Skip to content Skip to navigation

LHMP #218 Williams 2010 Roman Homosexuality

Full citation: 

Williams, Craig A. 2010. Roman Homosexuality. Oxford University Press, Oxford. ISBN 978-0-19-538874-9

Williams ironically acknowledges that part of his entire program is to demonstrate that “Roman homosexuality” is an oxymoron, but that this can only be explained by taking an in-depth look at the topics and evidence that superficially appear to define that very topic. The rule of thumb holds true that any academic study written by a man that has the word “homosexuality” in the title will have only minimal passing reference to female homosexuality, but in this case that’s an inevitable result of the nature, focus, and authorship of the available textual sources. But in this case, more to the point, it’s because the “Roman sexual system” itself assumes the primacy of the dominant, penetrative man and defines all other persons and actions in relation to that concept. The very notion of sex with no man or male-analog present is nonsensical within the normative structures of classical Roman sexuality. But for that very reason, a consideration of the place of women’s relations with that system (as with a consideration of men who don’t fit the dominant paradigm) helps find the cracks and inconsistencies of the system.

* * *


This book was originally published in 1999 and has been reivised and updated in this 2010 edition. It takes a similarly broad-based approach to that Dover (1978) did for Greek homosexuality. Williams notes that female same-sex desire is “not a central theme in [this] book” which is something of an understatement. But an understanding of how sexual relations between men in classical Rome fail to align with modern concepts of homosexuality also sheds light on ways in which relations between women might have been viewed.

This study interrogates the accepted premise of Roman relations between men, i.e., that the strict alignment of active and passive roles with status differences of the participants was considered a moral issue. That “passive” partners were universally ridiculed or despised. One key question is how sexual status hierarchies between men were distinguished from “heterosexual” relations (just as much a misnomer as homosexual in this context) in which the premise was that women are universally lower in status than men. Williams points out that textual data indicates that Roman men were not encouraged to evaluate or judge sex acts based on the genders involved, except in the case of acts between women which stood outside the expected paradigms. He challenges whether the terms “homosexual” and “heterosexual” have any historical meaning when applied to classical Rome.

In this analysis, it’s important to understand the contextual meaning of words like stuprum (debauchery, a shameful act), pudicitia (chastity, modesty), cinaedus (man who takes a passive role in sex). For that reason, the Latin words are used in this book to avoid adding irrelevant shades of meaning, and English terms will be used only as abstractions and not to talk about historic individuals. Williams takes a Foucaultian position that the categories “heterosexual”, “bisexual”, and “homosexual” and the impulse to assign all people to them is specific to contemporary Western culture, and that even in Western culture something identifiable as [male] homosexual identity did not emerged until fairly late, e.g., the 17th century in England.

[Note: It is a general flaw in Foucault’s work that it is based almost entirely on studying male relations, with a blithe and tacit assumption that women’s experiences were either parallel or of no particular importance.]

This drive to categorize all persons into a homo/hetero binary (and he considers “bisexual” to be “reserved for intractable cases”) is parallel to categorizing all persons into a gender binary. In many cultures, the central prototype [my phrase] for “man” may have sex with specific classes of assigned-male persons [again, my phrase], but if those “permitted partners” are contextually defined as “not members of the central prototype for man” then in what sense is the relationship homosexual as opposed to representing a type of heterosexuality within a multi-gender system?

And what about relations between women? Ovid’s Metamorphoses includes many stories that glorify romantic and sexual relations between men, but in the tale of Iphis and Ianthe, Iphis is given a speech about how impossible and unnatural love between women is. [Note: But in all these we are inevitably working through the male gaze. Ovid himself discusses his desire for woman and for “boys” but he is not capable of experiencing desire for a woman as a woman. So is his claim of “impossibility” a genuine reflection of Roman society, or simply a personal failure of imagination--and of a specifically male imagination? When considering what sort of evidence Ovid and other male writers offer about Roman reality--i.e., that love between women necessarily involves a butch-femme dynamic with transgender implications--how is that different from 20th century heteronormative expectations of lesbian relationships that one partner must “be the man”? And, as we’ll see, on a symbolic and rhetorical level, Roman sexual systems always considered the “passive” partner in an act to be feminized, regardless of their biological sex.]

This book covers texts from the 2nd century BCE to the 2nd century CE and includes all genres: epigrams, graffiti, love poetry, rhetoric. Nearly all of it was written by men for a male readership and reflects a male understanding and experience of the world. These texts show the messy contradictions in relations between men because they focus on the whole range of experience and poke at those cases that don’t appear to fit the paradigm. But the nature of the texts means that there is not a similar impetus to explore how relations between women challenged the paradigm (because women’s sexual deviations did not call into question male status and privilege). Examining a wide variety of genres examines not only the lived experiences of Romans, but how those experiences were framed in texts and even how authors framed their own experiences and acts, either via poetic personae or for rhetorical purposes.

Chapter 1: Roman Traditions - no notes

Chapter 2: Greece and Rome

Even within the limited temporal scope, the possibility should be considered that there was change over time in sexual attitudes or practices. Roman texts portrayed a major wave of Greek influence around the 2nd century BCE with regard to pederasty (i.e., age-differentiated relationships between men with the older partner taking the “dominant” role), which in Greek culture was carried out openly and celebrated. But Roman texts show no actual change in the accepted sexual codes during the transition from the Republic to the Empire, only an increase in concerns and accusations that men whose social status required them to take only the active role were instead taking passive parts. Even so, there’s no clear evidence for a change in behavior, only the degree of rhetoric about it.

Chapter 3: The Concept of Stuprum

The official Roman sexual system was organized around penetrative acts and the orifice that was being penetrated (vagina, anus, mouth). The only exception to this penis-oriented system was cunnilingus, which was heavily stigmatized (and was considered to be a penetration of the mouth by the cunnus). The texts are unconcerned with non-penetrative sex acts (having defined oral sex as “penetrative” of the mouth) such as mutual masturbation, as well as ignoring non-sexual and emotion-based life partnerships.

Williams distills down three “rules” for the sexual behavior of the Roman vir, the high-status man. 1) He must only be the penetrator in sex, never penetrated; 2) Other than his lawful wife, he must never engage in sex with a member of the Roman citizen class, whether male or female; 3) The physical ideal for a partner is smooth and youthful. How do these rules apply to women of the same class? #1 is irrelevant, the Roman sexual system places women at the other end of the pole: never penetrating, available in every way for penetration. Rule #2 has some odd quirks. In theory, the citizen-class Roman woman should never engage in sex with anyone but her husband. But when you get to “degrees of badness” it’s worse for her to have sex with a non-citizen than a citizen. It’s unclear how #3 would apply, since we don’t have a significant body of evidence on what women were expected to consider attractive.

Chapter 4: Effeminacy and Masculinity

Other characteristics than simply playing a “passive” role in sex could be associated with effeminacy. The evaluation was not necessarily related to one’s sexual partner. “Feminine” was defined in opposition to accepted male virtues, not in relation to feminine virtues. So something could “feminize” a man that wouldn’t be considered virtuous in a woman. This included a concern with physical appearance and grooming, “softness” in general, walking delicately, particular ways of talking, wearing loose colorful clothing, using perfume, curling the hair, depilation, “foreign” luxuries, fine dining and drunkenness, excess emotional display, uncontrolled lust (of any type). Similarly, a man was “feminized” by performing cunnilingus without that implying that it was acceptable for women to perform oral sex on each other. It was the framing of oral sex as a “receptive/passive” role that aligned it with femininity.

What emerges in literature however is a clear thread of “counter-culture” of men rejecting  or downplaying these rules by embracing otherwise deprecated behaviors and roles. [Note: what we don’t have to the same extent -- because that body of literature is unconcerned with women -- is the same sort of evidence for women rejecting the Roman sexual system, even if only symbolically in writing.]

Chapter 5: Sexual Roles and Identities

[Note: In filtering for information even vaguely relevant to f/f contexts, I’m going to spend a lot of time talking about oral sex here -- not necessarily because this was the primary form of sexual interaction between Roman women, but because it was the act that had implications within the “official Roman sexual system” so it’s the one that men wrote about.]

The basic Roman sexual system can be understood as a matrix with one scale for insertive/receptive and another scale for the orifice involved with “higher status” falling higher and to the left in the table. (Williams uses “insertive” and “receptive” rather than “active” and “passive”.) Obviously, the acts default to assuming the presence of a penis. The verbs (from which other vocabulary is derived) are as follows:

Vaginal - (insertive) futuere; (receptive) crisare (referring to moving the body in response)
Anal - (insertive) pedicare; (receptive) cevere (referring to moving the body in response)
Oral -(insertive) irrumare; (receptive) fellare

The corresponding nouns get somewhat complicated by the question of gender.

Vaginal - (insertive) fututor (m), fututrix (f); (receptive) woman (no special term needed)
Anal - (insertive) pedicator (m, not sure if any f. examples); (receptive) pathicus, cinaedus* (m), (no clear female equialent, though pathica may be used but is not necessarily specific to this act)
Oral - (insertive) irrumator (m, no f. equivalent); (receptive)  fellator (m), fellatrix (f), cunnilinctor** (m, the f. would be cunnilinctrix but I don’t know that it occurs)
*cinaedus isn’t actually specific to this sexual role, but rather means “a man who doesn’t live up to the expectations of a vir.
**This is, of course, the exception to the assumption that there is a penis involved somewhere in the act.

The focus by male authors on penetration means that even when sex between women is mentioned, the question is “who penetrated whom?” often with the assumption of a dildo being used. Oral sex is something of an anomaly, but the framework for oral sex can be seen for fellatio which is still classified and treated as a penetrative act. The problematic position of cunnilingus in Roman texts is in part because of the inability to fit it neatly into a penetrative frame. Looking at the overall system, we can make sense of how oral sex was treated.

In description and especially self-description, the Latin word vir (man) emphasized a man who adhered to the official code and rules of Roman masculinity. Thus we can make sense of Martial’s epigram on Bassa, which frames her as behaving as a man (vir) by the simple act of engaging in sex with women. Or rather, that her genitals (venus) “falsely plays the man (vir)”. This implies penetration -- the defining characteristic of a vir. Considering the Roman sexual system as a penetrative/insertive system rather than an active-passive system helps in understanding. Lingo (licking) didn’t count as “penetration” no matter who did it, and despite the fact that it might be thought of as more “active” than the person experiencing the act. Even if only by analogy, the mouth was considered a receptive orifice. Thus, even if Bassa is having oral sex performed on her by a woman, being the recipient of the act inherently masculinizes her.

The default expectation was that a man might have preferences for particular sex acts, but it was not expected for him to prefer a particular gender as a partner. It was as unusual for a man to exclusively prefer female partners as for him to exclusively prefer male partners. Within this framework, all manner of personal preferences were recognized and even considered innate, but they were not considered to constitute some sort of personal identity in the same way that alignment on the insertive/receptive scale did.

Williams argues against viewing these preferences as “orientations” in the modern sense, but more equivalent to a preference for a particular body type or feature like hair color. A man might prefer blondes, but that doesn’t mean that all people can be categorized in terms of which hair color they’re “oriented towards.” So while Williams dismisses the claim by writers such as Boswell that descriptions of personal preference of this type are evidence of “sexual orientation” as we understand it, he grants that classical authors recognized the concept of an innate preference for certain types of sexual partners and certain types of sex acts.

Laws such as the Lex Scantinia, which touched on matters of sexual status and offense, were rarely actually invoked and had very limited scope regarding personal behavior, as long as stuprum was not inflicted on a freeborn Roman man or woman.

With regard to individual sexual preferences, it was implied that men who were known to be fellators (i.e., who habitually took the receptive role in oral sex with other men) were also likely to perform cunnilingus on women. This was considered even less reputable than accepting anal penetration. Performing oral sex of any type was considered “unmanly.” Romans considered oral sex to “foul” the mouth, and there are comments that a person who performs oral sex should not be kissed or share drinking vessels, or should wash their mouth out. The topic is discussed in the language of “impurity.”

Most textual references to oral sex are to fellatio (either m/m or m/f). Men who performed cunnilingus were the subject of as much or more stigma than a fellator. (According to the system, being on the receptive side of a sex act with a woman was more degrading than from a man because it inverted the assumed status relationship even more.) Oral sex was not treated as a mutual exchange between lovers. The poet Martial wrote a poem about a sexually eager female lover who agreed to accept all types of penetration from him but only if he would return the favor with cunnilingus, but he refused in disgust.

All of this brings us to Martial’s epigram on Philaenis, which frames her sexual activity in masculine terms. She penetrates boys and girls, she exercises at the wrestling school, she eats and drinks excessively [though note that excessive eating and drinking in a man would be “unmanly”]. She refuses to perform fellatio because it’s “unmanly” but Martial’s punchline is that she “thinks it manly to perform cunnilingus on girls.” The satire here is focusing on her upside down values. She, a woman, does all these masculine things, but fails because she embraces the most unmanly act of all. A man performing cunnilingus is framed as “submitting” to the woman.

[Note: This leaves open a lot of questions that one suspects the male Roman writers had no interest in. Would it be shameful for a free Roman woman to receive cunnilingus from an unfree or lower status woman? Did women care as much about the status dynamics of sex acts in the same way that men did, given that they were coming from an official position at the bottom of the ranking? Is it possible that Philaenis could, simultaneously, reject the role of fellator--sexually subjugating herself to a man--as part of her personal identity, and yet not consider performing cunnilingus to be a similar (or even worse) subjugation?]

Williams suggests interpreting the cinaedus (i.e., a man who habitually takes the receptive role in anal sex or in some other way goes against the expectations of virility) as a gender category rather than a sexual orientation. This is illustrated by the fable related by Phaedrus about how, when Prometheus was creating human beings, he drunkenly attached sexual organs to the wrong bodies in some cases. Thus the cinaedus was created by attaching a penis to a “female” body while the tribas was created by attaching a vagina to a “male” body. That is, their behavior is viewed as resulting from a type of gender dysphoria [my term] within an obligatorily heteronormative system. [Note: This must be distinguished from a theory of transgender identity as it assumes a fixed relationship of gender identity and preferred sex acts.] This system assumes that the partner of the cinaedus and the tribas are behaving “normally” in accordance with the sexual desires that their bodies dictate. This theory is echoed by the 5th century medical writer Caelius Aurelianus who suggests that molles (another term for cinaedus) and tribades also experience an excess of lust, which leads to other sexual vices besides taking the “wrong” role in insertive sex. He writes:

nam sicut feminae tribades appellatae, quod utramque venerem exerceant, mulieribus magis quam viris misceri festinant et easdem inidentia paene virili sectantur...

“Just as those women called ‘tribades’--because they engage in both kinds of sexual practice--seek intercourse with women more than with men and pursue women with almost a man’s jealousy...”

By this definition, a tribade might just as easily choose a passive man as her sexual partner. Her identity comes, not from the nature of her sexual partner (just as men’s nature is not determined by the gender of their sexual partner), but from the nature of the acts she desires to perform with them and the fact of taking the active role. Thus, Seneca describes masculine women as “drinking to excess and penetrating men” (apparently unconcerned with what they might be doing with women).

[Note: One fall-out from this understanding that I have seen in some writings on Roman sexuality is that it becomes possible to doubt the very existence of sexual acts between women unless the text is very explicit about it. This occurs in Adams’ The Latin Sexual Vocabulary where he dismisses same-sex interpretations of terms for sexually active women unless no other possible interpretation is available. This is something of a conundrum: within the context of the above understanding of the Roman sexual system, it’s true that terms like “tribas” or “cunnilinctor” or even “fututrix” do not automatically imply sex acts between women. But by the very same argument, there’s no reason why the possibility of a female partner should be dismissed. The flaw in Adams is not that he points out that a tribas might have a male partner, but that he requires a higher standard of proof that she might have a female one.]

Williams considers interpretations of the evidence that Rome had a “subculture” of male-male relations equivalent, for example, to the molly houses of 18th century England. These interpretations are based on references to men meeting in certain locations for sex, to particular fashions or habits associated with cinaedi, and so forth. But he argues that to have a subculture, you need to have people identifying as sharing an identity, and a social context where their interactions could not otherwise be engaged in freely or openly. This wasn’t the case in Rome. Rather than a "subculture" of male-male relations, his position is that male-male relations were simply part of the default culture.

Afterword to the Second Edition

Williams mentions as new data a Pompeii wall painting that Clarke (1998) thinks may represent sex between women. The image is part of a series depicting deprecated sexual practices.

Appendix 1: The Rhetoric of Nature

One of the contributions to later rhetoric about sexual morality comes from Roman texts about the concept of things being “according to Nature” or “against Nature”. Natura expresses not only what people believed existed but also how they believed they should be. Thus, the simple existence of something or some practice was not a defense against it being identified as “against nature.”

Seneca defines “vice” as anything that is “against nature” but includes clearly cultural practices among “nature”. But other authors--whether seriously or in satire--point out the arbitrary or ambiguous definitions of “nature.” For example, one could say that the design of an anus indicates that it’s “natural” for it to be penetrated.

Cultural fables like the one by Phaedrus about Prometheus assumes that molles and tribades have always been part of the human race, and therefore could reasonably be included in Natura. Seeking “nature” in the behavior of animals, we see Ovid’s speech given to Iphis where she claims [erroneously, as it happens] that no where among animals does a female desire a female. But at the same time, Ovid never questions the “natural” desire of men for boys.

Appendix 2: Marriage between Males

How are we to interpret various references in classical texts to marriages between men, especially as these are usually brought up in the context of political satire or personal attacks? Williams gives some credence to the practice of marriage between men, although in some cases the references were probably satirical. It appears that such marriages would not be entered into the official registry (but neither were all marriages between men and women). Marriages between men didn’t fit into the formal structures of Roman marriage because those structures were concerned with the begetting of legitimate freeborn children. Martial wrote a number of satires about “male brides” where the “bride” is made an object of scorn, but the marriage itself is not. Williams concludes that such marriages happened and the men considered themselves spouses, but the relationship was treated as anomalous and always involved treating one partner as being feminized by the relationship.

Appendix 3: A Note on the Sources -- no notes

Appendix 4: Pompeiian Graffiti in Context

Williams notes that among the wide variety of sexual “advertisements” in this genre, the only combination not attested is a woman unambiguously selling sexual services to another woman. However, there are references to a woman identified (possibly by herself) as a fututrix (a woman who fucks) which--by the normal understanding of the word--would imply a female receptive partner. (See discussion in Adams 1982.)

Time period: