Main SourcebooksAncientMedievalModern

Subsidiary SourcebooksAfricanEast AsianGlobalIndianJewishIslamicLesbian/GayScienceWomen

About IHSP Help Page IHSP Credits

Internet Medieval Sourcebook

John Thorp:

The Social Construction of Homosexuality (review article) (1992)

from Phoenix 46.1 (1992) p54-65

Introductory Note

FOUCAULT'S IDEA that even the deepest-lying sexual categories are social constructs has surely been one of the most powerful developments on the intellectual horizon in recent decades. It gives a path of explanation for much that was inexplicable; it liberates us from some of our own most mysterious values; it allows us to see things in other societies and ages that were simply overlooked before. It is an idea that has generated a cascade of work in ancient social history in the last few years, most notably from David Halperin, Froma Zeitlin, and the late John J. Winkler.[1] One of the most seductive claims to issue from this work is the claim that the very category "homosexual" is a social construct which is scarcely more than a hundred years old. This claim is made of course by Foucault himself, and is restated and defended with great clarity and vigour by David Halperin in his book One Hundred Years of Homosexuality (above, note 1). This book has a number of essays on ancient erotic subjects; it would be pleasant to recount them and dilate on them here. But the book's most important and striking claim -- its recurrent underlying thesis -- is that homosexuality is not a natural but a social category. It is only this thesis, and Halperin's defence of it, that I shall consider in this review article.

Think of the great number and variety of theories that have been produced in modern times to explain homosexuality, and think of their abject failure to crack the nut. On the surface homosexuality is extremely odd; "nature does nothing without a purpose," said Aristotle, and we are inclined to believe him. And while it is easy to understand why nature set the looser of limbs among us in the heterosexual case, it is very hard to descry her reason in the homosexual case.

Let me risk oversimplification and divide the theories that have been proposed into two categories; I will call them the planophysical theories and the doliophysical theories. The planophysical theories are those which cast homosexuality as an error of nature, a freak -- produced, no doubt, by nature, but not in accordance with her grand plan. To put it in Aristotelian language, homosexuality is produced when efficient causality fails to push in the direction that final causality seems to pull -- as in an automatic pilot that is broken. The general tradition of psychoanalytic theories of homosexuality falls into this line, it seems to me. Arrested maturation, unresolved Oedipal complexes, and the like all depict homosexuality as something that in each individual case should not have happened; homosexuality arises because a fundamental psychic development that normally does occur, in this case, for some reason did not. These psychoanalytic theories have in general failed to command assent because they have failed to produce anything approaching causal laws. Thus, a weak father may cause gender confusion in a son because he fails to counterbalance the mother's influence, or the same weak father may cause strong male identification in the son by way of compensation: it so often seems that one and the same cause can produce quite contrary results, and this has not the makings of a causal law.[2] Also among the planophysical theories are the hormonal ones; they are perhaps the most promising in this category, but they seem to be systematically dogged by disputes about experimental method.[3]

The doliophysical theories are another matter. They are the theories that see homosexuality not as out of kilter with the grand plan of nature, but as definitely within it, produced by the very subtlety of nature's devices to achieve her ends: homosexuality in some part of a population might be a ruse of nature to promote that population's survival. I have in mind the sociobiological theories: the superior heterozygote fitness theory, the kin-selection theory, and the parental manipulation theory.[4] Each of these theories is designed to show, with minute attention to the detail of the workings, how it could be that a gene for homosexuality[5] might be encouraged in a population by natural selection, up to a certain equilibrium state. I admire these theories very much for their brilliant ingenuity, but I believe that there are difficulties of detail with each of them. Moreover, I also think that the whole project of understanding homosexuality as genetically based does not get over an initial hurdle of implausibility: each of these theories would require homosexuality to run in families in ways that nobody would seem to have any prima facie reason to think that it does.

The theoretical state is thus very unsatisfactory; neither the planophysical nor the doliophysical seems a really hopeful path of enquiry. And if homosexuality has not got there either by the oversight of nature or by a ruse of nature, how has it got there? Into this puzzling dilemma steps Foucault with a sort of Copernican revolution, proposing that homosexuality got there because we put it there. We created the category and gave it the importance which it seems to have; we drew together a series of practices and tastes, gave them a single name, and postulated their psychic depth. And now we seem to be mystified as to how this strange creature got there. Foucault writes: "homosexuality appeared as one of the forms of sexuality when it was transposed from the practice of sodomy into a kind of interior androgyny, a hermaphrodism of the soul. The sodomite had been a temporary aberration; the homosexual was now a species."[6] The very word "homosexual" came into English only in 1892, formed after a German neologism coined about twenty years earlier.[7] Homosexuality, then, is a social construct of our own culture, and virtually even of our own century. What we mean by "homosexuality" did not exist in Greece; there is no such thing as Greek homosexuality; the title of Dover's book designates the null set.

This thesis comes in both a strong and a weak form. The weak form is that different people naturally have a whole array of different sexual tastes and desires; what we have done is to categorize and label these in such a way that the great divide is established upon the gender of the object of desire, rather than upon its shape, size, vigour, colour, or social class. We have drawn the conceptual lines, and now are puzzled by them. The stronger form of the thesis is that the desires themselves have been socially produced: the category does not just group the desires, it creates them. The sociologist Jeffrey Weeks writes: "Social processes construct subjectivities not just as categories but at the level of individual desires."[8]

The defence of this thesis requires that one fight off counterexamples that come to mind, and the principal counterexample which comes to our minds, schooled as we are, is that of ancient Greece, which seemed to admit a category of homosexuality much like our own (though, to be sure, there was no settled word for it). Foucault engaged this counterexample, to mixed reviews, and David Halperin, in his 1990 book One Hundred Years of Homosexuality, has done the same, very convincingly and impressively. (The book is in many ways a thrilling one, not least because it is a work in classics by a classicist which pertains directly to understanding a socially very important modern phenomenon: it is relevant!) It does indeed seem that our category of homosexuality does not map onto -- is not congruent with -- any ancient Greek category of sexual behaviour. There is much zest and dazzle in both Foucault's and Halperin's treatment, and one does wonder, at the end, if despite some striking differences, the Greek category was really profoundly other than ours. What I want to do here is to sort out what I think are the main strands of difference that Foucault and Halperin argue, consider them, and ask how much of an overall difference, in the end, they amount to.

I think that at least three important differences are alleged. The first is the matter of the psychic depth of the phenomenon of males' erotic desire for males and females' for females. Foucault, in the striking remark which I quoted above, claims that what had been merely the practice of sodomy got transformed, in the thought of the nineteenth century, into an interior androgyny, a hermaphrodism of the soul. What Foucault seems to be suggesting is that sexual preference may itself be nothing more than a superficial matter of taste and practice, like the practice of opening breakfast eggs at the big or the little end, or the preference for white or dark meat of fowl, and that it need have no deeper roots in the soul than we generally suppose these cases to have. If this is true, it is no wonder that the quest to understand the deep cause of homosexual preference has been unsuccessful: it is a quest for something that isn't there.

But there are certainly some passages in the literature of antiquity which suggest that sexual preference was seen as a very profound matter. The one that comes first to mind is Aristophanes' speech in the Symposium, in which, after all, we are being offered, playfully,[9] an archeogenealogical explanation of the three different sexual orientations. Halperin's book has an extensive discussion of this passage, in which he defends his thesis against the apparent counterexample.

I would like to make three observations about Aristophanes' explanation, or rather about his explanandum. The first is that he doesn't seem to want to explain just an anatomical or mechanical sexual preference, but a whole way or tenor of life. The males who desire males do not just want to copulate with them, but to spend their lives with them (192c); they marry and have children only in deference to the usage of society (*u'po\ tou^ no/mou a'nagka/zontai*, 192b). The second point is that Aristophanes explicitly says that the sexual pleasure cannot explain the desire of these males to be together; their souls are longing for something else which they cannot name (192d). Far from being a superficial point of taste, the homosexual desire and activity is an expression of something which lies deep in the soul -- so deep that its nature is altogether unclear. The third point is that with this genealogical explanation, Aristophanes is effectively assigning the three groups to something like different races or species; he is undertaking to tell the story of the origins of three different races. (He even speaks of the three original sexes as *ge/nh*, 190b.) Foucault's idea that homosexuality used to be a superficial matter of practice and became a deep psychic category only in the nineteenth century seems to me pretty securely contradicted by Aristophanes' speech in the Symposium.

The second important difference that is alleged, notably by Halperin, between what we call homosexuality and the groups that are portrayed by Aristophanes is that where we have two categories of sexual orientation, homosexual and heterosexual, the speech of Aristophanes, at least, recognized three, corresponding to the three original *ge/nh*. I must say that I am inclined to regard this as a weak argument, not because Halperin seems wrong about Aristophanes' categories, but because he seems wrong about ours. I don't know how one would prove or even test this, but it seems to me that an exhaustive distinction between homosexual and heterosexual is for us very much an abstract and "scientific" distinction, which does not really correspond to the categories we commonly think in. We speak easily of gays and lesbians, and it seems somehow artificial to make the category of "gay" include that of "lesbian." Again, consider the widespread false etymology which derives the word "homosexual" from homo, hominis. I think that, at least at a common level, we think in three categories here just as Aristophanes did: gay, lesbian, and straight.

The third point of difference alleged by Halperin is that while we think of a homosexual couple as social equals and a homosexual relationship as a symmetrical one the Greek conception was altogether different. There was a difference of age and of sexual role between *e'ra/sths* and *e'rw/menos*. A strong tabu seems to have surrounded sexual relations between adult male citizens; and Halperin is probably right in his brilliant analysis of this tabu as having a political basis: a pathic is by definition submissive, and the fabric of the state required that citizens not be submissive. A citizen may penetrate whom or whatever he pleases, but a citizen must not be penetrated. However that may be, our symmetrical concept of homosexuality contrasts rather starkly with the Greek asymmetrical concept of pederasty. I would like to make two comments about this difference.

Those who write about this subject frequently adduce examples from wider anthropological studies, in an effort to jog us out of the universal projection of our own categories. One of the practices that is freely mentioned is that of the New Guinea tribe in which boys from the ages of eight to fifteen are orally inseminated on a daily basis by young warriors. Or again there is the Cretan practice, recorded by Ephorus, of a homosexual relationship as an initiation rite.[10] The implication seems to be that all the male population is involved in such rituals first in boyhood, on the one side, and then in young manhood, on the other. Now in reaching for an anthropological understanding of homosexuality we need to distinguish sexual arrangements in which the whole male population is involved in same-sex relations in limited contexts or for limited times, from arrangements in which only some part of the male population is involved in same-sex relations throughout their lives. Is it all the people some of the time, or some of the people all the time? In the case of classical Greek practice there is a strong current of scholarship which sees the same-sex relations as pretty well universal in the male population, but limited in time and context: the relic of an initiation rite. (One detects a sense that being an initiation rite somehow makes homosexuality acceptable -- boys will be boys, and moreover, they'll get over it!) It is important to notice that Aristophanes' account is rather precisely at variance with this. He says of the slices of the original male that when they are boys they love men and like to lie with them and be entwined in their limbs (*filou^si tou\s a'/ndras kai\ xai/rousi sugkatakei/menoi kai\ sumpeplegme/noi toi^s a'ndra/si*, 191e8); and then when they grow up they in turn love boys. So, while it is true that the approved homosexual relations are not between social equals, it is also true that, in Aristophanes' modelling at least, there is a category of males who by nature (*fu/sei*, 192b1) prefer males all their lives.

Secondly, while it may be true that the approved homosexual relationship is not between coevals, we know perfectly well that such relationships existed: indeed we have one right here in the Symposium: that between Agathon and Pausanias. (I leave aside the attempted seduction of Socrates by Alcibiades; the levels of irony and ├╗play in this incident are so hard to untangle that it is dangerous to try to draw any lessons from it.) As Aristophanes gets into the peroration of his speech it becomes clear that his theory has the rather shocking consequence that a life-long love bond between two men is the most blessed state of all -- that is precisely the socially disapproved form of homosexual liaison. At this point Eryximachus, whose speech in praise of love had been an exercise in solemn wit and ponderous banality, and whose humour is stuck in the locker room (Eryximachus the medicus aeternus), is told by Aristophanes not to make fun of the speech by saying that it is about Agathon and Pausanias. And the reason he should not mock the speech in this way is not that it is not about Agathon and Pausanias, but that perhaps it is: perhaps even these (*i'/sws ... kai\ ou`^toi*) are of the blessed number who have found their other halves, and are both really (*th\n fu/sin*) male (193b6). Much can be read from this phrase *i'/sws ... kai\ ou`^toi*: the relationship between Agathon and Pausanias is looked at somewhat askance; indeed they are perhaps notorious[11] (though they are scarcely pariahs, for they are, after all, present at the party); but one morally daring result of Aristophanes' archaeogenealogical theory is that such relationships are elevated to sanctity. But whether such relationships should be praised or not it is at least unquestionable that they existed, that everyone knew they existed, and that they were not, in fact, approved.

The disapproval of homosexual relationships between adult male citizens is well-known, of course -- and no doubt Halperin is right to locate the disapproval in the implication of such a relationship that at least one citizen of the pair is a pathic. I think we find this disapproval strikingly conveyed by Aristotle in Book 7 of the Nicomachean Ethics, though the point has been missed in many of the translations. In this chapter he is describing morbid and brutish states of character, states of character which are beyond vice and weakness of will. He instances the habit of plucking out the hair, of gnawing the nails, of eating coal, and *h` tw^n a'frodisiw^n toi^s a'/rresin e`/cis* (1148b25). The Oxford translation of this phrase is "paederasty," and Barnes has kept it, but I am convinced it is quite wrong. It is not pederasty that Aristotle has in mind here, but the habit of the adult catamite, the pathic, the *ki/naidos*.[12] The whole passage is somewhat loose in structure, but it seems clear that Aristotle regards these states as arising either from habit or by nature: and he goes on to say that, when they arise by nature, no one would ascribe them to weakness of will, any more than one would say that it is through weak will that women play the passive part in copulation (*ou'k o'pu/ousin a'll' o'pu/ontai*). This analogy with the role of women in copulation seems to make it clear that # t#n #frodisi#n to#w #rresin #jiw is the predilection not of the pederast but of the catamite. And whether the predilection arise by custom or by nature, it is clear that Aristotle does not admire it. (There is an interesting passage in the Problems 4.26, which explains a natural disposition to the passive role as stemming from having a physiology which delivers moisture to the fundament -- as the Oxford translation so delicately puts it -- rather than to the penis and testicles.)

I have been considering the thesis of Foucault and Halperin and others that homosexuality is a modern category and that the Greeks did not have it. It is true that they did not have the word. And it is true that there are differences over what kinds of homosexual activity are or are not approved and why. But the speech of Aristophanes in the Symposium seems strong evidence that a) it was reasonable to portray males' desire for males as psychically deep -- the notion of hermaphrodism of the soul would have made sense to the Greeks; b) whatever may or may not have been the case about universal initiation rite pederasty, there clearly was taken to be a class of males who had a life-long predilection for males, and moreover a further, less approved, class whose predilection did not respect the age asymmetry that is part of pederasty. Greek homosexuality seems very close to our own category in fundamental ways.

Of course, it may still be the case that homosexuality is a social construct: but if so it is striking that the Greeks and we have constructed it so similarly.



This review article began as a talk to the Classical Association of Canada, in May 1990, under the title "Was there homosexuality in classical Greece?" I am grateful to the audience of that day for their questions and suggestions.
John J. Winkler, The Constraints of Desire: The Anthropology of Sex and Gender in Ancient Greece (New York and London 1990); David Halperin, John Winkler, and Froma Zeitlin (eds.), Before Sexuality: The Construction of Erotic Experience in the Ancient Greek World (Princeton 1990); David Halperin, One Hundred Years of Homosexuality: And Other Essays on Greek Love (New York and London 1990).
For an interesting discussion of the inadequacy of Jungian theory on the question of homosexuality see Robert H. Hopcke, Jung, Jungians, and Homosexuality (Boston 1989).
See the discussion in Michael Ruse, Homosexuality: A Philosophical Inquiry (Oxford 1988) 84-129.
See, for example, Michael Ruse, Sociobiology: Sense or Nonsense? (Dordrecht 1979) 90-93.
Or a gene for homosexual-rearing. The parental manipulation theory holds that the phenotype which is passed on genetically is not homosexuality, but the parental behaviour which rears a child as homosexual. The sociobiological theories are therefore, as such, committed neither to nature nor to nurture as the source of homosexual leanings.
Michel Foucault, The History of Sexuality. 1: An Introduction, tr. Robert Hurley (New York 1980) 43.
In 1869 "homosexuelle" was coined by Karl Maria Kertbeny; in the same year Carl Westphal introduced the phrase "die conträre Sexualempfindung." The terminological history is told by Halperin, 155, n. 2.
Jeffrey Weeks, "Discourse, Desire and Sexual Deviance: Some Problems in a History of Homosexuality," in Kenneth Plummer (ed.), The Making of the Modern Homosexual (London 1981) 111; quoted in Halperin 40.
Although the tone is playful and the dramatic purpose is to advance a theory which is subsequently to be rejected, still certain inferences are possible about the social phenomenon which Plato's Aristophanes took himself to be describing. Thus far, if no further, the speech is a useful source for social history. See Sir Kenneth Dover, "Aristophanes' Speech in Plato's Symposium," The Greeks and Their Legacy: Collected Papers, Vol. 2 (Oxford and New York 1988) 102-114.
See Jan Bremmer, "An Enigmatic Indo-European Rite: Paederasty," Arethusa 13 (1980) 279-298; Dover (above, n. 9).
Agathon, certainly, was notoriously effeminate. See Aristophanes Thesmophoriazousai 130 ff.
The German translation of Olof Gigon (Zurich and Stuttgart 1967) gives die Knabenliebe; H. Apostle (Dordrecht and Boston 1975), J. A. K. Thomson (Harmondsworth 1955), J. Tricot (Paris 1967), and J. Voilquin (Paris 1965) all translate the phrase by some form of "homosexual"; Aquinas in his commentary gives the neutral uti coitu masculorum, and he is followed in this neutrality by R. E. Gauthier and J. Y. Jolif (Louvain and Paris 1970), who have faire l'amour avec les mâles, and by A. G. Robledo (Mexico 1972): el comercio sexual entre los machos. D. P. Chase (London and New York 1942) translates the phrase by suspension points.


The Internet History Sourcebooks Project is located at the History Department of  Fordham University, New York. The Internet Medieval Sourcebook, and other medieval components of the project, are located at the Fordham University Center for Medieval Studies.The IHSP recognizes the contribution of Fordham University, the Fordham University History Department, and the Fordham Center for Medieval Studies in providing web space and server support for the project. The IHSP is a project independent of Fordham University.  Although the IHSP seeks to follow all applicable copyright law, Fordham University is not the institutional owner, and is not liable as the result of any legal action.

© Site Concept and Design: Paul Halsall created 26 Jan 1996: latest revision 6 October 2023 [CV]