STRAIGHT TALK ABOUT CURVED HORNS AND GAY MARRIAGE: A NEW READING OF JUVENAL'S SECOND SATIRE

Abstract This article argues that one of our only pieces of evidence for Roman marriage between cinaedi, Juvenal's second satire, has been consistently misread and in fact describes a marriage between a cinaedus and a sex worker. It begins by providing the context for the passage in question and its traditional reading, and then demonstrates that the critical phrase siue hic recto cantauerat aere refers to financial, not erotic, exchanges. The article finally discusses the implications of this correction, which are far more substantial than one might expect for a contentious ablative.

Having explained what this article sets out to accomplish, however, I should also explain what it does not.Juvenal is an author of no small craft, and his dense web of allusion, intertext and invective has drawn a great deal of attention from scholars of imperial literature 2 and especially of satire. 3That is not my project; my interests are primarily historical. 4This is not to say that I will here take Juvenal as an eyewitness to the events he claims to describe, but rather that I assume he knew what his audience would find funny.For Juvenal's jokes to land, he would have had to describe events his audience found plausible or logical (even if the logic were that of a dream); his writings can tell us, if not about how people actually lived in Rome, about how they thought other people secretly lived, or how they would live if they could.Those are the questions I use Juvenal to answer.As such, I am not particularly concerned with Juvenal's imbrication within other literary traditions except inasmuch as they explain otherwise puzzling representational or lexical choices, and-for the purposes of this article, at least-I do not argue that Juvenal did or did not mean what he said. 5I refer to the narrator of the satires as 'Juvenal'; that does not indicate a commitment to reading the author and speaker as coterminous, so much as an interest in different questions (and a desire to be easily understood).

I
When Juvenal describes a wedding reception for two men at 2.117-42, he references an established literary-if not social-practice.While we have no evidence of Romans using marital terminology to refer to their own same-sex partners, a number of imperial writers talk about other people's homosexual marriages, or pointedly cast those people's 2 I think in particular of S.H. Braund, Beyond Anger: A Study of Juvenal's Third Book of Satires (Cambridge, 1988); T. Geue, Juvenal and the Poetics of Anonymity (Cambridge, 2017); C. Keane, Figuring Genre in Roman Satire (New York and Oxford, 2006); Juvenal and the Satiric Emotions (New York and Oxford, 2015); D.H.J. Larmour, The Arena of Satire: Juvenal's Search for Rome (Norman, OK, 2016); C. Nappa, Making Men Ridiculous: Juvenal and the Anxieties of the Individual (Ann Arbor, 2018); J. Uden, The Invisible Satirist: Juvenal and Second-Century Rome (New York and Oxford, 2015). 3Especially K. Freudenburg, Satires of Rome: Threatening Poses from Lucilius to Juvenal (Cambridge, 2001), 229-97; M. Plaza, The Function of Humour in Roman Verse Satire: Laughing and Lying (New York and Oxford, 2006). 4For a more historical analysis of Juvenal's Satires-albeit one about which I hold some methodological reservations-see J. Gérard, Juvenal et la réalité contemporaine (Lille, 1976); see also F. Bellandi, Etica diatribica e protesta sociale nelle satire di Giovenale (Bologna, 1980); A. Galimberti, 'Adriano e Giovenale', MedAnt 16 (2013), 87-99; L. Perelli, Protesta sociale e poetica nelle satire di Giovenale (Turin, 1972). 5An enormous and influential strain of scholarship takes the narrator of the Satires not as Juvenal but instead as a persona: see the essays collected at W.S. Anderson, Essays on Roman Satire (Princeton, 1982), 197-486, as well as M.M. Winkler, The Persona in Three Satires of Juvenal (New York, 1983) and L.C. Watson and P.A. Watson, Juvenal Satire 6 (Cambridge, 2014), 35-48.In particular, taking Juvenal as mocking his narrator's disgust for his subjects has allowed readers to imagine him as deceptively sympathetic to contemporary audiences; see, for example, M. Broder, 'The most obscene satires: a queer/camp approach to Juvenal 2, 6, and 9', in D. Dutsch and A. Suter (edd.),Ancient Obscenities: Their Nature and Use in the Ancient Greek and Roman Worlds (Ann Arbor, 2015), 283-309.For a more critical perspective on Juvenal and persona theory, see J.W. Iddeng, 'Juvenal, satire, and the persona theory: some critical remarks', SO 75 (2000), 107-29; Geue (n. 2), 7-11 and Uden (n. 2), 3-8 offer an overview of the debate around Juvenal's deployment of persona (as well as its ethical implications) and a more extensive bibliography.
erotic choices as those of a 'bride' and a 'groom'. 6Suetonius describes Nero making his eunuch Sporus into a bride and parading him through the streets of Rome (Ner.28 per sollemnia nuptiarum celeberrimo officio deductum ad se pro uxore habuit) before himself taking on the female role in his marriage with Doryphorus.'Just as he had married Sporus, so was he married to Doryphorus, and imitated the cries and forceful shouts of maidens being deflowered' (Ner.29 sicut ipsi Sporus, ita ipse denupsit, uoces quoque et heiulatus uim patientium uirginum imitantes). 7The Historia Augusta's uita Heliogabali-likely based on the work of the Severan historian Marius Maximus 8 -refers to courtiers boasting about their own homosexual marriages (Heliogab.11 maritos se habere iactarent), while Elagabalus' paramour Zoticus is described as so powerful that courtiers treated him as the emperor's husband (Heliogab.10 Zoticus sub eo tantum ualuit ut ab omnibus officiorum principibus sic haberetur quasi domini maritus). 9Lucian's Dialogues of the Courtesans provides one of very few depictions of marriage between women, with Megilla insisting on being known as Megillus and boasting of her marriage to Demonassa (Dial.meret.5.3 Μέγιλλος γὰρ ἐγὼ λέγομαι καὶ γεγάμηκα πρόπαλαι ταύτην τὴν Δημώνασσαν, καὶ ἔστιν ἐμὴ γυνή). 10Furthermore, Lucian's True Histories suggests at 1.21 that marriage between men is a common practice in the fantastical all-male society of the moon.Finally (for now), Martial twice refers to men who perform masculinity to the outside world while playing the bride in private (1.24 and 12.42): aspicis incomptis illum, Deciane, capillis, cuius et ipse times triste supercilium, qui loquitur Curios adsertoresque Camillos?nolito fronti credere: nupsit heri.Decianus, do you see that man with unkempt hair whose gloomy brow you fear, who speaks of the Curii and the Camilli, champions of liberty?Don't trust appearances; yesterday he was a bride. 11barbatus rigido nupsit Callistratus Afro, hac qua lege uiro nubere uirgo solet.praeluxere faces, uelarunt flammea uultus, 6 See, for example, R. Taylor, 'Two pathic subcultures in ancient Rome', JHSex 7 (1997), 319-71; Williams (n.1), 245-52.On the strict differentiation between these two roles in Roman discussions of same-sex weddings, see G. Gellérfi, 'Nubit amicus: same-sex weddings in imperial Rome', GLB 25 (2020), 89-100, at 95-8. 7See also Tac.Ann.15.37 (describing a similar relationship between Nero and a Pythagoras). 8See T. Barnes, The Sources of the Historia Augusta (Brussels, 1978); A. Birley, 'Marius Maximus: the consular biographer', ANRW 2.34.3 (1997), 2678-757. 9To be clear, our sources on the reign of Elagabalus-particularly textual ones-are unreliable at best; I offer this more as an example of discursive tendencies in Latin literature than in the service of any claim about Elagabalus or his court.A debate persists about the meaning of the rituals these authors describe.While few argue that same-sex marriages were formally recognized in imperial Rome-despite Martial's use of lex, Roman marriage law was too gendered to neatly apply to unions of two men or of two women12 -classicists such as Bruce Frier, Amy Richlin and Craig Williams (as well as legal historians such as William Eskridge) have argued that male/male weddings could have happened regardless, as expressions of affection and communal celebrations. 13his account of same-sex weddings is familiar to modern readers: perhaps too familiar.When we imagine a modern same-sex marriage, we imagine both parties as part of the same stigmatized group (specifically, that of non-heterosexuals or queer people). 14odern accounts of same-sex weddings as sites of resistance or community-building implicitly presume this endogamy, and accordingly that when a society refuses to recognize such unions it does so because of distaste for, or discrimination against, both parties.By contrast, our surviving accounts of Roman same-sex weddings typically describe only one partner as failing to meet social expectations, with the other (the masculine partner in a marriage between men, or the feminine in a marriage between women) coming in for far less abuse. 15Elagabalus' courtiers wear feminine hairstyles (caput reticulo componerent) as they discuss their maritos, about whom nothing else is said.Lucian's object of mockery, Megillus, masculinizes her name while referring to Demonassa as ἐμὴ γυνή.Finally, Martial uses nubo to clearly indicate role differentiation in the marriages he describes. 16In each case the receptive or penetrated male is portrayed as a cinaedus, complete with the abusive tropes that accompany such a designation (moral incontinence, hedonism, effeminacy),17 whereas his insertive or penetrating counterpart-the 'husband'-is portrayed as conventional in his desires, if occasionally mocked as a stud for hire18 or as a lovestruck youth. 19If these two spouses were as different from each other in their tastes, pleasures and social lives as men and women were thought to be, it is hard to see them as part of the same 'subculture' (to quote Amy Richlin) and thus to see the relationships these authors describe as specific to, or contained within, that subculture.20 II Enter Juvenal.Juvenal's second satire is, by far, our most extended account of a Roman marriage between men.It invokes many of the tropes seen in the marriages discussed above: Juvenal consistently describes one party with feminizing language (for example at line 120 noua nupta), while showing him paying a dowry to the other (117 quadringenta dedit Gracchus sestertia dotem), and feminizing his appearance.The poet even makes a crass joke about the effeminate partner's inability to bear children at lines 137-8.Juvenal innovates on the form, however, in some important ways.For one, Juvenal's marriage is an unusually public affair; it is communally celebrated by the cabal of hypocritical cinaedi who recur throughout the text. 21While Juvenal refers to the wedding as a secret from the majority of unsuspecting Romans, he claims that such ceremonies will soon become a feature of the city's social calendar: 'these things will happen publicly, people will even want them entered into records' (136 fient ista palam, cupient et in acta referri).No other author is nearly so explicit in claiming that cinaedi fraternize and imagine themselves as a cohesive subculture.While it may be tempting to say that this portrayal simply reflects the broader obsessions of the second satire, it is nevertheless remarkable that Juvenal felt his charge would stick.The second satire complicates our understanding of cinaedi as social fact, moving from the language of pathologized individual desire to something more organized: from antisocial to contrasocial. 22n addition, the second satire has been taken as our best surviving evidence for cinaedic endogamy: in other words, for one cinaedus marrying another. 23Cinaedic endogamy would transform our understanding both of the lived experiences of the Roman cinaedus (since cinaedi would be able to find romantic partners within their own ranks, rather than seducing or hiring outsiders) and of the basic criteria that marked a man as cinaedus within Roman imaginations.After all, penetrative sex requires a subject as well as an object; if Juvenal thought that two men who enjoyed being penetrated would marry each other-and presumably engage in marital relationsthen there was vastly more space for what we might call 'role egalitarianism' in Roman ideas of homosexual relationships than previously assumed. 24his reading rests on a brief and nasty joke at line 118.The full passage (2.117-21) reads as follows (with the critical clause left untranslated): quadringenta dedit Gracchus sestertia dotem cornicini, siue hic recto cantauerat aere; signatae tabulae, dictum 'feliciter', ingens cena sedet, gremio iacuit noua nupta mariti.o proceres, censore opus est an haruspice nobis?Gracchus gave 400,000 sesterces as a dowry to a horn-player, or perhaps he recto cantauerat aere: The tablets have been signed, congratulations wished, The huge feast is laid, and the new bride sits in the husband's lap.O Chiefs, have we need of a censor or an oracle?That there is a pun is clear enough.The most apparent meaning of line 118 is that Gracchus' husband may not have played the curved cornu but instead the straight, trumpet-like tuba or aes rectum. 25However, this observation is too banal, and too disconnected from the poem's broader invective project, to serve as anything other than the staging ground for a double entendre.As for that double entendre, Braund translates the line 'or it may have been on a straight horn that this man played', 26 and Williams and Nappa (among others) follow her in taking Juvenal to suggest that the cornicen is himself a fellator. 27The syntactical claims required to justify this reading are clear enough to a contemporary audience.The verb canto primarily denotes musical performance and specifically oral performance; for a musician like the cornicen such performance might entail putting an instrument in one's mouth, and Braund argues that Juvenal uses recto aere to mean 'something straight, hard, and pointy'-namely, an erection.I propose, instead, that Juvenal refers to the cornicen performing or feigning romantic interest in exchange for aes rectum: 'or perhaps he played the part [of the husband] for the right price'.
Scholia are of no use here, 28 but my proposed reading-with its accompanying implications for how we understand cinaedi to have lived and loved in imperial Rome-is better supported by the text on both philological and literary grounds.First, cantauerat.While the verb canto is occasionally used for instrumental performance, this usage is not at all common and appears nowhere else in Juvenal; it is by far the greatest semantic stretch in the pun of line 118. 29Instead, canto typically denotes oral performance or singing. 30Juvenal follows this pattern; if we examine other instances where Juvenal uses canto, we can see that he de-emphasizes instruments-the critical component for analogizing a cornicen cantans to a fellator-almost completely, and instead uses the term to describe public speech and song. 31In fact, Juvenal frequently juxtaposes performances described with forms of canto with instrumental performance, suggesting that he views them as distinct musical modes. 32recto cantauerat aere is unusual, but it seems more likely that Juvenal would use this idiosyncratic locution to analogize tuba-playing to something further within the semantic range of cantare (such as performing or singing) than to something equally outside of it (such as fellatio).
That semantic range is particularly restrictive in Juvenal.More than other authors, Juvenal frequently uses canto to describe false or misleading public statements.My first example of Juvenal's use of canto, in the fourth satire, is perhaps the most obvious: 'Begin, Calliope.You may also sit down; it is no time for singing, the truth is being discussed' (4.34-5 incipe, Calliope.licet et considere: non est | cantandum, res uera agitur).Juvenal here tells Calliope (Muse of the oral genre of epic) that the time is not right for performing (cantandum), because, instead, the circumstance calls for true speech (res uera agitur).In the eighth satire Juvenal argues that Orestes' matricide was less obscene than that of Nero, because 'he did not pollute himself with [slitting] Electra's throat or with the blood of his Spartan wife, he poisoned none of his relatives, he never performed onstage as Orestes' (218-20 in scaena numquam cantauit Orestes). 33This is not the only time when Juvenal will use cantare to refer to stage acting; at 6.73-5 he claims that 'some women would even stop Chrysogonus from performing' (sunt quae | Chrysogonum cantare uetent) of whom his example, 28 Ancient scholiasts do not comment on the passage at all, and the fifteenth-century Valla only suggests that the cornicen is expertus, ut ait Probus, naturam eum quasi grande habere cornu.See P. Wessner, Scholia in Iuvenalem vetustiora (Leipzig, 1967), 26. 29The TLL identifies only nine other such uses in all of extant Latin: TLL 3.289.12-21.Interestingly, Martial twice uses canto with reference to musical instruments, but each time the cantans is a poet and not an instrumental performer (Martial himself at 8.3.21 and Lucan at 10.64.4), and thus likely performs with musical accompaniment. 30 We believe that Athos was once crossed by sea and whatever else lying Greece dares in its history, that it was covered with those same ships and that a solid sea was placed under wheels; we believe that the deep rivers failed and the seas were drained at the meals of the Mede, and whatever else wet-sleeved Sostratus sings about.
Juvenal here explicitly concerns himself with truth and lies.He claims that the story of Mt Athos having been crossed by sea is believed (creditur) along with all of the other fictions told by lying Greece (Graecia mendax), so when he next says that we believe (credimus) the other tall tales of which Sostratus sings, we understand the content of Sostratus' songs to be no less fictional.To be fair, this usage of canto is idiosyncratic to Juvenal; by way of comparison, while Martial uses canto twenty-four times in his extant corpus he typically deploys it to emphasize the oral nature of the performance described and never in contexts that suggest falsity. 34But it should not surprise us that Juvenal in particular might blur the lines between Roman public life and more traditional performance contexts; after all, the first satire explicitly parallels the mythical subjects of most poetry to the quotidian subjects of Juvenal's own (1.85-6).Given Juvenal's particular usage of canto, taking cantauerat to indicate fellatio would be very strange within Juvenal's corpus.Simply put, blowing on a horn is not the kind of performance that Juvenal envisages for canto anywhere else.
Another reading is available, however, which accords with Juvenal's focus on false public statements.Line 118 describes a wedding reception, almost certainly following some kind of ritual act. 35Understanding the cornicen as feigning affectio maritalis in exchange for money requires him to have at some point falsely proclaimed that affectio, and that action accords far more closely with Juvenal's other uses of canto.
My proposed reading also explains the tense of cantauerat.Taking cantauerat to refer to fellatio would imply not that Gracchus' husband performed fellatio in the 34 For example contrasting it with other types of speaking at 1.89.4-5(quereris and ploras); with dancing at 2.7.5 (saltas); and most frequently with the silent enjoyment of poetry at 5.16.3 (legis et tota cantas mea carmina Roma), 6.60.1 (laudat, amat, cantat nostros mea Roma libellos), 7.51.7 (referring to memorization: sic tenet absentes nostros cantatque libellos) and 8.61.3 (orbe cantor et legor toto).Martial also uses canto at 1.76.3, 1.94.1-2,3.93.9,4.55.5, 4.61.9, 7.14.5, 7.17.10,7.29.7,7.88.5, 8.3.21,8.61.3, 9.49.1, 10.64.4,11.3.5, 11.20.8,12.40.2,12.53.4,14.183.1 and 14.216.1.On similarities in language between Juvenal and Martial, see R.E.Colton, 'Juvenal's second satire and Martial', CJ 61 (1965), 68-71. 35While we do not know the precise contours of the Roman wedding, a number of sources (see, for example, Cic.Mur.27; Plut.Quaest.Rom.30; Val.Max.10.7) refer to a spoken component.While the particular fragment of speech there attested is spoken by the bride, it seems inconceivable that the groom played no part in his own wedding.See K.K. Hersch, The Roman Wedding: Ritual and Meaning in Antiquity (Cambridge, 2010), 187-90.context of his marriage to Gracchus but only, and specifically, that he had done so at some point before.This would be an odd claim to make about a man whose past is discussed nowhere else in the second satire; it says nothing about the actual marriage itself, but instead about the sexual history of a character who is otherwise not even given a name and who is otherwise described only with reference to his present sexual relationship. 36However, any public profession of matrimonial intent would be completed by the time of the scene Juvenal describes; the cornicen had, at that point, 'sung'.Now, recto aere.Braund takes aes rectum as referring to the erect, 'pointy' penis upon which the cornicen had, perhaps, performed. 37While my reservations here are more subjective than they are with cantauerat, taking recto aere as phallic requires a very laboured pun.cornu frequently serves as a metonym for the penis, from the 'horny bow' of Penelope's suitors in Ovid's Amores (1.8.48 corneus arcus) to Oenothea's promise in Petronius' Satyrica to give Giton 'something hard as a horn' (134 illud tam rigidum quam cornum). 38Juvenal himself explicitly references this connection only a few lines above; at 2.90 a group of cinaedi expel women from their mock Bona Dea festival by announcing that 'no flautress groans on a horn here' (nullo gemit hic tibicina cornu). 39By contrast, aes and aes rectum are not attested as metaphors for the penis.For such a pun to work, Juvenal's reader would have to travel from the usual instrument of the cornicen (which Romans imagined as phallic) to its opposite (the aes rectum or the tuba), then analogize it back to the cornu and thus to an erection.In other words, the punchline to the joke-line 118, which Juvenal offers as a possibility that might reveal some hidden truth about the reality of the line before-is less phallic than its setup.While Juvenal clearly intends his reader to think of a musical instrument at 2.118, there is no reason why an aes rectum would be seen as more phallic than the horns cornicines usually blew.
Furthermore, even if recto has fairly obvious sexual connotations, specifying that the new instrument of the cornicen is made of bronze, in the absence of any established metaphoric vocabulary about the metal, simply confuses matters. 40By contrast, if aes refers to the money that Gracchus gave to his new husband-to the dowry of quadrigenta sestertia Juvenal references at 2.117-as the price of the man's cantatio, the joke makes sense.recto aere remains as a double entendre between the traditional instrument and another more scandalous reading, but that reading refers to money and not sex; like Martial's Afer, Gracchus' new husband has accepted money in exchange for (or in contemplation of) sexual services and Juvenal specifies that the price is fair. 4136 On the bodily exposure of unnamed abject figures such as the cornicen, see Geue (n. 2), 74. 37Braund (n.26), 154; on pointiness in phallic metaphor, see J.N. Adams, The Latin Sexual Vocabulary (London, 1982), 14-25. 38See Adams (n.37), 21-2. 39Notably, Juvenal describes one of these men as wearing a golden hairnet (reticulumque comis auratum ingentibus implet), like both the 'hair-netted adulterer' (reticulatus adulter) of Juv.6(Ox.).22 and the courtiers of Elagabalus whom the Historia Augusta would describe centuries later.While the latter text is not explicit in its reference to Juvenal, commentators have noticed the intensely literary and referential character of the uita Heliogabali; see M. Bittarello, 'Otho, Elagabalus and the judgment of Paris: the literary construction of the unmanly emperor', DHA 37 (2011), 93-113; G. Mader, 'History as carnival, or method and madness in the vita Heliogabali', ClAnt 24 (2005), 131-72; Sommer (n.9). 40This is especially true if we take Gracchus-the 'bride'-as the object of the affections of the cornicen.Surviving invective against cinaedi suggests bodily debilities that would render aere spectacularly inapt, at least in the sense a sexual reading of 2.118 would seem to imply.Cf.Mart.6.26. 41This sense of recto as 'correct or fitting' may also reflect that Gracchus' 'dowry' would be exactly enough to lift the cornicen into the class of equites: Braund (n.26), 154; Courtney (n.21), 142-3.This reading has several philological parallels.aes as a metonym for money is extremely well attested, 42 and rectum often denotes fairness, accuracy or precision (as at, for example, Cic.Verr.2.3.168,Livy 45.37.5 and Tac.Hist.1.14). 43While ablatives of price usually accompany verbs of buying and selling, they not uncommonly qualify a good or service (as at Cic. De or.1.126 or Livy 27.3.1). 44y final observation is a broader one; construing the cornicen as a fellator goes against everything we know about cinaedic sexuality.In all of the examples discussed above, male/male 'marital' relationships feature strictly differentiated sexual roles mapping onto (or often dictating) broader differences in the partners' gendered presentation.The cinaedus plays the bridal role in the couple, often using language that is lexically available to only one of the two men.Consider how Martial exploits the gendered connotations of nubo to clearly differentiate the two parties to the marriage.Even Nero is an exception that proves the rule: when Suetonius claims that Nero married Doryphorus as Sporus had been married to him, the biographer makes clear that both relationships featured strict role differentiation, even as Nero played opposite roles in each.Role-egalitarian homosexual relationships are simply not attested elsewhere, which militates against interpreting the dubious evidence of line 118 to constitute our sole surviving account. 45o be fair, were I to rest my argument here I could be fairly accused of begging the question.If we take a practice as the norm within a society on the grounds that it is the only practice attested, but also treat sceptically accounts of other ways of living because they describe acts that make no social sense, we risk creating the unanimity from which we claim to reason. 46That risk is heightened here, because Juvenal's account in the second satire really is unique.The second satire describes cinaedic sociality in far more detail than any other extant source, and Juvenal's account of in-group solidarity-concordia inter molles-bears no obvious parallel.Furthermore, Juvenal is perhaps the Latin author most likely to pull a sex joke out of thin air.But jokes require context, and it seems inconceivable that a reader without access to a long, modern tradition of role-egalitarian homosexuality would see line 118 as a reference to mutual fellatio without clearer guideposts than Juvenal here provides.
Instead, that reader would presumably contextualize the pun within Juvenal's broader description of the relationship between Gracchus and his cornicen.Juvenal describes that relationship in ways that militate against reading the cornicen as a 42 TLL 1.1074-7. 43While this use of rectum does not elsewhere accompany the metonymic aes, the phrase aes rectum is sufficiently well attested as a synonym for the tuba that an author would only combine the two terms if-as I argue happens here-they were intentionally conflating the musical and the financial readings of the phrase. 44See also H. Pinkster, The Oxford Latin Syntax (Oxford, 2015), 1.1200-1. 45There are two possible exceptions to this rule, both cryptic.Suetonius describes Caligula as 'sparing neither his own sexual honour nor that of anyone else' ( pudicitiae neque suae neque alienae pepercit) and as having reciprocally pleased a M. Lepidus, a pantomime-actor named Mnester, and a number of Roman hostages (M.Lepidum, Mnesterem pantomimum, quosdam obsides dilexisse fertur commercio mutui stupri), while Seneca refers to wellborn young men who 'wickedly satisfy their own lusts and those of others by exchange' (suam alienamque libidinem exercent mutuo inpudici): Suet.Calig.36.1;Sen. Ep. 99.13.The sheer numbers of partners listed in each, however, suggest that this boundary violation occurred in shorter encounters rather than in role-egalitarian relationships.See Richlin (n.13), 539-40. 46T. Ilan, 'Premarital cohabitation in ancient Judea: the evidence of the Babatha archive and the Mishnah (Ketubbot 1.4)', HThR 86 (1993), 247-64, at 251-62 presents a useful introduction to this problem, albeit in the context of Jewish practice.
cinaedus and point instead towards more meretricious interpretations.For example, when Juvenal describes Gracchus at 2.120 as 'newly married and reclining in her husband's lap' (gremio iacuit noua nupta mariti), Gracchus is explicitly feminized while his husband, just as explicitly, is not.Similarly, Juvenal differentiates at 2.137-8 between the terrified brides (nubentibus) and the husbands whom they wish, but are unable, to keep by becoming pregnant (nequeant parere et partu retinere maritos).Both jokes rely on the two men in these marriages being categorically different.Juvenal attacks the mariti in these marriages in all sorts of terms-in particular, as excessively socially mobile and as exchanging sex for money, both allegations visible throughout Juvenal's work-but never feminizes them. 47Furthermore, when Juvenal discusses the infertility of these marriages, he seems to suggest that the fault lies entirely in the unsuitability of the body of the cinaedus.This assignation of blame only makes sense if the sexual behaviour of the maritus is more 'correct' than his partner's-that is to say, if he plays the insertive and inseminative role.In other words, taking the marriage of Gracchus as a union of cinaedi would require not only reading it as different in kind from every other male/male marriage described in extant literature but also reading Juvenal's other claims about the marriages of Gracchus and his friends out of the text.
So, to sum up.Juvenal describes Gracchus' husband, an unnamed cornicen, in harsh and somewhat laconic terms.I take Juvenal to accuse the cornicen of feigning marital affection in exchange for money, in contrast to other modern readers who-perhaps informed by modern homoerotic and homoromantic relationships-have understood the cornicen as a fellator. 48I have argued that this latter claim is unsupported by the text.It requires Juvenal to use canto in a highly idiosyncratic fashion, accompanied by a sexual metonym otherwise unattested in Latin literature, in order to depict a sort of relationship of which no other example survives and which differs radically from that ascribed to Gracchus and his groom in the rest of the second satire.Gracchus' cornicen is not a cinaedus but a mercenary; prior to the banquet Juvenal depicts, the cornicen has 'performed' (that is, professed his affection for, and marital intentions towards, the wealthy Gracchus) in exchange for a good amount, or an appropriate amount, of money.

III
So what are the stakes of this contested pun?I here suggest some implications of a correct reading of Juv.2.118.The first is historical; the second satire has been taken as our most important evidence for cinaedic endogamy, and it should not be so taken.With Juvenal taken off the table, it seems almost certain that the cinaeduswhether taken as a discursive category Roman elites used to map the boundaries of normatively acceptable sexual desire, 49 or as an actual social group organized around that categorization and finding meaning within it-sought out non-cinaedic men as 47 See Braund (n. 2), 164-8; T.P. Malnati, 'Juvenal and Martial on social mobility', CJ 83 (1987), 133-41, at 133-4; Nappa (n. 2), 186-8. 48Craig Williams raised a similar point in his review of Braund's commentary, specifically noting that Braund's use of modern terminology such as 'homosexual' in translating cinaedus subtly affected her reading of the second satire: C. Williams, 'Review of Juvenal Satires Book 1', BMCRev (1997): https://bmcr.brynmawr.edu/1997/1997.07.09. 49Thus Gleason (n.17), 392: 'We infer the existence of a cultural norm of masculinity from accounts of deviations from it.'sexual or romantic partners.As a result, understanding Juv.2.118 as a financial reference complicates the connections that scholars such as Amy Richlin have drawn between the cinaedus and the contemporary 'passive homosexual'. 50Endogamy is one of the defining features of contemporary gay life, particularly regarding the sorts of rituals that Juvenal describes and that scholars have seen as resembling modern same-sex weddings.More specifically, marriage rituals between two members of a disfavoured group offer opportunities not only for celebration or pleasure but also for reinforcing group identity and strengthening communal bonds. 51By contrast, the marriage described in Juvenal is one between members of two discrete groups; while both kinds of wedding invoke the forms (to a greater or lesser degree) of more widely practiced marriage ceremonies, they do not do anything like the same work.
Even if we take a more abstracted view of the Rome that Juvenal depicts-as a monstrous reflection in the mind of a man too furious to be taken literally 52 or alternately as, per Susanna Braund, a 'channel through which views of Roman morality are constituted, rehearsed, reinforced, and transmitted' 53 -Gracchus emerges as a very different character if we understand his affections to aim primarily outside of his own subaltern group.Gracchus becomes a cinaedus who actively seeks 54 to corrupt other sexual partners, and this (for example) accords far better with the predatory cinaedi of the sixth satire or the eunuch groom of the first. 55Juvenal depicts cinaedi as a menace, not because they made their sexual partners into cinaedi themselves but because the pleasures they offered were categorically different from those that other Roman men could hope to provide, and thus destabilized the patriarchal erotics that the satirist claims to prefer. 56imilarly, while I care more about Juvenal as a historical source than as a literary figure in his own right, understanding the cornicen as a sex worker opens up broader avenues for understanding the author's invective project.Gracchus' wedding is particularly important for Juvenal's corpus because Gracchus himself is one of its few recurring characters; Juvenal describes him fighting as a gladiator at 2.143-8, and returns to the theme at 8.199-210.Effeminate gladiators may seem striking to a contemporary reader, but in the latter poem Juvenal calls attention not to Gracchus' marriage but to his rank; he analogizes Gracchus pugnans to patricians acting onstage (8.190-2) and to the 'musician/emperor' (citharoedus princeps) of 8.198. 57Gracchus' 'neglegit atque alium bipedem sibi quaerit asellum.' 'But, as you feign ignorance of and dismiss my other services, what price do you reckon for the fact that, had I not been your attached and devoted client, your wife would remain a virgin?' … 'Do you reckon it worth nothing, ungrateful wretch, nothing at all that your son or daughter was conceived by my effort?' … 'You have a parent's rights, and on my account you are inscribed as an heir and take the whole legacy, plus escheats besides.Many extra goods will be joined to those estates if I fill out the household and make it three.''You have good reason to be upset, Naevolus, how does he reply?' 'He ignores me and seeks another donkey on two legs.'Naevolus is one of the very few characters whom Juvenal allows to speak at length, and most scholarship on the ninth satire has viewed his abject position (as a poor cliens forced into sex work) as analogous to Juvenal's own perceived social marginalization. 62f we take the second satire's marriage to be based on mutual affection, then Naevolus has no parallel within Juvenal's corpus.Both Naevolus and the cornicen, however, trade sexual penetration-the privilege of socially dominant Roman men-for money. 63The cornicen publicly marries Gracchus, while Naevolus privately satisfies his patron and impregnates his wife.
Both transactions are masked by Roman institutions with deep historical roots: marriage in the second satire, clientela in the ninth.In each poem the actual inversion that Juvenal critiques, the dependency of the insertive man on his nominally subordinated receptive partner, is hidden from public view.That perverse relationship, however, hides behind different fictions in the two poems.Gracchus presents himself as the weaker partner in his marriage, sitting in his husband's lap and frantically seeking potions to stop him from leaving, but it is actually the cornicen who is unable to leave without financial ruin.By contrast, Naevolus' dependence on his patron is a publicly established social fact, while that patron's passivity and unwillingness to inseminate his wife remain hidden.Reading both passages in tandem reveals Juvenal's concern with broader social distortions; these two relationships, similar in their lived realities, masquerade as, and thus corrupt, two very different institutions.The range of social practices encompassed by older, more upright forms of social life boils down to the one; whether marriage or patronage, Juvenal seems to say, it's prostitution all the way down. 64his is, of course, a truly weird thing to think about imperial Rome, and we might rather Juvenal thought something more familiar.Juvenal's complex poetics often force his reader to resolve semantic ambiguities, and it can be tempting to do so in ways that produce a morality we recognize.This sort of domestication often invokes the Juvenalian persona 65 but can instead take the form it does here: of a satire cantata in the voice of someone like a modern homophobe attacking something like a modern homosexuality.I worry, though, about what we lose when we approach Juvenal in this way.The singular choleric worldview of the satires (whether Juvenal's own or a poetic project) makes them a remarkable artefact, and we should be cautious about sanding down their edges.
Furthermore and finally, accepting the radical strangeness of Juvenal's cinaedi can help us see the contingency of our own attitudes.Amy Richlin is right to note that the lives recorded in even hostile authors can help later readers better situate themselves within histories of human friendship and desire. 66But this tool does not belong only to readers with the right kinds of friendship or desire.Even if Juvenal does not, in fact, depict a role-egalitarian homosexual relationship of the kind with which most of us are familiar today, there are other ways in which queer people have lived.Juvenal's concordia inter molles implies a mutual friendship and support that does not require romantic affection, and that engages closely with other marginalized people (particularly the enslaved or freed, or those engaged in sex work).I make no claim here about whether a subculture like the one Juvenal describes ever existed in Rome, but if it did it would have clear historical parallels; for much of 'modern' history, queer societies consisted of gender-nonconforming people who formed close social bonds with each other, while looking outside of that circle to find sexual partners who were often financially compensated and were understood as far less unusual in their desires and self-presentations. 67o, then, earlier readings of the second satire have not simply read modern social practices 'into' the text, 68 but have read other social practices out.Katherine Franke and Michael Warner, among others, have written about contemporary society's desire to inscribe a narrow space for 'dignified' non-normative sexuality, and how this dignity-granting process necessarily consigns other ways of being to a greater abjection. 69When we find sexualities like our own in ancient texts we foreclose these possibilities, and join in a discourse that implicitly ranks some ways of living above others.To end on a prescriptive note: if we wish to treat Juvenal as a chronicle of Roman practices that dared not speak their name, we ought to take those practices as we find them.