. 9.25, 2.10.

15. This effect of studied casualness was much sought after in Catullus' time: Cicero, for instance, recommends that the orator adopt a "neglegentia diligens" in the use of short phrases ( Orator 78).

16. The lightness of the poet's relation to his own work, and the invitation to audience members to make of it what they will (perhaps more than the poet does himself), are constitutive elements of what later becomes the genre of nugae or ineptiae . Pliny, sending a friend his "hendecasyllables," cuts short his introduction of his work with the words:

Sed quid ego plura? Nam longa praefatione vel excusare vel commendare ineptias ineptissimum est. Unum illud praedicendum videtur, cogitare me has nugas ita inscribers "hendecasyllabi," qui titulus sola metri lege constringitur. Proinde, sive epigrammata sive idyllia sive eclogas, sive, ut multi, poematia seu quod aliud vocare malueris, licebit voces; ego tantum hendecasyllabos praesto (4.14.8-9).

17. Richlin 1992a, 162, speaks of Catullus in this poem as a literary Priapus who "pimps his book."

18. The fullest exposition of this poem as an Alexandrian manifesto is given by Cairns 1969. See also Wiseman 1979, Chapter 11. An excellent corrective to the usual focus on the learned, philological aspect of Roman Alexandrianism is provided by Hutchinson 1988, Chapter 6, and especially pages 282-83, where Hutchinson points out the sensuous and seductive aspects of the Roman poets' descriptions of their own work: "The poets fit Callimachus into their own aesthetic language, and use that language to develop their own conception of their poetry, a conception far richer and more insidious than Callimachus' overt conception of his" (283). Hutchinson is speaking here of the elegists, but what he says is also applicable to Catullus.

19. See Latta 1974, 209.

20. Putnam 1982, 32 (especially note 1), has an interesting discussion of words meaning "smooth" or "polished" in Latin. He cites a number of passages that castigate effeminates for their use of the pumex , but does not seem to think this nuance of any importance in Catullus. Seneca's "Ite nunc et in istis vulsis atque expolitis et nusquam nisi in libidinibus viris quaerite oratores" ( Contr. 1, praef. 10, emphasis mine) tells us something about the implications of Catullus' "pumice expolitum,'' and to this one might add Persius 1.85-87 (emphasis mine): "Pedius quid? crimina rasis / librat in antithesis, doctas posuisse figuras / laudatur: 'bellum hoc.' hoc bellum? an, Romule, ceves?" Richlin 1988, 168, 188-89, has some useful material on the theme of male depilation in Roman literature.

21. "To tease" is one of the meanings of delicias facere ; compare Plautus, Cas. 528, Poen. 296. Compare also Martial 1.14.1.

22. The view that Catullus is making an invidious, Alexandrian comparison between Nepos' huge work and his own slim but polished volume is no longer held by most scholars. Three volumes do not in fact constitute a large work, and Nepos seems to have been something of a neoteric himself (Wiseman 1979, Chapter 10). The word laboriosis (7), however, does suggest that he is teasing Nepos a bit; it is, as Quinn points out in his commentary, "a word that suggests going out of one's way to make work for oneself (90); compare Cicero, Cael. 1, "vos laboriosos existumet."

23. The fact that the Muse is addressed as patrona also has the effect of withdrawing the book from Nepos. Some scholars have suggested emending what is evidently a corrupt text in this line to give the expected reference to Nepos as patronus (e.g., patroni ut ergo), but see the arguments against this in Gold 1987, 56-57. I disagree with Gold only in that I see this reference to the patron Muse as a deliberate replacement of the more usual human patron.

24. It has been argued that the sparrow would have been understood by the Romans as a phallic symbol. The evidence for this seems good enough that a suspicion of double entendre should enter the mind of the reader, but this does not mean that the whole poem becomes an elaborate allegory in which every statement has a specific sexual translation. Thomas 1993 is the most recent to put the case for the double entendre and he does so very judiciously. For the most part, those who have seen a sexual meaning in the sparrow have taken the allegorical path; see Genovese 1974, 121-25; Giangrande 1975, 137-46; Nadeau 1980, 879-90 (discreetly written in Latin!). Against the first two of these, see Jocelyn 1980, 421-24.

25. Catullus' confusion of surface and depth in c.2 is nicely illustrated by his use of the word solaciolum (a little consolation, 7). On the one hand, this word belongs to the psychological background of Lesbia's behavior: we watch Lesbia play with the sparrow and conclude that she is sublimating her desire in this game. But the grammar and form of the word assign it to the surface of the scene: Lesbia plays "carum nescioquid" (some dear thing), and she plays a solaciolum , the diminutive emphasizing the delight of the observer in what he sees. The potential narrative of frustration and consolation behind Lesbia's game is blocked by the diminutive form of the word that suggests it, and replaced by a voyeuristic delight. The same is true of the "little golden apple" (aureolum malum, 12) whose narrative potential in the myth in which it brings Atalanta's virginity to an end is blocked by the use to which it is put in the simile, where it illustrates Catullus' desire to play with the sparrow.

26. The Alexandrian poets represented Atalanta as having already fallen in love with Hippomenes and so welcoming defeat. They played up the symbolism of the apple as a love token (Fordyce 1961, 91). Pearcy 1980, 157-62, who supports the integrity of 2a and 2b, emphasizes this interpretation of the Atalanta story in order to argue that, for Catullus to see Lesbia playing with the passer , and therefore to know that she is in love with him, is as pleasing as was the apple that declared Hippomenes' love to Atalanta. Pearcy has to transpose 9-10 to the end of the poem in order to get around the awkward fact that it is playing with the passer that would be pleasing to Catullus.

27. Copley 1949, 22-40, is a classic statement of the thesis that Catullus' experience of love was on a level that could not be reciprocated by the depraved and trivial Lesbia; for an application of this model to c.2, see Williams 1968, 140-43; Bignone 1945, 360; Buechner 1977, 66; Lyne 1980, 51-52.

28. This poem presents a number of textual problems of which the only one that affects my reading is the question of the relation of 11-13 to the rest of the poem. In the ancient manuscripts, 1-13 are presented without a break, and the modern tradition of printing 11-13 as a separate fragment (like 14b, a clearer case) was begun by the renaissance editor Guarinus. Grammatically, there is nothing impossible about taking possem in line 9 as conditional with "tam gratum est" as apodosis (could I but play with you, it would be as welcome). Fordyce 1961, who cites this solution of Ellis's (91) along with Ellis's parallels for the indicative in the apodosis (Plautus, Poen. 921, Martial 2.63.3), rejects this interpretation because of the inappropriateness of the simile that would result. This supposed inappropriateness seems to be the main stumbling block to taking 1-13 as a single poem. Quinn, who has no qualms about changing "cum . . . acquiescet" in line 8 to "tum . . . acquiescat," describes Vossius' conjecture posse for possem in line 9 as "a brilliant attempt to weld 1-10 to 11-13 '' but "rather too good to be true" (95).

29. Most scholars have noticed that the language Catullus uses of the effects on him of Calvus' lepos is erotic (Finamore 1974, 11-19, especially the citations on 11-12). Others have obviously felt uncomfortable with the homosexuality of this passage, most notably Clack 1976, 50-53, who claims that the erotic language is not addressed to Calvus at all but derives from the fact that Calvus and Catullus had spent the afternoon writing about Lesbia; "Calvus had a number of amusing things to say about Roman ladies in general and Clodia especially" (52). This totally gratuitous diversion of the expression of the erotic effect of one man on another to lockerroom conversation between two men about a woman shows what it takes to save Catullus' honor!

30. Aulus Gellius' anecdote (19.9.7) about the Greeks who taunted Antonius Julianus that the Romans, with their barbarous language, had no erotic poetry of any grace apart from Catullus and Calvus is a good indication of the atmosphere in which upper-class Romans produced homosexual verse. Julianus quotes, by way of response, epigrams by Valerius Aedituus, Porcius Licinus, and Quintus Catulus, all on homosexual themes.

31. The question of Roman attitudes to homosexuality is controversial; MacMullen 1982 disagrees with scholars like John Boswell and Veyne, who have seen a tolerance of homosexuality in ancient Rome. The view that "relations with boys, provided that they were not ingenui , were both very common and very lightly viewed" has been supported more recently by Griffin 1985, 25-26. Clearly, no single answer can be given to the question of the Roman attitude to homosexuality; the various statements cited by scholars have to be considered in their context, for at Rome it very much depends on who is talking to whom and in what context. Cicero, for instance, expresses very different attitudes to augury in different works. For our purposes, MacMullen's quotation of Minucius Felix (28.10) to the effect that Romans tolerated all sexual license as urbanitas is most apposite. Wiseman 1985, 3-4, reminds us that the "The 'Julio-Claudians' were actually Iulii Caesares and Claudii Nerones, wealthy aristocrats who behaved as such," and that the attitudes of the Roman elite did not suddenly change with the end of the Republic. The sexual attitudes and practices of the emperors, amply documented by Suetonius, may tell us something about those of their republican forbears. Most recently, Richlin 1993 has presented the evidence for homophobia at Rome in a fascinating study of the cinaedus that has an excellent discussion of recent controversies about ancient sexuality.

32. A whole book of Cicero's letters, Book 13, consists of epistulae commendaticiae in which Cicero praises the commendee and stresses the closeness of his relationship with him, and how important it is for Cicero that the addressee should look out for the interests of the commendee. The particular kind of commendatio in Catullus' poem is best illustrated by Cicero, Pro Caelio 17, where he speaks of how Caelius' father entrusted (commendavit et tradidit) the young Caelius to Cicero as a mentor. For commando meaning "render agreeable or attractive," see OLD s.v. 6.

33. Compare castum and pudice in 15.4-5 to "parum pudicum" and castum in 16.4-5.

34. Veyne 1988, 18-19.

35. This pair features in one other poem of Catullus (c.11), where their long-winded protestations of friendship are answered with a request to take a short and unpleasant message to Lesbia. In c.16 the duo appear as readers who have judged Catullus from his versiculi (16.3), and again a problem arises from the fact that Catullus must entrust his words to them.

36. My interpretation of pruriat and "movere lumbos" as referring respectively to the sexual arousal and wiggling buttocks of the pathic and not, as is more usual, to erections is in agreement with Wiseman 1987, 222-24, and Buchheit 1976b, 342-44. Both authors provide ample parallels for these meanings. See also Selden 1992, 485, and especially note 117 on cevere . For the sexual stimulation caused by effeminate poetry, see Aristophanes, Thesm. 130-34, and Persius 1. 19-21, both of whom locate the excitement in the anus. For movere lumbos used of the movement of the pathic's buttocks, see Vergil, Cat. 13. 21, and for prurire used of the sexual excitement of the pathic (usque ad umbilicum), see Martial 6.37.3. Notice the use of prurit in this sense in the graffito quoted below ( CIL 4.2360). Richlin 1992a, 248, note 9, remains unconvinced by Buchheit and holds to the traditional interpretation of prurit as active rather than pathic.

37. CIL 4.2360; compare 13.10017. On this graffito, see Housman 1931. Housman interprets prurit in line 2 to mean the same as I have argued it does in Catullus 16.9. Svenbro 1988, 207-218, has recently studied the Greek parallels to this graffito and argues that pederasty is one of the models through which the Greeks understood the relation between writing and reader; according to this model, the reader is cast as the eromene * .

38. For a similar analysis of the relation between reader and author in Catullus 16, see Selden 1992, 485; Batstone 1993, 150-55, is an interesting study of the relation between literal and figurative in this poem that also raises the issue of the position of the reader. An extraordinary passage from Seneca's Epistles (46. 1-2), in which Seneca describes his encounter with a text of Lucilius, is cited by Habinek 1992b, 196-97, and it provides a fascinating example of the use of a sexual model for reading. The passage begins with Seneca the reader exploring a text that is levis (smooth), unlike his own body (and Lucilius'). But drawn on by the text's "sweetness" (dulcedo), Seneca gradually comes to take the feminine position in relation to a text that is "virile", "big," and "taut," and which he swallows whole (exhausi totum)! The process is exactly the same as in Catullus 16, where the effeminacy of the text paradoxically penetrates the reader, except that in this case the reader is actually rejoicing in his passivity.

39. Cato's words about the old Roman attitude to poetry convey an attitude that never completely disappeared: "Poeticae artis honos non erat; si quis in ea re studebat aut sese ad convivia adplicabat, grassator vocabatur" ( Mor. 2). Cicero's equation of words like urbanus, festivus , and elegans with effeminacy in his attack on Clodius ( In P. Clod. et C. Cur. , frag. 22) is a good example of the questionable reputation of these fashionable social values. As Edwards 1993, 96, points out, elite Romans of the late Republic and early principate had to perform a balancing act between urbanity, which always ran the danger of tipping over into effeminacy, and rusticity, which gave virility a bad name.

40. Pace Quinn 1970, 143, who cites c.48 and other homosexual verse. To the Roman mind, a partner of the opposite sex did not preclude effeminacy (see note 7).

41. Petronius' impotent Encolpius makes a distinction between kissing and real sex when he describes his dalliance with Circe as follows: "In hoc gramine pariter compositi mille osculis lusimus, quaerentes voluptatem robustam" (127).

42. Inambulatio would seem to be a technical term for the pacing of an orator; see Rhet. Her. 3.27 and Cicero, Brut. 158. Argutus is usually a positive term for speakers and poets ( OLD s.v. 2. and 6b), but it can also mean "garrulous" ( OLD s.v. 4).

43. The Greek word cinaedus that the Romans used for "pathic" also means "dancer."

44. See Veyne 1979, 16-19, on the problems the Romans had accepting Hellenism and the cultural indulgences with which it was associated. Though Hellenism was always seen as a threat to the Roman duty of "virile tension," it eventually became acceptable to the Romans, even if their ambivalent attitude is reflected in the pervasive theme of the "senator whose private life is nothing but soft but who shows himself nevertheless to be a man of energetic action" (19).

45. On Naevius and this exchange with the Metelli, see Jocelyn 1969.

46. Laevius (frag. 13 in Courtney 1993) compares the writing on a sheet of paper to a back marked by tattoos or a beating: "fac papyrin . . . haec terga habeant stigmata." Unfortunately, we do not know the context of this line.

47. On the status of the actor at Rome, see Edwards 1993, 123-26. Catullus' contemporary (and the dedicatee of his book) Cornelius Nepos declared that to exhibit oneself on stage to the people is considered by the Romans to be inconsistent with respectability, even though the Greeks did not find it a source of shame (quoted by Edwards 1993, 98). Dupont 1985, 95-110, argues that for the Romans the problem with the actor was that his self-display and his creation of a persona were dangerously close to the theatrics of the political arena in which the Roman elite acted their virtus to and for the people. This is confirmed by Richlin 1992b, who has shown how close the orator came to the actor in his gestures, use of voice, and even application of makeup; she also shows how this proximity was a considerable source of anxiety to Roman rhetorical theorists striving to distinguish the two: the orator must avoid being too dry and jejune while maintaining a clear boundary between his effects and the effeminate antics of the actor. In contrast, we have stories of Roman nobles dancing at their own banquets: according to Velleius Paterculus (2.83.2), Lucius Plancus danced the role of the sea god Glaucus, naked except for a fishtail and blue paint. Wiseman 1985, 47, who cites these stories (note III), comments, "When admiration of the professional's art came to mean more than the dignity of his own status, at a private party the sophisticated Roman could let his hair down and act as a mimus himself." Martial makes a connection between poets and actors when he describes the poet blowing kisses in response to applause (1.3.7, 1.76.14), a custom associated with actors (see Citroni 1975 ad loc.)

48. Pseudolus himself reminds us that the trickery by which he has outwitted his master is ultimately no protection from the absolute power that his master holds over him. When Simo promises that he will get his revenge sometime, Pseudolus answers, "Why threaten? I've got a back" (quid minitare? habeo tergum, 1325).

49. For an excellent study of Plautine "metatheater," see Slater 1985, especially 12-18.