by Greyson Gove
This project deals with the topic of male prostitution in ancient Rome, and the tangled narratives, biases, and scholarly practices which surround its study. Researchers in classics and other fields may approach the topic for a variety of reasons — queer and women’s studies scholars may examine prostitution as a lens for Roman sexuality or gender inequality, while the more normative and prevalent cisgender heterosexual males in the field may study the topic for (supposedly) purely economic or historical reasons. Whatever the impetus behind its study, prostitution, and specifically male prostitution, exists at a complex and interesting crux of scholarly and personal biases. The project of this paper is, among other things, to map and attempt to untangle these webs of discourse and ideas. In the long term, this research will tease apart three distinct lenses through which male prostitution is analyzed and interpreted: material culture, language, and “theory.” While this paper deals primarily with the first of these, all three lenses are important in capturing a full picture of the current state of the field.
First, I will provide some context on the nature of these three “lenses,” and my rationale behind dividing them as I have here. Unlike cultural anthropology and sociology, which (despite their own biases) have the luxury of dealing with extant cultures and conducting field research with living people, classics and archaeology are by nature limited in the information they have access to. As such, classicists have, in my view, three options for collecting data and making their arguments. The first is material culture. Most prevalent in the study of archaeology, this kind of work — studying material remains, artifacts, and ancient art and architecture — gives us access to tangible pieces of the past. The second lens, language, is particularly common when studying cultures like Greece and Rome which left behind an extensive written record. Through translating and analyzing poetry, laws, and other documents from the past, we may find some insight into how they functioned. The final lens, “theory,” is how I label any argument which uses the field itself as evidence. Arguing that a piece of art should be interpreted a certain way because other scholars have interpreted other works similarly, or that a certain modern identity did or did not exist in antiquity: these kinds of claims, in my model, would fall under theory.
Pointing out these lenses and their interactions with each other is by no means novel, but what I would like to emphasize here is that, despite the veneer of objectivity, all three of these lenses are subject to bias and misinterpretation. As I will outline throughout this paper, while a certain artifact or structure may physically exist, the ways material culture is used to enforce an argument or prove a point are very often tenuous at best, and outright misleading at worst. My project here is not to dismantle the field — or, not to dismantle it out of spite, or a kind of skepticism which refuses the value of this type of work. Rather, I would like to examine how, in such a fraught and complicated topic as male prostitution, the narratives around these lenses are not simple. They are convoluted, intertwined, and often betray or hint at a larger scholarly bias. Whether or not my research sheds any light on the actual workings of prostitution in ancient Rome, it is still important to capture the current discourse in all its complexities and implicit and explicit bias.
II. Background and “Theory”
While prostitution is a fairly well-studied topic within the field of classics, with scholars like Thomas A. McGinn publishing multiple books on the topic, male prostitution in particular is somewhat missing from the narrative. On the one hand, scholars take for granted that male prostitutes existed and were even common in ancient Rome — McGinn, in The Economy of Prostitution in the Roman World, dives into an explanation of why his focus is female prostitution without feeling the need to first assert that male prostitution existed. On the other, though this very “taking for granted” in some ways contributes to a lack of scholarly attention on the subject. While scholars tend to agree that male prostitution was prevalent in ancient Rome, few actively engage with the subject beyond a footnote. McGinn offers an explanation for his focus on female prostitution in The Economy of Prostitution in the Roman World: he writes that “the greater share of the evidence” on prostitution concerns women, and that “male prostitution is an important subject nonetheless and thus is deserving of separate treatment” (2). This verbal maneuver, affirming the existence and even importance of a topic before dismissing it as someone else’s problem, raises a question: where is this separate treatment on male prostitution in ancient Rome, and, further, why exactly must the study of classical sex work be separated explicitly by gender?
McGinn does offer some compelling evidence on these points over the course of the text. Some Roman emperors and other officials attempted to explicitly ban male prostitution (97), though this is more true of the Christian era of Rome than its predecessors, and the fact that Roman law at times separated sex work by gender is one argument for separating them in scholarship. Both these explanations, that the evidence is primarily on female sex work or that the topic of male prostitution should be treated as its own study, are all true to a point, but I argue that they ignore the complexities and the societal and theoretical biases around this particular topic. For better for for worse, the subject of male prostitution is inexorably linked to homosexuality. In Economics, Sexuality, and Male Sex Work, a book detailing the history of male sex work from the ancient to the modern world, scholar Trevon D. Logan also asserts that male prostitution clearly existed in cultures like ancient Rome, but, in the same breath, already ties it to homosexuality. In the very first chapter of the book, he writes: “Male sex work as an occupation is as old as its female counterpart . . . . male sex work has always carried the added stigma of homosexuality, causing male sex to be socially distinct from the more widely practiced female sex work” (19).
Many scholars of classical Greece and Rome, McGinn included, seem to dance around the topic of sexuality, referring to male sex work a distinct and different without explaining precisely why. And perhaps these scholars are right to avoid mentioning homosexuality — one of the prevailing notions of Roman sexuality is that it wasn’t about gender, that it was all about penetration, dominance and submission, the active and passive roles taking on more significance than gender. Even if these dominance-submission-obsessed models are completely accurate, though, they fail to represent the way many scholars seem to frame the discussion. Take McGinn’s claim that male sex work should be a distinct and separate topic: if the only thing that matters in Roman sexuality is dominance and penetration, why should this be the case? Male prostitutes often behaved passively, often were penetrated, even occupied lower social classes than their clients. If Rome itself apparently didn’t care about gender in sexual relations, why should we? This is why the quote from Logan, that “male sex work has always carried the added stigma of homosexuality,” rings true to me despite the fact that its historical accuracy is up for debate. Whether or not Roman men were “always” stigmatized for “homosexuality,” the topic of male sex work is. Tacit in McGinn’s focus on female sex work and insistence that male prostitution must be treated separately, carefully, and as its own distinct topic, I see the specter of homosexuality.
This is not to say that I am any more objective than Logan or McGinn. I am far from a neutral party on this topic: as a gay man with an interest in classics and the Greco-Roman world, I have a vested interest in finding representations of queerness in the real and mythological past, in actively queering the narratives around Rome. My positionality makes me quick to push back against the subtle erasure of male sex work and sexuality it McGinn, and to question and poke holes in the dominance-submission model of Roman sexuality; it makes me perhaps more lenient with scholars like Logan who use modern labels and notions of sexuality to describe the distant past. The point here is not that they are biased and I am not, but rather that the bias around this topic has been hidden in layers of jargon and theoretical discourse. The topic of male prostitution is clearly marked as “gay,” as queer, as non-normative because we view it that way in our own society. As I continue to unravel the discourse around this topic, I hope to expose these inconsistencies, and to map the biases and assumptions, conscious or not, which inform the way we discuss sex work and sexuality. Whether we like it or not, sexuality and sex work are linked. In trying to ignore the influence the idea of homosexuality has on our view of male prostitution in Rome, the narrative itself becomes dishonest and incomplete. As I examine the complex and tenuous ways material culture and other more “objective” pieces of evidence are invoked around this topic, these biases and my own positionality are vital to keep in mind.
III. Notes on “Language”
While this paper deals most explicitly with material culture, my overall project is to map the discourse around male prostitution in ancient Rome through all three lenses, and one of these is “language.” Like material culture, language is something that tangibly exists: ancient texts, epic poems, and legal documents from ancient Rome all have words which were written by ancient people. While the words themselves are immutable to a point, the way we interpret them ranges from tenuous to outright biased, especially when dealing with topics like male prostitution and homosexuality wherein scholars and translators may have a vested interest in a certain type of translation. Once again, this is not to say that translation is useless. Far from it — it is one of the most direct and important portals we have into the past. Rather, I would like to examine the way individual words with more literal meanings take on a life of their own within the discourse, and become animated and altered by the scholarly narratives around Roman sexuality.
Homosexuality in particular is a rather convoluted topic when it comes to translation and linguistic evidence. While the prevailing notion in some scholarship is that homosexuality did not exist in ancient Rome, there are nevertheless a constellation of words and labels which are taken to refer to various types of sexually non-normative men. In “Some Myths and Anomalies in the Study of Roman Sexuality,” for instance, James L. Butrica brings up the words cinaedus, concubinus, puer delicatus, and exoletus — all of which are taken to refer to various kinds of sexually “passive” men who sleep with other men. Even disregarding the accuracy of the various translations of these words, I would like to push back for a moment against the weight we put behind these words at all. Regardless of the historical and textual evidence we have for any given definition of a word, scholars sometimes invoke labels like cinaedus as if they refer to a legible and unchangeable caste of people.
Cinaedus, which Collins Latin Concise Dictionary defines as “sodomite; lewd dancer,” seems to me more a pejorative or descriptor than a concrete social label, but this is not how many scholars treat the word. Take Butrica’s self-described task of “[arguing against the common belief] that the cinaedus cannot be the same as the modern male homosexual because the cinaedus was thought capable of performing cunnilingus” (209) — even the language used here, “the” cinaedus being “capable” of various acts, almost seems to treat the label as a separate caste or species. The word does certainly appear in Latin texts, but compare it for a moment to present-day pejoratives like “nancy boy” or social categories like “twink” within the modern gay male community. These terms do have reliable meanings, and are social labels to some extent, but they describe appearance and perceived behavior, not a distinct caste of men. Speaking of what “the twink is capable of” in the bedroom sounds ludicrous. While these words are certainly not in a one-to-one relationship with those like cinaedus, archaeologists and linguists attempting to reconstruct what the modern gay community looked like with limited textual evidence might treat them as such. Existing translations and analyses of these words may indeed have merit, but constructing monolithic sexual categories — especially categories which are already designed to fit within the existing framework of Roman sexuality as all about dominance and submission — has the potential to cause harm, and to distort the research.
Until now I have examined the linguistic discourse around male homosexuality in ancient Rome from a more general or hypothetical standpoint. Now, I will frame my analysis around a specific word, and one directly related to male prostitution: exoletus. Commonly defined as “older male prostitute,” Butrica devotes considerable time to untangling the word’s etymology in “Some Myths and Anomalies in the Study of Roman Sexuality.” Butrica’s central claim about exoleti is that, rather than being strictly adult male prostitutes, they were “adult sexual partners of adult males” (223) who “have outgrown . . . in theory at least, their sexual attractiveness . . . [but] continue to be sexual partners of men” (225). Butrica bolsters this claim by tracing the word’s usage in existing Latin texts, and examining in each case whether the common understanding of exoletus as “male prostitute” makes sense. In his analysis, only one historical usage of the word undeniably refers to sex workers, and many others — such as the word’s use to describe Ganymede in Prudentius’s telling of the myth of Jupiter and Ganymede (228) — make no sense whatsoever when forced into that definition.
Butrica is not necessarily free from bias or scholarly agenda; “Some Myths and Anomalies in the Study of Roman Sexuality” was published in the Journal of Homosexuality, and he makes clear from the onset of his text that one of his goals is to push back against the notion that various Roman identities like the cinaedus cannot correspond to modern gay men. His positionality does not, however, discredit the careful textual work he does over the course of the paper, nor make the claims of the scholars he critiques any more well-founded. My point is not that Butrica is an unbiased and therefor superior scholar, but that other biases are often presented as unbiased within the discourse. Butrica’s work to trouble and undermine our current understanding of words like exoletus is astute and valuable in particular because the scholarly bias which assumes that Roman sexuality was never based on gender is often taken as objectivity.
Once again, let us turn to McGinn; in The Economy of Prostitution in the Roman World, he calls using words like “homosexual” to describe Roman sexuality a “radical constructionist” view, and generally avoids using the term throughout his text (230). One the face of it this is a perfectly reasonable argument to make, especially since it follows from the general model of Roman sexuality as entirely centered on the active-passive dichotomy. This is troubled, however, when we examine the way he talks about opposite-sex relations, referring to them as “heterosexual couples making love” (164) and mentioning “all-heterosexual graffiti” (229). Making a conscious effort to refer to graffiti describing gay sex acts as depicting “same-sex relations” while being perfectly comfortable using the term “heterosexual” and even euphemistically saying straight couples “made love,” to me, betrays a bias. The bias here is simply a more normalized and therefor subtle one: heterosexual is taken as the norm. Rather than avoiding any terms which reference gender preference, McGinn specifically avoids using the term homosexual. This is a pattern in much of the narrative around ancient sexuality — even while insisting that gender and gender preference are not and should not be part of the conversation, the labels of “homosexual” and “gay” are still treated as abnormal exceptions to the heterosexual rule, and straight gender preference is not given the same careful deconstructionist treatment that same-sex historical couples and relations are.
IV. Material Culture and the House of Jupiter and Ganymede
As detailed above, Thomas McGinn falls into some linguistic biases when he approaches translation and terminology like “homosexual” in his study of Rome, but his work is more revealing here as an examination of how material culture is used to provide evidence on male prostitution and sexuality in ancient Rome. Material culture and archaeology in general can be just as deceptively objective as language. On the one hand, the material world does tangibly exist: artifacts found in Pompeii and Rome have a fixed physical existence which we can analyze, and which give us a somewhat concrete portal into the past. On the other hand, though, the interpretation of these artifacts is just as subject to bias as translation and other theoretical work. Especially around subjects such as non-normative sexual behavior and sex work, surviving physical evidence is at times limited, and scholars can and do extrapolate a great deal from a fairly minimal amount of actual material. One interesting example of this concerning male prostitution is McGinn’s analysis of an archaeological site in Pompeii called the House of Jupiter and Ganymede.
McGinn has access to limited material evidence about the House: the archaeological site, like many in Pompeii, has survived, but the main evidence that it may have been a brothel is the presence of sexually explicit graffiti on the outside. Interestingly, and central to the argument McGinn pushes back against, all of this graffiti describes or depicts same-sex relations (229). McGinn’s work in this portion of The Economy of Prostitution in the Roman World is generally to identify and map the various brothels in Pompeii (with a near-exclusive focus on female prostitution), and, as such, there are two claims he wants to make about the House of Jupiter and Ganymede: first that it was indeed a brothel, and second that it was most likely a “co-ed” brothel, housing both male and female sex workers despite the skewed nature of the graffiti. In supporting these claims, he calls on both his own interpretation of the graffiti and archaeological work by John R. Clarke, a colleague of his.
The first of these two supporting arguments, i.e. his interpretation of the graffiti, is more hypothetical than anything else. Over the course of the passage, McGinn argues that “the fact that the graffiti refer exclusively to same-sex relations does not inevitably mean women did not sell sex there as well, any more than the presence of all-heterosexual graffiti at a brothel means that male prostitutes did not work there” (229); in other words, if all “heterosexual” graffiti does not rule out the possibility of male sex workers, “same-sex” graffiti does not rule out that of female ones. This claim is complicated by his later assertion that “most of the Pompeiian brothels had only female prostitutes . . . . [and] all-female brothels were the norm elsewhere as well” (229). McGinn’s claim about the House of Jupiter and Ganymede is built off of a conditional — if “heterosexual” graffiti does not rule out the possibility of male sex workers — and his later claims make this conditional uncertain at the very least. Once again, like in his asymmetrical approach to the terms “homosexual” and “heterosexual,” McGinn seems to see no theoretical problem in the prevalence of strictly “heterosexual” brothels, and reserves his skepticism for establishments which may have catered to same-sex tastes.
More tenuous than this, however, is his invocation of John R. Clarke. McGinn describes Clarke as having “withdrawn his identification of [the House] as a hotel for homosexuals, evidently out of concerns grounded in orthodox social constructionism” (229), citing a passage from Clarke’s “Looking at Lovemaking: Constructions of Sexuality in Roman Art.” Examining Clarke’s original text, however, there are two notable discrepancies between his account and McGinn’s. First, Clarke did not describe the establishment as “a hotel for homosexuals,” but rather “a hotel for gay men” (88). While these two wordings are fairly similar in meaning, it is nonetheless interesting that McGinn opted to paraphrase Clarke rather than quote him directly or otherwise use the word “gay.” More pressing, however, is the second difference between the two texts: while McGinn describes Clarke as withdrawing a claim about the House of Jupiter and Ganymede, Clarke does not reference the establishment at any point during his entire article. Rather, Clarke’s argument centers around how to interpret a scene from an archaeological find: namely, a satirical sex scene engraved on an artifact called the Warren cup.
The Warren cup depicts a total of five figures, four involved in sex acts and the fifth acting as (according to Clarke) either a transitional figure, an attendant, or a voyeur. On side A of the cup, two adult men of seemingly equal status make love (in Clarke’s wording); side B depicts the same between a man and a younger boy, and a second boy with a distinctly different appearance is a medial figure who seems to be observing the couple on side A. There are two problems for analysis here: first the depiction of sex between adult men, and second the role of the onlooker boy between the two sides. Clarke puts forward two possible claims as to what exactly these scenes depict. First, he poses that the cup may depict a brothel, with the medial boy figure as a kind of attendant; second he briefly suggests the idea that the scene may depict a kind of “gay hotel.” Along the lines of McGinn, he quickly rejects this second idea, calling it “naively anachronistic” (88), but he notably does not land on a singular interpretation of the cup, instead offering various other scholars’ analyses for the consideration of the reader.
It is possible that Clarke has made comments, either in personal correspondences with McGinn as a colleague or in un-cited works, about the nature of the House of Jupiter and Ganymede, but, whether or not this is the case, McGinn’s entire use of his argument hinges on a citation which has nearly nothing to do with the actual archaeological site at hand. Whether or not the Warren cup depicts a brothel — and this idea is tenuous at best — it is a work of art and likely of satire, and not, as McGinn intentionally or unintentionally implies, a work commenting on the nature of the House. While this particular example is a kind of rabbit-hole into the minutia of archaeological analysis, it is in some ways representative of the role material culture has in any work within the field. There are physical, objective realities involved — the House has well-preserved graffiti, and the Warren cup depicts a clearly visible scene — but this material evidence is interpreted, extrapolated, and even warped to prove a theoretical point. Whether or not McGinn in particular is influenced by scholarly or personal bias, the potential for that bias is there.
Over the course of this paper, I have outlined a small subsection of the current academic discourse and scholarly narrative around male prostitution in ancient Rome, and specifically how it relates to and tangles with biases around Roman sexuality. The three lenses I outline — the broader lens of “theory” and the two more specific lenses of language and material culture — are far from the only ways to interface with this topic, but they do cover the majority of the evidence classicists and archaeologists have at their disposal. “Theory” as such is by nature a scholarly framework, and a subjective if well-supported lens for analysis, but my hope is that this paper has demonstrated that the other two lenses are no more objective than theory. As this project continues, I hope to map more of the discourse around male sex work in ancient Rome, and more of the convoluted interactions of theory, tangible evidence, and scholarly and personal bias in the field. Ultimately, many of the questions we have about Roman sexuality and sex work will always go unanswered. Like the illustrations on the Warren cup, there will always be multiple interpretations for anything we find, no matter how tangible, and no matter how concrete the evidence itself is. By examining our own positionality, decentralizing the assumptions framed an objective and unbiased, and paying close attention to the discourse itself, however, we may at least attain a more nuanced perspective.
Butrica, James L. “Some Myths and Anomalies in the Study of Roman Sexuality.” Journal of Homosexuality, no. 49, 2005, pp. 209-69.
Clarke, John R. “Looking at Lovemaking: Constructions of Sexuality in Roman Art, 100 B.c.-A.d. 250.” University of California Press, 1998.
Collins Latin Concise Dictionary. First US edition, HarperCollins Publishers, 2006.
Logan, Trevon D. Economics, Sexuality, and Male Sex Work. Cambridge University Press, 2017.
McGinn, Thomas A. The Economy of Prostitution in the Roman World: A Study of Social History and the Brothel. University of Michigan Press, 2004.