One of the revolutions in the study of history in the twentieth century might be called "minority history": the effort to recover the histories of groups previously overlooked or excluded from mainstream historiography. Minority history has provoked predictable skepticism on the part of some traditional historians, partly because of its novelty - which will, of course, inevitably wear off - and partly because the attitudes that previously induced neglect or distortion of minority history still prevail in many quarters. The most reasonable criticism of minority history (aside from the objection that it is sometimes very poor scholarship, against which no discipline is proof) is that it lends itself to political use, which may distort scholarly integrity.


    As a point about minority history as a genre this is not cogent: Since the exclusion of minorities from much historiography prior to the twentieth century was related to or caused by concerns other than purely scholarly interest, their inclusion now, even for purely political ends, not only corrects a previous "political" distortion but also provides a more complete data base for judgment about the historical issues involved. Such truth as is yielded by historical analysis generally emerges from the broadest possible synthesis of the greatest number of viewpoints and vantages: The addition of minority history and viewpoints to twentieth-century historiography is a net gain for all concerned.

       But at a more particular level political struggles can cause serious problems for scholars, and a curious debate now taking place among those interested in the history of gay people provides a relevant and timely example of a type of difficulty that could subvert minority history altogether if not addressed intelligently. To avoid contributing further to the undue political freight the issue has lately been forced to bear, I propose to approach it by way of another historical controversy, one that was - in its day - no less heated or urgent, but that is now sufficiently distant to be viewed with dispassion by all sides.

The conflict in question is as old as Plato and as modern as cladism, and although the most violent struggles over it took place in the twelfth and thirteenth centuries, the arguments of the ancients on the subject are still in use today. Stated as briefly and baldly as possible, the issues are these: Do categories exist because humans recognize real distinctions in the world around them, or are categories arbitrary conventions, simply names for things that have categorical force because humans agree to use them in certain ways? The two traditional sides in this controversy, which is called "the problem of universals," are "realists" and "nominalists." Realists consider categories to be the footprints of reality ("universals"): They exist because humans perceive a real order in the universe and name it. The order is present without human observation, according to realists; the human contribution is simply the naming and describing of it. Most scientists operate - tacitly - in a realist mode, on the assumption that they are discovering, not inventing, the relationships within the physical world. The scientific method is, in fact, predicated on realist attitudes. On the other hand, the philosophical structure of the modern West is closer to nominalism: the belief that categories are only the names (Latin: nomina) of things agreed upon by humans, and that the "order" people see is their creation rather than their perception. Most modern philosophy and language theory is essentially nominalist, and even the more theoretical sciences are nominalist to some degree: In biology, for example, taxonomists disagree strongly about whether they are discovering (realists) or inventing (nominalists) distinctions among phyla, genera, species, etc. (When, for example, a biologist announces that bats, being mammals, are "more closely related to" humans than to birds, is he expressing some real relationship, present in nature and detected by humans, or is he employing an arbitrary convention, something that helps humans organize and sort information but that bears no "truth" or significance beyond this utility?)

This seemingly arcane struggle now underlies an epistemological controversy raging among those studying the history of gay people. The "universals" in this case are categories of sexual preference or orientation (the difference is crucial). Nominalists ("social constructionists" in the current debate) in the matter aver that categories of sexual preference and behavior are created by humans and human societies. Whatever reality they have is the consequence of the power they exert in those societies and the socialization processes that make them seem real to persons influenced by them. People consider themselves "homosexual" or "heterosexual" because they are induced to believe that humans are either "homosexual" or "heterosexual." Left to their own devices, without such processes of socialization, people would simply be sexual. The category "heterosexuality," in other words, does not so much describe a pattern of behavior inherent in human beings as it creates and establishes it.

Realists ("essentialists") hold that this is not the case. Humans are, they insist, differentiated sexually. Many categories might be devised to characterize human sexual taxonomy, some more or less apt than others, but the accuracy of human perceptions does not affect reality. The heterosexual/homosexual dichotomy exists in speech and thought because it exists in reality: It was not invented by sexual taxonomists, but observed by them.[1]

Neither of these positions is usually held absolutely: Most nominalists would be willing to admit that some aspects of sexuality are present, and might be distinguished, without direction from society. And most realists are happy to admit that the same real phenomenon might be described by various systems of categorization, some more accurate and helpful than others. One might suppose that "moderate nominalists" and "moderate realists" could therefore engage in a useful dialogue on those areas where they agree and, by careful analysis of their differences, promote discussion and understanding of these issues.

Political ramifications hinder this. Realism has historically been viewed by the nominalist camp as conservative, if not reactionary, in its implicit recognition of the value and/or immutability of the status quo; and nominalism has generally been regarded by realists as an obscurantist radical ideology designed more to undercut and subvert human values than to clarify them. Precisely these political overtones can be seen to operate today in scholarly debate over issues of sexuality. The efforts of sociobiology to demonstrate an evolutionary etiology of homosexuality have been vehemently denounced by many who regard the enterprise as reactionary realism, an effort to persuade people that social categories are fixed and unchangeable, while on the other side, psychiatric "cures" of homosexuality are bitterly resented by many as the cynical folly of nominalist pseudoscience: Convince someone he shouldn't want to be a homosexual, persuade him to think of himself as a "heterosexual," and - presto!
- he is a heterosexual. The category is the person.

Whether or not there are "homosexual" or "heterosexual" persons, as opposed to persons called "homosexual" or "heterosexual" by society, is obviously a matter of substantial import to the gay community, since it brings into question the nature and even the existence of such a community. It is, moreover, of substantial epistemological urgency to nearly all of society,[2] and the gravity and extent of this can be seen in the case of the problems it creates for history and historians.

The history of minorities poses ferocious difficulties: censorship and distortion, absence or destruction of records, the difficulty of writing about essentially personal and private aspects of human feelings and behavior, problems of definition, political dangers attendant on choosing certain subjects, etc. But if the nominalists are correct and the realists wrong, the problems in regard to the history of gay people are of an entirely different order: If the categories "homosexual/heterosexual" and "gay/straight" are the inventions of particular societies rather than real aspects of the human psyche, there is no gay history.[3] If "homosexuality" exists only when and where people are persuaded to believe in it, "homosexual" persons will have a "history" only in those particular societies and cultures.

In its most extreme form, this nominalist view has argued that only early modern and contemporary industrial societies have produced "homosexuality," and it is futile and misguided to look for "homosexuality" in earlier human history.

"What we call 'homosexuality' (in the sense of the distinguishing traits of 'homosexuals'), for example, was not considered a unified set of acts, much less a set of qualities defining particular persons, in pre-capitalist societies… Heterosexuals and homosexuals are involved in social 'roles' and attitudes which pertain to a particular society, modern capitalism."[4]

If this position is sustained, it will permanently alter, for better or worse, the nature and extent of minority history.

Clearly it has much to recommend it. No characteristics interact with the society around them uniformly through time. Perceptions of, reactions to, and social response regarding blackness, blindness, left-handedness, Jewishness, or any other distinguishing (or distinguished) aspect of persons and peoples must necessarily vary as widely as the social circumstances in which they occur, and for this reason alone it could be reasonably argued that being Jewish, black, blind, left-handed, etc., is essentially different from one age and place to another. In some cultures, for example, Jews are categorized chiefly as an ethnic minority; in others they are not or are not perceived to be ethnically distinct from the peoples around them, and are distinguished solely by their religious beliefs. Similarly, in some societies anyone darker than average is considered "black"; in others, a complex and highly technical system of racial categorization classes some persons as black even when they are lighter in color than many "whites." In both cases, moreover, the differences in attitudes held by the majority must affect profoundly the self-perception of the minority itself, and its patterns of life and behavior are in all probability different from those of "black" or "Jewish" people in other circumstances.

There can be no question that if minority history is to merit respect it must carefully weigh such fundamental subtleties of context: Merely cataloguing references to "Jews" or to "Blacks" may distort more than it reveals of human history if due attention is not paid to the meaning, in their historical setting, of such words and the concepts to which they apply. Do such reservations, on the other hand, uphold the claim that categories such as "Jew," "black," or "gay" are not diachronic and can not, even with apposite qualification, be applied to ages and times other than those in which the terms themselves were used in precisely their modern sense? Extreme realists, without posing the question, have assumed the answer was no; extreme nominalists seem to be saying yes.

The question can not be addressed intelligently without first noting three points. First, the positions are not in fact as clearly separable as this schema implies. It could well be argued, for example, that Padgug, Weeks, et. al., are in fact extreme realists in assuming that modern homosexuality is not simply one of a series of conventions designated under the same rubric, but is instead a "real" phenomenon that has no "real" antecedent in human history. Demonstrate to us the "reality" of this homosexuality, their opponents might legitimately demand, and prove to us that it has a unity and cohesiveness that justifies your considering it a single, unparalleled entity rather than a loose congeries of behaviors. Modern scientific literature increasingly assumes that what is at issue is not "homosexuality" but "homosexualities"; if these disparate patterns of sexuality can be grouped together under a single heading in the present, why make such a fuss about a diachronic grouping?

Second, adherents of both schools fall prey to anachronism. Nearly all of the most prominent nominalists are historians of the modern U. S., modern Britain, or modern Europe, and it is difficult to eschew the suspicion that they are concentrating their search where the light is best rather than where the answers are to be found, and formulating a theoretical position to justify their approach. On the other hand, nominalist objections are in part a response to an extreme realist position that has been predicated on the unquestioned, unproven, and overwhelmingly unlikely assumption that exactly the same categories and patterns of sexuality have always existed, pure and unchanged by the systems of thought and behavior in which they were enmeshed.

Third, both extremes appear to be paralyzed by words. The nominalists are determined that the same word can not apply to a wide range of meaning and still be used productively in scholarly discourse: In order to have meaning, "gay," for example, must be applied only as the speaker would apply it, with all the precise ramifications he associates with it. This insistence follows understandably from the implicit assumption that the speaker is generating the category himself, or in concert with certain contemporaries, rather than receiving it from a human experience of great longevity and adjusting it to fit his own understanding. Realist extremists, conversely, assume that lexical equivalence betokens experiential equality, and that the occurrence of a word that "means" "homosexual" demonstrates the existence of "homosexuality," as the modern realist understands it, at the time the text was composed.

It is my aim to circumvent these difficulties as far as possible in the following remarks, and my hope that in doing so I may reduce the rhetorical struggle over "universals" in these matters and promote thereby some more useful dialogue among the partisans. Let it be agreed at the outset that something can be discussed, by modern historians or ancient writers, without being named or defined. (Ten people in a room might argue endlessly about proper definitions of "blue" and "red," but could probably agree instantly whether a given object was one or the other [or a combination of both].) "Gravity" offers a useful historical example. A nominalist position would be that gravity did not exist before Newton invented it, and a nominalist historian might be able to mount a convincing case that there is no mention of gravity in any texts before Newton. "Nonsense," realists would object. "The Latin gravitas, which is common in Roman literature, describes the very properties of matter Newton called 'gravity.' Of course gravity existed before Newton discovered it."

Both, of course, are wrong. Lack of attention to something in historical sources can in no wise be taken as evidence of its nonexistence, and discovery can not be equated with creation or invention. But gravitas does not mean "gravity"; it means "heaviness," and the two are not at all the same thing. Noting that objects have heaviness is entirely different from understanding the nature and operations of gravity. For adherents of these two positions to understand each other each would have to abandon specific nomenclature, and agree instead on questions to be asked of the sources. If the proper questions were addressed, the nominalist could easily be persuaded that the sources prove that gravity existed before Newton, in the sense that the operations of the force now designated gravity are well chronicled in nearly all ancient literature. And the realist could be persuaded that despite this fact the nature of gravity was not clearly articulated - whether or not it was apprehended - before Newton.

The problem is rendered more difficult in the present case by the fact that the equivalent of gravity has not yet been discovered: There is still no essential agreement in the scientific community about the nature of human sexuality. Whether humans are "homosexual" or "heterosexual" or "bisexual" by birth, by training, by choice, or at all is still an open question.[5] Neither realists nor nominalists can, therefore, establish any clear correlation - positive or negative - between modern sexuality and its ancient counterparts. But it is still possible to discuss whether modern conceptualizations of sexuality are novel and completely socially relative, or correspond to constants of human epistemology which can be documented in the past.

To simplify discussion, three broad types of sexual taxonomy are abbreviated here as types A, B, and C. According to Type A theories, all humans are polymorphously sexual, i. e., capable of erotic and sexual interaction with either gender. External accidents, such as social pressure, legal sanctions, religious beliefs, historical or personal circumstances determine the actual expression of each person's sexual feelings. Type B theories posit two or more sexual categories, usually but not always based on sexual object choice, to which all humans belong, though external pressures or circumstance may induce individuals in a given society to pretend (or even to believe) that they belong to a category other than their native one. The most common form of Type B taxonomy assumes that humans are heterosexual, homosexual, and bisexual, but that not all societies allow expression of all varieties of erotic disposition. Subsets or other versions of Type B categorize on the basis of other characteristics, e. g., a predilection for a particular role in intercourse. Type C theories consider one type of sexual response normal (or "natural" or "moral" or all three) and all other variants abnormal ("unnatural," "immoral").

It will be seen that Type A theories are nominalist to the extent that they regard categorizations like "homosexual" and "heterosexual" as arbitrary conventions applied to a sexual reality that is at bottom undifferentiated. Type B theories are conversely realist in predicating categories that underlie human sexual experience even when obscured by social constraints or particular circumstances. Type C theories are essentially normative rather than epistemological, but borrow from both sides of the universals question in assuming, by and large, that people are born into the normal category but become members of a deviant grouping by an act of the will, although some Type C adherents regard "deviants" as inculpably belonging to an "abnormal" category through mental or physical illness or defect.

That no two social structures are identical should require no proof; and since sexual categories are inevitably conditioned by social structure, no two systems of sexual taxonomy should be expected to be identical. A slight chronological or geographical shift would render one Type A system quite different from another one. But to state this is not to demonstrate that there are no constants in human sexual epistemology. The frequency with which these theories or variations on them appear in Western history is striking.

The apparent gender blindness of the ancient world has often been adduced as proof that Type B theories were unknown before comparatively recent times. In Plutarch's Dialogue on Love it is asserted that

"the noble lover of beauty engages in love wherever he sees excellence and splendid natural endowment without regard for any difference in physiological detail. The lover of human beauty [will] be fairly and equably disposed toward both sexes, instead of supposing that males and females are as different in the matter of love as they are in their clothes."[6]

Such statements are commonplaces of ancient lore about love and eroticism, to the extent that one is inclined to believe that much of the ancient world was completely unaware of differentiation among humans in sexual object choice, as I have myself pointed out at length elsewhere.[7] But my statements and the evidence on which they rest can easily be misapprehended. Their purport is that ancient societies did not distinguish heterosexuality from homosexuality, not that all, or even most, individuals failed to make such a distinction.

A distinction can be present and generally recognized in a society without forming any part of its social structure. In some cultures skin color is a major determinant of social status; in others it is irrelevant. But it would be fatuous to assume that societies that did not "discriminate on the basis of" [i. e., make inviduous distinctions concerning] skin color could not "discriminate" [distinguish] such differences. This same paranomastic subtlety must be understood in regard to ancient views of sexuality: City-states of the ancient world did not, for the most part, discriminate on the basis of sexual orientation, and, as societies, appear to have been blind to the issue of sexual object choice, but it is not clear that individuals were unaware of distinctions in the matter.

It should be obvious, for instance, that in the passage cited above Plutarch is arguing against precisely that notion that Padgug claims had not existed in precapitalist societies, i. e., Type B theories. Plutarch believes that a normal human being is susceptible to attraction to either gender, but his comments are manifestly directed against the contrary view. Which attitude was more common in his day is not apparent, but it is clearly inaccurate to use his comments as demonstration that there was only one view. The polemical tone of his remarks, in fact, seems good evidence that the position he opposes was of considerable importance. The whole genre of debates about the types of love of which this dialogue is a representative[8] cuts both ways on the issue: On the one hand, arguing about the matter and adducing reasons for preferring one gender to the other suggests a kind of polymorphous sexuality that is not predirected by heredity or experience toward one gender or the other. On the other, in each of the debates there are factions that are clearly on one side or the other of the dichotomy not supposed to have existed before modern times: Some disputants argue for attraction to males only; some for attraction to females only. Each side derogates the preference of the other side as distasteful. Sometimes bisexuality is admitted, but as a third preference, not as the general nature of human sexuality:

"Zeus came as an eagle to god-like Ganymede, as a swan came he to the fair-haired mother of Helen. So there is no comparison between the two things: one person likes one, another likes the other; I like both."[9]

This formulation of the range of human sexuality is almost identical to popular modern conceptions of Type B: Some people prefer their own gender; some the opposite; some both. Similar distinctions abound in ancient literature. The myth of Aristophanes in Plato's Symposium is perhaps the most familiar example: Its manifest and stated purpose is to explain why humans are divided into groups of predominantly homosexual or heterosexual interest. It is strongly implied that these interests are both exclusive and innate; that is stated outright by Longus, who describes a character as "homosexual by nature [physei]."[10]

[Note: Among many complex aspects of Aristophanes' speech in the Symposium as an indication of contemporary sexual constructs, two are especially notable. (1) Although it is the sole attic reference to lesbianism as a concept, male homosexuality is of much greater concern as an erotic disposition in the discussion than either female homosexuality or heterosexuality. (2) It is this, in my view, which accounts for the additional subtlety of age distinctions in male-male relations, suggesting a general pattern of older erastes and younger eromenonos. Age differential was unquestionably a part of the construct of sexuality among elements of the population in Athens, but it can easily be given more weight than it deserves. "Romantic love" of any sort was thought to be provoked by and directed toward the young, as is clearly demonstrated in Agathon's speech a little further on, where he uses the greater beauty of young males and females interchangeably to prove that Love is a young god. In fact, most Athenian males married women considerably younger than themselves, but since marriage was not imagined to follow upon romantic attachment, this discrepancy does not appear in dialogues on eros. David Dalperin argues in "Sex Before Sexuality" (in this volume) that the speech does not indicate a taxonomy comparable to modern ones, chiefly because of the age differential, although in fact the creatures described by Aristophanes must have been seeking a partner of the same age, since, joined at birth, they were coeval. What is clear is that Aristophanes does not imagine a populace undifferentiated in experience or desire, responding circumstantially to individuals of either gender, but persons with lifelong preferences arising from innate character (or a mythic prehistory).]

It is true that there were no terms in common use in Greece or Rome to describe categories of sexual preference, but it does not follow that such terms were wholly unknown: Plato, Athenaeus, and other writers who dealt with the subject at length developed terms to describe predominant or exclusive interest in the apposite gender.[11] Many writers, moreover, found it possible to characterize homosexuality as a distinct mode of erotic expression without naming it. Plautus, for example, characterized homosexual activity as the "mores of Marseilles," suggesting that he considered it a variant on ordinary human sexuality.[12] Martial found it possible to describe an exclusively heterosexual male, even though he had no terminology available to do so and was himself apparently interested in both genders.[13]

One even finds expressions of solidarity among adherents of one preference or another in ancient literature, as when Clodius Albinus, noted for his exclusively heterosexual interest, persecutes those involved in homosexual behavior,[14] or when a character who has spoken on behalf of love between men in one of the debates bursts out, "We are like strangers cut off in a foreign land…; nevertheless, we shall not be overcome by fear and betray the truth,"[15] or when Propertius writes, "Let him who would be our enemy love girls; he who would be our friend enjoy boys."[16] That there is a jocular tone to some of these statements, especially the last, is certainly attributable to the fact that the distinctions involved in no way affected the well-being, happiness, or social status of the individuals, owing to the extreme sexual tolerance of ancient societies; but it does not cast doubt on the existence of the distinctions. Even when preferences are attributed ironically, as is likely the case in Plato's placing the myth of sexual etiology in the mouth of Aristophanes, the joke depends on the familiarity of the distinctions.

Subtler indications of Type B taxonomies can also be found. In the Ephesiaca, a Hellenistic love novel by Xenophon of Ephesus, sexual categories are never discussed, and are clearly not absolute, but they do seem to be well understood and constitute an organizing principle of individual lives. Habrocomes is involved throughout only with women, and when, after his long separation from his true love Anthia, she desires to know if he has been faithful to her, she inquires only if he has slept with other women, although she knows that men have been interested in him, and it is clear that sex with a man would also constitute infidelity (as with Corymbus). It seems clear that Habrocomes is, in fact, heterosexual, at least in Anthia's opinion. Another character, Hippothoos, had been married to an older woman and attracted to Anthia, but is apparently mostly gay: The two great loves of his life are males (Hyperanthes and Habrocomes); he left all to follow each of these, and at the end of the story he erects a stature to the former and establishes his residence near that of the latter. The author tidies up all the couples at the end by reuniting Anthea and Habrocomes and introducing a new male lover (Clisthenes) for Hippothoos. This entire scenario corresponds almost exactly to modern conceptualizations: Some people are heterosexual, some homosexual, some bisexual; the categories are not absolute, but they are important and make a substantial difference in people's lives.

Almost the very same constellation of opinions can be found in many other preindustrial societies. In medieval Islam one encounters an even more overwhelming emphasis on homosexual eroticism than in classical Greek or Roman writing. It is probably fair to say that most premodern Arabic poetry is ostensibly homosexual, and it is clear that this is more than a literary convention. When Saadia Gaon, a Jew living in Muslim society in the tenth century, discusses the desirability of "passionate love,"[17] he apparently refers only to homosexual passion. There is the sort of love men have for their wives, which is good but not passionate; and there is the sort of love men have for each other, which is passionate but not good. (And what of the wives' loves? We are not told.) That Saadia assumes the ubiquity of homosexual passion is the more striking because he is familiar with Plato's discussion of homosexual and heterosexual varieties of love in the Symposium.[18]

Does this mean that classical Islamic society uniformly entertained Type A theories of human sexuality and regarded eroticism as inherently pansexual? No. There is much evidence in Arabic literature for the very same Type B dichotomies known in other cultures. Saadia himself cites various theories about the determination of particular erotic interests (e. g., astrological lore),[19] and in the ninth century Jahiz wrote a debate involving partisans of homosexual and heterosexual desire, in which each disputant, like his Hellenistic counterpart, expresses distaste for the preference of the other.[20] Three debates of this sort occur in the Thousand and One Nights, a classic of Arabic popular literature.[21] "Homosexuals" are frequently (and neutrally) mentioned in classical Arabic writings as a distinct type of human being. That the "type" referred to involves predominant or exclusive preference is often suggested: In tale 142 of the Nights, for example, it is mentioned as noteworthy that a male homosexual does not dislike women; in Night 419 a woman observes a man staring longingly at some boys and remarks to him, "I perceive that you are among those who prefer men to women."

A ninth-century text of human psychology by Qusta ibn Luqa treats twenty areas in which humans may be distinguished psychologically.[22] One area is sexual object-choice: Some men, Qusta explains, are "disposed towards" [yamilu ila] women, some toward other men, and some toward both.[23] Qusta has no terminology at hand for these categories; indeed, for the second category he employs the euphemism that such men are disposed toward "sexual partners other than women"[24]: obviously lack of terminology for the homosexual/heterosexual dichotomy should not be taken as a sign of ignorance of it. Qusta, in fact, believed that homosexuality was often inherited, as did ar-Razi and many other Muslim scientific writers.[25]

It has been claimed that "homosexuality" was viewed in medieval Europe "not as a particular attribute of a certain type of person but as a potential in all sinful creatures."[26] It is certainly true that some medieval writers evinced Type A attitudes of this sort: Patristic authors often address to their audiences warnings concerning homosexual attraction predicated on the assumption that any male might be attracted to another.[27] The Anglo-Saxon life of Saint Eufrasia[28] recounts the saint's efforts to live in a monastery disguised as a monk and the turmoil that ensued: The other monks were greatly attracted by Agapitus (the name she took as a monk), and reproached the abbot for bringing "so beautiful a man into their minister" ["forþam swa wlitigne man into heora mynstre gelædde," p. 344]. Although it is in fact a woman to whom the monks are drawn, the account evinces no surprise on anyone's part that the monks should experience intense sexual attraction toward a person ostensibly of their own gender.

Some theologians clearly regarded homosexual activity as a vice open to all rather than as the peculiar sexual outlet of a portion of the population, but this attitude was not universal and was often ambiguously or inconsistently held even by those who did most to promulgate it. Albertus Magnus and Thomas Aquinas both wrote of homosexual acts as sins that presumably anyone might commit, but both also recognized that it was somewhat more complex than this: Aquinas, following Aristotle, believed that some men were "naturally inclined" to desire sexual relations with other men - clearly a theory of type B - and Albertus Magnus considered homosexual desire to be a manifestation of a contagious disease, particularly common among the wealthy, and curable through the application of medicine.[29] This attitude is highly reminiscent of psychiatric opinion in late Victorian times, and a far cry from categorizing homosexuality simply as a vice.

"Sodomy" was defined by many clerics as the improper emission of semen - the gender of the parties and their sexual appetites being irrelevant - but many others understood sodomita to apply specifically to men who preferred sexual contact with other men, generally or exclusively, and sodomia to apply only to the sexual acts performed in this context.[30]

Medieval literature abounds in suggestions that there is something special about homosexuality, that it is not simply an ordinary sin. Many writers view it as the special characteristic of certain peoples; others argue that it is completely unknown among their own kind. There are constant association of homosexual preference with certain occupation or social positions, clearly indicating that it is linked in some way to personality or experience. The modern association of homosexuality with the arts had as its medieval counterpart a regular link with the religious life: When Bernard of Clairvaux was asked to restore life to the dead son of a Marquess of Burgundy he had the boy taken to a private room and lay down upon him. No cure transpired; the boy remained lifeless. The chronicler, who had been present, nonetheless found humor in the incident and remarked, "That was the unhappiest monk of all. For I've never heard of any monk who lay down upon a boy that did not straightaway rise up after him. The abbot blushed and they went out as many laughed."[31]

Chaucer's pardoner, also a cleric, appears to be innately sexually atypical, and his association with the hare has led many to supposed that it is homosexuality that distinguishes him.[32] Even non-Christians linked the Christian clergy with homosexuality.[33]

Much of the literature of the High Middle Ages that deals with sexual-object choice assumes distinct dispositions, most often exclusive. A long passage in the Roman d'Énéas characterizes homosexual males as devoid of interest in women and notable in regard to dress, habits, decorum, and behavior.[34] Debates of the period characterize homosexual preference as innate or God-given, and in the well-known poem "Ganymede and Helen" it is made pellucidly clear that Ganymede is exclusively gay (before the intervention of the gods): It is Helen's frustration at his inability to respond properly to her advances that prompts the debate.[35] In a similar poem, "Ganymede and Hebe," homosexual relations are characterized as "decreed by fate," suggesting something quite different from an occasional vice.[36] Indeed, the mere existence of debates of this sort suggests very strongly a general conceptualization of sexuality as bifurcated into two camps distinguished by sexual object-choice. Popular terminology of the period corroborates this: as opposed to words like sodomita, which might designate indulgence in a specific activity by any human, writers of the High Middle Ages were inclined to use designations like "Ganymede," whose associations were exclusively homosexual, and to draw analogies with animals like the hare and the hyena, which were thought to be naturally inclined to sexual relations with their own gender.

Akkain of Lille invokes precisely the taxonomy of sexual orientation used in the modern West in writing about sexuality among his twelfth-century contemporaries: "Of those men who employ the grammar of Venus there are some who embrace the masculine, others who embrace the feminine, and some who embrace both..."[37]

Clearly all three types of taxonomy were known in Western Europe and the Middle East before the advent of modern capitalist societies. It is, on the other hand, equally clear that in different times and places one type of theory has often predominated over the others, and for long periods in many areas one or two of the three may have been quite rare. Does the prevalence of one theory over another in given times and places reveal something about human sexuality? Possibly, but many factors other than sexuality itself may influence, deform, alter, or transform conceptualizations of sexuality among peoples and individuals, and much attention must be devoted to analyzing such factors and their effects before it will be possible to use them effectively in analyzing the bedrock of sexuality beneath them.

Nearly all societies, for example, regulate sexual behavior in some way; most sophisticated cultures articulate rationalizations for their restrictions. The nature of such rationalizations will inevitably affect sexual taxonomy. If "the good" in matters sexual is equated with procreation, homosexual relations may be categorically distinguished from heterosexual ones as necessarily excluding the chief good of sexuality. Such a moral taxonomy might create a homosexual/heterosexual taxonomy in and of itself, independent of underlying personal attitudes. This appears, in fact, to have played some role in the Christian West. That some heterosexual relations also exclude procreation is less significant (though much heterosexual eroticism has been restricted in the West), because there is not an easily demonstrable generic incompatibility with procreative purpose. (Compare the association of chest hair with maleness: Not all men have hairy chests, but only men have chest hair; hence, chest hair is thought of as essentially masculine; though not all heterosexual couplings are procreative, only heterosexual acts could be procreative, so heterosexuality seems essentially procreative and homosexuality essentially not.)

In a society where pleasure or the enjoyment of beauty are recognized as legitimate aims of sexual activity, this dichotomy should seem less urgent. And in the Hellenistic and Islamic worlds, where sexuality has traditionally been restricted on the basis of standards of decorum and propriety[38] rather than procreative purpose, the homosexual/heterosexual dichotomy has been largely absent from public discourse. Just as the presence of the dichotomy might be traceable to aspects of social organization unrelated to sexual preference, however, its absence must likewise be seen as a moot datum: As has been shown, individual Greek and Muslim writers were often acutely conscious of such a taxonomy. The prevalence of either Type A or Type B concepts at the social level, in other words, may be related more to other social structures than to personal perceptions of or beliefs about the nature of sexuality.

Another factor, wholly overlooked in previous literature on this subject, is the triangular relationship of mediated desire, beauty, and sexual stereotypes. It seems safe enough to assume that most humans are influenced to some degree by the values of the society in which they live. Many desires are "mediated" by the valorization accorded things by surrounding society, rather than generated exclusively by the desiring individual. If one posits for the sake of argument two opposed sets of social values regarding beauty and sex roles, it is easy to see how conceptualizations of sexual desire might be transformed to fit "mediated desire" resulting from either pole. At one extreme, beauty is conceived as a male attribute: Standards and ideals of beauty are predicated on male models, art emphasizes male beauty, and males take pride in their own physical attractions. Greece and the Muslim world approach this extreme: Greek legend abounds in examples of males pursued for their beauty, standards of beauty are often predicated on male archetypes (Adonis, Apollo, Ganymede, Antinous), and beauty in males is considered a major good, for the individual and for his society. Likewise, in the Muslim world, archetypes of beauty are more often seen in masculine than in feminine terms, beauty is thought to be a great asset to a man, and the universal archetype of beauty, to which even beautiful women are compared, is Joseph.

This pole can be contrasted with societies in which "maleness" and beauty are thought unrelated or even contradictory, and beauty is generally predicated only of females. In such societies "maleness" is generally idealized in terms of social roles, as comprising, for example, forcefulness, strength, the exercise of power, aggression, etc. In the latter type of society, which the modern West approaches, "beauty" would generally seem inappropriate, perhaps even embarrassing in males, and males possessing it would be regarded as "effeminate" or sexually suspect to some degree.

In nearly all cultures some linkage is expressed between eroticism and beauty, and it should not therefore be surprising that in societies of the former type there will be greater emphasis on males as sex objects than in those of the latter type. Since beauty is conceptualized as a good, and it is recognized to subsist on a large scale - perhaps even primarily - among men, men can be admired even by other men for their beauty, and this admiration is often indistinguishable (at the literary level, if not in reality) from erotic interest. In cultures of the latter type, however, men are not admired for their beauty; sexual interest is generally imagined to be applied by men (who are strong, forceful, powerful, etc., but not beautiful) to women, whose beauty may be considered their chief - or even sole - asset. In the latter case, expressions of admiration for male beauty will be rare, even among women, who will prize other attributes in men they desire.

These descriptions are deliberate oversimplifications to make a point: In fact, no society is exclusively one or the other, and elements of both are present in all Western cultures. But it would be easy to show that many societies tend more toward one extreme than the other, and it is not hard to see how this might affect the prominence of the homosexual/heterosexual dichotomy: In a culture where male beauty was generally a source of admiration, the dividing line between what some taxonomies would define as homosexual and heterosexual interest would be considerably blurred by common usage and expression. Expressions of admiration and even attraction to male beauty would be so familiar that they would not provoke surprise or require designation as a peculiar category. Persons in such a society might be uninterested in genital interaction with persons of their own sex, might even disapprove of it, but they would tend not to see romantic interest in male beauty - by males or females - as bizarre or odd or as necessetating special categorization.

In cultures that deemphasize male beauty, however, expressions of interest in it by men or women might be suspect. In a society that has established no place for such interest in its esthetic structures, mere admiration for a man's physical attraction, without genital acts, could be sharply stigmatized, and a strict division between homosexual and heterosexual desire would be easy to promulgate and maintain.

Female roles would also be affected by such differences: If women are thought of as moved by beauty, even if it is chiefly male beauty, the adoption of the role of the admirer by the woman will nor seem odd or peculiar. If women are viewed, however, as the beautiful but passive objects of a sexual interest largely limited to men, their expressing sexual interest - in men or women - may be disapproved.[39] George Chauncey has documented precisely this sort of disapproval in Victorian medical literature on "homosexuality": At the outset sexual deviance is perceived only in women who violate the sex role expected of them by playing an active part in a female-female romantic relationship. The "passive" female, who does not violate the expectations of sex role by receiving, as females are thought naturally to do, the attentions of her "husband," is not considered abnormal. Gradually, as attitudes and the needs of society to define more precisely the limits of approved sexuality change, attention is transferred from the role of the female "husband" plays to the sexual object choice of both women, and both come to be categorized as "homosexual" on the basis of the gender to which they are attracted.[40]

Shifts of this sort, relating to conceptions of beauty, rationalization of sexual limitations, etc., are supported, affected, and overlaid by more specific elements of social organization. These include patterns of sexual interaction (between men and women, the old and young, the rich and the poor, etc.), specific sexual taboos, and what might be called "secondary" sexual behavior. Close attention must be devoted to such factors in their historical context in assessing sexual conceptualizations of any type.

Ancient "pederasty," for example, seems to many to constitute a form of sexual organization entirely unrelated to modern homosexuality. Possibly this is so, but the differences seem much less pronounced when one takes into account the sexual context in which "pederasty" occurs. The age differential idealized in descriptions of relations between the "lover" and the "beloved" is less than the disparity in age between heterosexual lovers as recommended, for example, by Aristotle (nineteen years). "Pederasty" may often represent no more than the homosexual side of a general pattern of cross-generation romance.[41] Issues of subordination and power likewise offer parallel structures that must be collated before any arguments about ancient "homosexuality" or "heterosexuality" can be mounted. Artemidorus Daldianus aptly encapsulates the conflation of sexual and social roles of his contemporaries in the second century A. D. in his discussion of the significance of sexual dreams: "For a man to be penetrated [in a dream] by a richer and older man is good: for it is customary to receive from such men. To be penetrated by a younger and poorer is bad: for it is the custom to give to such persons. It signifies the same [i. e., is bad] if the penetrator is older and poor."[42] Note that these comments do not presuppose either Type A or Type B theories: They might be applied to persons who regard either gender as sexually apposite, or to persons who feel a predisposition to one or the other. But they do suggest the social matrix of a system of sexual distinctions that might override, alter, or disguise other taxonomies.

The special position of passive homosexual behavior, involving the most common premodern form of Type C theory, deserves a separate study, but it might be noted briefly that its effect on sexual taxonomies is related not only to status considerations about penetration, as indicated above, but also to specific sexual taboos that may be highly culturally variable. Among Romans, for instance, two roles were decorous for a free adult male, expressed by the verbs irrumo, to offer the penis for sucking, and futuo, to penetrate a female, or pedico, to penetrate a male.[43] Indecorous roles for citizen males, permissible for anyone else, were expressed in particular by the verbs fello, to fellate, and ceveo, not translatable into English.[44] The distinction between roles approved for male citizens and others appears to center on the giving of seed
(as opposed to the receiving of it) rather than on the more familiar modern active/passive division. (American prison slang expresses a similar dichotomy with the terms "catchers" and "pitchers.") It will be seen that this division obviates to a large degree both the active/passive split - since both the irrumator and the fellator are conceptually active[45] - and the homosexual/heterosexual one, since individuals are categorized not according to the gender to which they are drawn but to the role they play in activities that could take place between persons of either gender. It is not clear that Romans had no interest in the gender of sexual partners, only that the division of labor, as it were, was a more pressing concern and attracted more analytical attention.

Artemidorus, on the other hand, considered both "active" and "passive" fellatio to be categorically distinct from other forms of sexuality. He divided his treatment of sexuality into three sections - the natural and the legal, the illegal, and the unnatural - and he placed fellatio, in any form, among illegal activities, along with incest. In the ninth-century translation of his work by Hunain ibn Ishaq (the major transmitter of Aristotelian learning to the West), a further shift is evident: Hunain created a separate chapter for fellatio, which he called "that vileness of which it is not decent even to speak."[46]

In both the Greek and Arabic versions of this work the fellatio that is objurgated is both homosexual and heterosexual, and in both, anal intercourse between men is spoken of with indifference or approval. Yet in the Christian West the most hostile legislation regarding sexual behavior has been directed specifically against homosexual anal intercourse: Fellatio has generally received milder treatment. Is this because fellatio is more wildly practised among heterosexuals in the West, and therefore seems less bizarre (i. e., less distinctly homosexual)? Or is it because passivity and the adoption of what seems a female role in anal intercourse is particularly objectionable in societies dominated by rigid ideals of "masculine" behavior? It may be revealing, in this context, that many modern languages, including English, have skewed the donor/recipient dichotomy by introducing a chiastic active/passive division: The recipient (i. e., of semen) in anal intercourse is "passive"; in oral intercourse he is "active." Could the blurring of the active/passive division in the case of fellatio render it less obnoxious to legislative sensibilities?

Beliefs about sexual categories in the modern West vary wildly, from the notion that sexual behavior is entirely a matter of conscious choice to the conviction that all sexual behavior is determined by heredity or environment. The same individual may, in fact, entertain with apparent equanimity contradictory ideas on the subject. It is striking that many ardent proponents of Type C etiological theories who regard homosexual behavior as pathological and/or depraved nonetheless imply in their statements about the necessity for legal repression of homosexual behavior that it is potentially ubiquitous in the human population, and that if legal sanctions are not maintained everyone may suddenly become homosexual.

Humans of previous ages were probably not, as a whole, more logical or consistent than their modern descendants. To pretend that a single system of sexual categorization obtained at any previous moment in Western history is to maintain the unlikely in the face of substantial evidence to the contrary. Most of the current spectrum of belief appears to have been represented in previous societies. What the spectrum reveals about the inner nature of human sexuality remains, for the time being, moot and susceptible of many divergent interpretations. But if the revolution in modern historical writing - and the recovery of whatever past the "gay community" may be said to have- is not to be stillborn, the problem of universals must be sidestepped or at least approached with fewer doctrinaire assumptions. Both realists and nominalists must lower their voices. Reconstructing the monuments of the past from the rubble of the present requires quiet concentration.

Postscript

This essay was written five years ago, and several of the points it raises now require clarification or revision. I would no longer characterize the constructionist-essentialist controversy as a "debate" in any strict sense: One of its ironies is that no one involved in it actually identifies him- or herself as an "essentialist," although constructionists (of whom, in contrast, there are many)[47] sometimes so label other writers. Even when applied by its opponents the label seems to fit extremely few contemporary scholars.[48] This fact is revealing, and provides a basis for understanding the controversy more accurately not as a dialogue between two schools of thought, but as a revisionist (and largely one-sided) critique of assumptions believed to underlie traditional historiography. This understanding is not unrelated to my nominalist/realist analogy: One might describe constructionism (with some oversimplification) as a nominalist rejection of a tendency to "realism" in the traditional historiography of sexuality. The latter treated "homosexuality" as a diachronic, empirical entity (not quite a "universal," but "real" apart from social structures bearing on it); constructionists regard it as a culturally dependent phenomenon or, as some would have it, not a "real" phenomenon at all. It is not, nonetheless, a debate, since no current historians consciously defend an essentialist point of view.

Second, although it is probably still accurate to say that "most" constructionists are historians of the nineteenth and twentieth centuries, a number of classicists have now added their perspective to constructionist theory. This has broadened and deepened the discussion, although, strikingly, few if any historians of periods between Periclean Athens and the late nineteenth century articulate constructionist views.[49]

Third my own position, perhaps never well understood, has changed. In my book, Christianity, Social Tolerance and Homosexuality I defined "gay persons"[50] as those "conscious of erotic inclination toward their own gender as a distinguishing characteristic" (p. 44). It was the supposition of the book that such persons have been widely and identifiably present in Western society at least since Greco-Roman times, and this prompted many constructionists to label the work "essentialist." I would now define "gay persons" more simply as those whose erotic interest is predominantly directed toward their own gender (i. e., regardless of how conscious they are of this as a distinguishing characteristic). This is the sense in which, I believe, it is used by most American speakers, and although experts in the field may well wish to employ specialized language, when communicating with the public it seems to me counterproductive to use common words in senses different from or opposed to their ordinary meanings.

In this sense, I would still argue that there have been "gay persons" in most Western societies. It is not clear to me that this is an "essentialist" position. Even if societies formulate or create "sexualities" that are highly particular in some ways, it might happen that different societies would construct similar ones, as they often construct political or class structures similar enough to be subsumed under the same rubric (democracy, oligarchy, proletariat, aristocracy, etc. - all of which are both particular and general).[51]

Most constructionist arguments assume that essentialist positions necessarily entail a further supposition: that society does not create erotic feelings, but only acts on them. Some other force - genes, psychological forces, etc. - creates "sexuality," which is essentially independent of culture. This was not a working hypothesis of Christianity, Social Tolerance and Homosexuality. I was and remain agnostic about the origins and etiology of human sexuality.

Notes

1. For particularly articulate examples of "nominalist" history, see Robert A. Padgug, "Sexual Matters: On Conceptualizing Sexuality in History," Radical History Review 20
(1979): 3-33, reprinted in this volume; and Jeffrey Weeks, Coming Out: Homosexual Politics in Britain from the Nineteenth Century to the Present (London, 1977). Most older studies of homosexuality in the past are essentially realist; see bibliography in John Boswell, Christianity, Social Tolerance and Homosexuality (London, 1980), p. 4, n. 3.

2. It is of substantial import to several moral traditions, e. g., whether or not homosexuality is a "condition" - an essentially "realist" position - or a "lifestyle" - basically a "nominalist" point of view. For a summary of shifting attitudes on these points within the Christian tradition, see Peter Coleman, Christian Attitudes to Homosexuality (London,
1980), or Edward Batchelor, Homosexuality and Ethics (New York, 1980).

3. Note that at this level the debate is to some extent concerned with the degree of convention that can be sustained without loss of accuracy. It is conventional, for instance, to include in a history of the United States treatment of the period before the inauguration of the system of government that bears that title, and even to speak of the "colonial U. S.," although while they were colonies they were not the United States. A history of Greece would likewise, by convention, concern itself with all the states that would someday constitute what is today called "Greece," although those states may have recognized no connection with each other (or even have been at war) at various points in the past. It is difficult to see why such conventions should not be allowed in the case of minority histories, so long as sufficient indication is provided as to the actual relationship of earlier forms to later ones.

4. Padgug, "Sexual Matters," p. 59.

5. For the variety of etiological explanations to date see the brief bibliography in Boswell, Christianity, p. 9, n. 9. To this list should now be added (in addition to many articles) three studies: Alan Bell and M. S. Weinberg, Homosexualities: A Study of Diversity Amond Men and Women (New York, 1978); idem, Sexual Preference: Its Development in Men and Women (Bloomington, Indiana, 1981); and James Weinrich, Sexual Landscapes (New York, 1987). An ingenious and highly revealing approach to the development of modern medical literature on the subject of homosexuality is proposed by George Chauncey, Jr., "From Sexual Inversion to Homosexuality: Medicine and the Changing Conceptualization of Female Deviance," Salmagundi, no. 58-59 (Fall 1982-Winter 1983): 114-46.

6. Moralia 767: Amatorius, tans. W. C. Helmhold (Cambridge, Mass., 1961), p. 415.

7. Boswell, Christianity, Part I passim, esp. pp. 50-59.

8. See Boswell, Christianity, pp. 125-27.

9. Greek Anthology, trans. W. R. Paton (Cambridge, Mass., 1918) 1.65.

10. Daphnis and Chloe, 4.11. The term paiderastes here can not be understood as a reference to what is now called paedophilia, since Daphnis - the object of Gnatho's interest - is full grown and on the point of marriage. It is obviously a conventional term for "homosexual."

11. For Plato and Pollianus, see Boswell, Christianity, p. 30, n. 56; Athenaeus uses philomeirax of Sophocles and philogynes of Euripides, apparently intending to indicate that the former was predominantly (if not exclusively) interested in males and the latter in females. Cf. Scriptores physiognomici, ed. R. Foerster (Leipzig, 1893), 1:29, p. 36, where the word philogynaioi, "woman lover," occurs.

12. Casina, V. 4.957.

13. Epigrams 2.47.

14. Capitolinus, 11.7.

15. Boswell, Christianity, p. 127.

16. 2.4: Hostis si quis erit nobis, amet ille puellas: gaudeat in puero si quis amicus erit.

17. Saadia Gaon, Kitab al-'Amanat wa'l-I c tikhadat, ed. S. Landauer (Leyden, 1880), 10.7, pp. 294-97 (English translation by S. Rosenblatt in Yale Judaica Series, vol. 1: The Book of Beliefs and Opinions).

18. Kitab, p. 295.

19. Ibid.

20. Kitab mufakharat al-jawari wa'l-ghilman, ed. Charles Pellat (Beirut, 1957).

21. See discussion in Boswell, Christianity, pp. 257-58.

22. "Le Livre des caractères de Qostâ," ed. and trans. Paul Sbath, Bulletin de l'institut d'Egypte 23 (1940-41): 103-39. Sbath's translation is loose and misleading, and must be read with caution.

23. Ibid., p. 112.

24. "…waminhim man yamilu ila ghairihinna mini 'lghilmani…," ibid. A treatment of the fascinating term ghulam (pl. ghilman), whose meanings range from "son" to "sexual partner," is beyond the scope of this essay.

25. Qusta discusses this at some length, pp. 133-36. Cf. F. Rosenthal, "ar-Râzî on the Hidden Illness," Bulletin of the History of Medicine 52, no. 1 (1978): 45-60, and the authorities cited there. Treating "passive sexual behavior" (i. e., the reception of semen in anal intercourse) in men as a hereditary condition generally implies a conflation of Types A and C taxonomies in which the role of insertor with either men or women is thought "normal," but the position of the "insertee" is regarded as bizarre or even pathological. Attitudes toward ubnah should be taken as a special aspect of Muslim sexual taxonomy rather than as indicative of attitudes toward "homosexuality." A comparable case is that of Caelius Aurelianus: see Boswell, Christianity, p. 53; cf. Remarkds on Roman sexual taboos, below.

26. Weeks, Coming Out, p. 12.

27. See Boswell, Christianity, pp. 159-61.

28. Aelfric's Lives of Saints, ed. and trans. W. W. Skeat (London, 1881), p. 33.

29. Discussed in Boswell, Christianity, pp. 316 ff.

30. "Sodomia" and "sodomita" are used so often and in so many competing senses in the High Middle Ages that a separate study would be required to present even a summary of this material. Note that in the modern West the term still has overlapping senses, even in law: In some American states "sodomy" applies to any inherently nonprocreative sex act (fellatio between husband and wife, e. g.), in others to all homosexual behavior, and in still others only to anal intercourse. Several "sodomy" statutes have in fact been overturned on grounds of unconstitutional vagueness. See, in addition to the material cited in Boswell, Christianity, pp. 52, 183-184; Giraldus Cambrensis, Descriptio Cambriae, 2.7; J. J. Tierney, "The Celtic Ethnography of Posidonius," Proceedings of the Royal Irish Academy 60 (1960): 252; and Carmina Burana: Die Lieder der Benediktbeurer Handschrift. Zweisprachige Ausgabe (Munich, 1979), 95.4, p. 334 ("Pura semper ab hac infamia/nostra fuit minor Britannia"; the ms. Has Bricciavia).

31. Walter Map, De nugis curialium 1.23, trans. John Mundy, Europe in the High Middle Ages, 1150-1309 (New York, 1973), p. 302. Cf. discussion of this theme in Boswell, Christianity, chapter 8.

32. Prologue, 669ss. Of several works on this issue now in print see especially Monica McAlpine, "The Pardoner's Homosexuality and How it Matters," PMLA, January
1980, pp. 8-22; and Edward Schweitzer, "Chaucer's Pardoner and the Hare," English Language Notes 4, no. 4 /1967):247-250 (not cited by McAlpine).

33. See Boswell, Christianity, p. 233.

34. 8565ss; cf. Roman de la Rose 2169-74, and Gerald Herman, "The 'Sin Against Nature' and its Echoes in Medieval French Literature," Annuale Mediaevale 17 (1976):
70-87.

35. "Altercatio Ganimedis et Helene: Kritische Edition mit Kommentar," ed. Rolf Lenzen, Mittellateinisches Jarbuch 7 (1972): 161-86; English translation in Boswell, Christianity, pp. 381-389.

36. Boswell, Christianity, pp. 392-98.

37. The Anglo-Latin Satirical Poets and Epigrammatists, ed. Thomas Wright (London, 1872), 2:463.

38. The relationship between the words "propriety" and "property" is not coincidental, and in this connection is highly revealing. Although social attitudes toward sexual propriety in pre-Christian Europe are often touted as more humane and liberal than those which followed upon the triumph of the Christian religion, it is often overlooked that the comparative sexual freedom of adult free males in the ancient world stemmed largely from the fact that all the members of their household were either legally or effectively their property, and hence could be used by them as they saw fit. For other members of society what has seemed to some in the modern West to have been sexual "freedom" might be more aptly viewed as "abuse" or "exploitation," although it is of course silly to assume that the ability to coerce necessarily results in coercion.

39. Lesbianism is often regarded as peculiar or even pathological in cultures which accept male homosexuality with equanimity. In the largely gay romance Affairs of the Heart (see Boswell, Christianity, pp. 126-27) lesbianism is characteried as "the tribadic disease" [tes tribakes aselgeias] (s. 28). A detailed analysis of the relationship of attitudes toward male and female homosexuality will comprise a portion of a study I am preparing on the phenomenology of homosexual behavior in ancient and medieval Europe.

40. Cf. n. 5, above.

41. Since the publication of my remarks on this issue in Christianity, pp. 28-30, several detailed studies of Greek homosexuality have appeared, most notably those of Félix Buffière, Eros adolescent: la pédérastie dans la Grèce antique (Paris, 1980); and K. J. Dover, Greek Homosexuality (Cambridge, Mass., 1978). Neither work has persuaded me to revise my estimate of the degree to which Greek fascination with "youth" was more than a romantic convention. A detailed assessment of both works and their relation to my own findings will appear in the study mentioned above, no. 39.

42. Artemidorus Daldianus, Onirocriticon libri quinque, ed. R. Park (Leipzig, 1963) 1.78, pp. 88-89. (An English translation of this work is available: The Interpretation of Dreams, trans. R. J. White [Park Ridge, N. J., 1975]).

43. "non est pedico maritus:/quae faciat duo sunt: irrumat aut futuit" Martial 2:47 (cf. n. 14, above: pedico is apparently Martial's own coinage).

44. Ceveo is, that is, to futuo or pedico what fello is to irrumo: It describes the activity of the party being entered. The vulgar English "put out" may be the closest equivalent, but nothing in English captures the actual meaning of the Latin.

45. Futuo/pedico and ceveo are likewise both active.

46. Hunayn ibn Ishaq, trans., Kitab Tacbir ar-Ru'ya, ed. Toufic Fahd (Damascus, 1964), pp. 175-76.

47. For an overview of this literature since the material cited in note 1, see most recently Steven Epstein, "Gay Politics, Ethnic Identity: The Limits of Social Constructivism," Socialist Review 93/94 (1987): 9-54; also John D'Emilio, Sexual Politics, Sexual Communities: The Making of a Homosexual Minority in the United States, 1940-1970
(Chicago, 1983); and the essays in Kenneth Plummer, ed., The Making of the Modern Homosexual (London, 1981). See also note 48.

48. Three recent writers on the controversy (Steven Murray, "Homosexual Characterization in Cross-Cultural Perspective,"in Murray, Social Theory, Homosexual Realities [Gai Saber Monograph, 3] [New York, 1984]; Epstein, "Gay Politics"; and David Halperin, "Sex before Sexuality: Pederasty, Politics, and Power in Classical Athens" [in this collection] identify among them a dozen or more "constructionist" historians, but Murray and Halperin adduce only a single historian (me) as an example of modern "essentialist" historiography; Epstein, the most sophisticated of the three, can add to this only Adrienne Rich, not usually thought of as a historian. As to whether my views are actually "essentialist" or not, see further.

49. See, for example, Halperin, "Sex before Sexuality." Much of the controversy is conducted through scholarly papers: at a conference on "Homosexuality in History and Culture" held at Brown University in February 1987, of six presentations four were explicitly constructionist; two of these were by classicists. On the other hand, the standard volume on Attic homosexuality, K. J. Dover, Greek Homosexuality (New York, 1985), defies easy classification, but falls closer to an "essentialist" point of view than a "constructionist" one, and Keith DeVries's Homosexuality and Athenian Society, when it appears, will be a nonconstructionist survey of great subtlety and sophistication. See also David Cohen, Law, Society and Homosexuality in Classical Athens," Past and Present 117 (1987): 3-21. For the (relatively few) recent studies of periods between Athens and the late nineteenth century, see Saara Lilja, Homosexuality in Republican and Augustan Rome (Helsinki, 1983) (Societas Scientarium Fennica, Commentationes Humanarum Litterarum, 74); Alan Bray, Homosexuality in Renaissance England (London, 1982); James Saslow, Ganymede in the Renaissance: Homosexuality in Art and Society (New Haven, 1986); Guido Ruggiero, The Boundaries of Eros: Sex, Crime and Sexuality in Renaissance Venice (New York, 1985); Claude Courouve, Vocabulaire de l'homosexualité masculine (Paris, 1985).

50. An expression I use to include both women and men.

51. Of course, if a constructionist position holds that "gay person" refers only to one particular modern identity, it is then, tautologically, not applicable to the past.