I’m currently reading Alice Dreger’s book ‘Hermaphrodites and the Medical Invention of Sex’ and trying to clarify for myself a series of questions about sexual dimorphism and social construction claims about biological sex. The book itself is excellent and highly readable, and surely a valuable resource for anybody who is interested in the history of how intersex people have been treated by society (and specifically the medical establishment) over the course of the past two centuries. Dreger is an advocate for intersex rights, and supports the rights and humanity of intersex people. For these reasons I would recommend the book unhesitatingly to anyone interested in this subject. My take on it here will be quite narrow and critical however, because I don’t agree with the strong version of the ‘sex as a social construction’ thesis which frames the historical narrative.
Dreger is an historian who does not question the Michel Foucault / Judith Butler inspired poststructuralist model of inquiry, and describes the history of medical definitions of sex through the nineteenth and twentieth century without any reference to a universal, history and culture independent account of ‘true’ sex. In a characteristic passage she states:
So, if we want to sort, what should we employ as the necessary and/or sufficient traits of malehood and femalehood? What makes a person male or a female or a hermaphrodite? This is the problem. Today my own students, college students in history classes, sometimes in exasperation ask these questions of me at the end of a discussion of the history of sex, as if I am holding the ‘real’ answer from them. ‘What really is the key to being male, female, or other?’ But, as I tell them, … the answer necessarily changes with time, with place, with technology, and with the many serious implications – theoretical and practical, scientific and political – of any given answer. The answer is, in a critical sense, historical – specific to time and place. There is no ‘back of the book’ final answer to what must count for humans as ‘truly’ male, female, or hermaphroditic, even though the decisions we make about such boundaries have important implications. Certainly we can observe some basic and important patterns in the bodies we call ‘male’ and the bodies we call ‘female’. And the patterns we notice depend in part on the cognitive and material tools available at a given moment. But the development of new tools doesn’t get us closer and closer to some final, definite answer of what it is to be ‘truly’ male, female or hermaphroditic. Instead it only alters the parameters of possible answers. A hundred years ago we could not point to ‘genes’ in the way we can today, but being able to point to genes doesn’t mean that we have found the ultimate, necessary, for-all-time answer to what it means to be of a certain sex[i].
The theoretical model is that of a continuum or spectrum, with more male traits on one side and female traits on the other. In the middle is a blurry bit which gets labelled ‘intersex’ or ‘hermaphroditic’ (I’ll use the 20th century term ‘intersex’). Because the points which define boundaries between male, intersex and female change over time according to changing medical definitions and theories, ‘sex’ is a socially constructed form of categorisation. To put things very crudely, the wider and more fluctuating the blurry middle part is, the more arbitrary and less ‘real’ sex classes such as ‘male’ and ‘female’ are. There are at least three distinct claims which typically employ the sex as continuum model as a fundamental premise:
1. Biological sex is socially constructed
2. Biological sex is not real
3. Biological sex is mutable and changeable
I'll come back to claim (2) at the end of this post, but for now I will restrict my focus to the question of the nature of the continuum. Specifically, if we accept that sex can be thought of as a continuum, how big and how ‘blurry’ is the intersex part in the middle?
The Frequency of Intersex Conditions
Alice Dreger’s discussion of the complexities and pitfalls of arriving at a firm and precise percentage are worth recounting and careful consideration. Before she gives any statistical estimate, Dreger lists a number of reasons why ‘it is almost impossible to provide with any confidence an overall statistic for the frequency of sexually ambiguous births[ii]’. Specialist medical texts offer different frequency estimates, so we don’t know which sources to trust. The samples upon which they are based might not be representative of the entire human population. Sometimes very rare intersex conditions cluster in particular geographic regions, so it is hard to work these local variations into a global statistic. Environmental factors such as hormone treatments complicate the picture further.
More fundamental considerations include the question of what exactly counts as an intersex condition. Some types of intersex conditions result in very clear anatomical ambiguity, whereas other types are less severe or less obvious. Particular conditions may or may not manifest themselves as cases of sexual ambiguity. Should we count all of the people with all of the various conditions, or only the cases in which sex is difficult to determine? Dreger’s account is also centred around the idea that the category of ‘intersex’ is culturally and historically relative:
… such a statistic [of intersex frequency] is always necessarily culture specific. It varies with gene pool isolation and environmental influences. It also varies according to what, in a given culture, counts as acceptable variations of malehood or femalehood as opposed to forms considered sexually ambiguous. And it varies according to what opportunities there are in a given culture for doubts to surface and be articulated on record. […] Frequency is specific to particular cultural spaces[iii].
Nevertheless Dreger does eventually offer us a statistical estimate:
When I am pressed for a rough statistic, I suggest that today, in the United States, probably about one to three in every two thousand people are born with an anatomical conformation not common to the so called typical male or female such that their unusual anatomies can result in confusion and disagreement about whether they should be considered female or male or something else. Anne Fausto-Sterling, through recent research, estimates the incidence of intersexed births to be in the range of 1 percent, although Fausto-Sterling warns that the figure ‘should be taken as an order of magnitude estimate rather than a precise count.’ (In other words, the number might be closer to one in a thousand.)[iv]
Although I appreciated the reasons for Dreger’s reluctance to state an exact figure, I found this statement to be very odd. Was it one in a thousand or one in a hundred? It looks like Dreger might be inclined to a more conservative figure than Fausto-Sterling, but she does not explain why.
It turns out that Anne Fausto-Sterling is one of the authors of the only academic study which attempts to rigorously answer this question:
We surveyed the medical literature from 1955 to the present for studies of the frequency of deviation from the ideal male or female. We conclude that this frequency may be as high as 2% of live births. The frequency of individuals receiving corrective genital surgery, however, probably runs between 1 and 2 per 1000 live births (0.1 – 0.2%)[v]
Again, there are two estimates: a liberal figure of 2% and a conservative estimate which differs by an order of magnitude – quite a big gap! In a paper written before this research was carried out, Fausto-Sterling provides an even more ambitious frequency estimate, together with an almost lyrical evocation of the ‘sex as a continuum’ thesis:
For some time medical investigators have recognized the concept of the intersexual body. But the standard medical literature uses the term intersex as a catch-all for three major subgroups with some mixture of male and female characteristics: the so-called true hermaphrodites, whom I call herms, who possess one testis and one ovary (the sperm- and egg-producing vessels, or gonads); the male pseudohermaphrodites (the "merms"), who have testes and some aspects of the female genitalia but no ovaries; and the female pseudohermaphrodites (the "ferms"), who have ovaries and some aspects of the male genitalia but lack testes. Each of those categories is in itself complex; the percentage of male and female characteristics, for instance, can vary enormously among members of the same subgroup. Moreover, the inner lives of the people in each subgroup-- their special needs and their problems, attractions and repulsions-- have gone unexplored by science. But on the basis of what is known about them I suggest that the three intersexes, herm, merm and ferm, deserve to be considered additional sexes each in its own right. Indeed, I would argue further that sex is a vast, infinitely malleable continuum that defies the constraints of even five categories.
Not surprisingly, it is extremely difficult to estimate the frequency of intersexuality, much less the frequency of each of the three additional sexes: it is not the sort of information one volunteers on a job application. The psychologist John Money of Johns Hopkins University, a specialist in the study of congenital sexual-organ defects, suggests intersexuals may constitute as many as 4 percent of births[vi].
Although the four percent figure has been shown to be unsupported by the available evidence, it is hard not to conclude that Anne Fausto Sterling is quite keen on the idea that the percentage of intersex people is a lot bigger than people would tend to expect. Vast infinitely malleable continuums need room to blur boundaries and break categories, hundredths are way better than thousandths for this purpose.
All of the intersex advocacy organisations I have looked at use Fausto-Sterling’s figures, and the OII Australia site contains a detailed list of distinct conditions with respective percentages. The most up to date quoted figure seems to be 1.7%, taken from a book written by Fausto-Sterling in 2000, which bases its figures on exactly the same academic source I quoted from above.
The OII Australia site also refers to an academic paper critical of Fausto-Sterling’s frequency estimate by Dr. Leonard Sax. He argues that Fausto-Sterling’s definition of intersex as including ‘anything that deviates from the Platonic ideal of male and female bodies’ is far too broad. His definition of what counts as intersex leads to a radically different frequency estimate:
‘The available data support the conclusion that human sexuality is a dichotomy, not a continuum. More than 99.98% of humans are either male or female. If the term intersex is to retain any clinical meaning, the use of this term should be restricted to those conditions in which chromosomal sex is inconsistent with phenotypic sex, or in which the phenotype is not classifiable as either male or female. The birth of an intersex child, far from being “a fairly common phenomenon,” is actually a rare event, occurring in fewer than 2 out of every 10,000 births[vii].’
The OII Australia site argues against Sax’s narrow conceptualisation, and favours a broad and inclusive category which
“…encapsulates a range of atypical physical or anatomical sex characteristics. These share in common their non-conformance with medical and social sex and gender norms. This non-conformance with stereotypical standards for male and female is why intersex differences are medicalised in the first place and, while that remains the case, it makes sense to us to include them in a definition of intersex.
The difference between narrow and broad definitions in medicine is somewhat ideological. The exclusion of some diagnoses that embody atypical sex characteristics but not others seems, at least to us, to be irrational. Intersex people do not share the same identities, but we share common ground in the stigmatisation of our atypical sex characteristics.”
Clearly there is a very important question here as to whether scientific distinctions and definitions are purely ‘ideological’ or not. Also, the suggestion here that stigmatisation itself might be a criteria for what counts as ‘intersex’ is surely problematic: flat chested women and men with high pitched voices may well face stigma, that doesn’t mean that they are ‘intersex’. Another consideration is how ethical considerations intersect with scientific questions. Intersex people face marginalisation and stigma both because they have bodies which do not conform to biological norms and also because such conditions are in fact unusual. If the level of ‘unusualness’ is reduced, this would arguably lessen the sense of marginalisation and stigma. Should we base our definition of ‘intersex’ on social justice considerations or scientific theories?
Putting these complex and contentious philosophical questions to one side, what types of specific ‘moderate’ intersex conditions are we talking about here? Leonard Sax rejects several of the various types of intersex condition identified by Fausto-Sterling because they do not fit his scientific definition. By far the biggest category he objects to is that of ‘Late Onset Congenital Adrenal Hyperplasia’ (LOCAH). This single category accounts for a massive 1.5% of Fausto-Sterling’s 1.7% figure. What is this condition and what are its symptoms?
The broad category of Congenital Adrenal Hyperplasia (CAH) is described by a popular internet medical site as:
…an inherited (genetic) condition causing swelling of the adrenal glands. The condition is associated with a decrease in the blood level of a hormone called cortisol and an increase in the level of male sex hormones (androgens) in both sexes. Some people get a mild condition that produces no symptoms. Others (mainly baby boys) develop a severe form that can be life-threatening. Medical treatment to correct hormone levels is available. Surgery to improve the appearance of unusual genitalia (in girls) is sometimes considered.
Leonard Sax accepts that the severe (and very rare) form of CAH is a genuine intersex condition, but he denies that the milder version of LOCAH is an intersex condition. To sum up his analysis and put it bluntly: all men who have LOCAH are unambiguously male. Sometimes they experience balding. Many women who have LOCAH have no symptoms. Of those that do, symptoms include excessive body hair, infrequent periods and acne. A small percentage of women with LOCAH have a larger than average clitoris. If you want details, refer to his article through the link above (it’s technical but not that hard to understand). I checked out a couple of the studies Sax refers to. Dreger is correct when she warns about radically different estimates from academic sources – the papers referred to by Sax quote a frequency of around 1 or 2 in 1000 for LOCAH, very different from the 1.5% figure. I’m not an expert in medical science, so it is hard for me to understand this massive discrepancy. The best explanation appears to be that there is a fairly wide spectrum of conditions which include higher than average levels of male hormones in females (androgens), including rare conditions like LOCAH but also including much more common conditions such as Polycystic Ovary Syndrome (PCOS).
The conclusion to be drawn from these observations is that there appears to be a contemporary trend towards a much broader and more liberal definition of intersex conditions. This trend is in conflict with a more conservative scientific definition of intersex. Radically different frequency estimates are a consequence of this ideological conflict.
The fact that this sort of conflict exists is consistent with Alice Dreger’s over-arching postmodern narrative. She convincingly describes how doctors and medical specialists strived throughout the late nineteenth and early twentieth century for a very narrow definition of intersex. By defining males in terms of the presence of testicular tissue, and females in terms of ovarian tissue, the existence of ‘hermaphrodites’ was squeezed into an almost non-existent category. People with ‘ovotestes’ are extremely rare (0.0012% according to Fausto-Sterling), so by defining ‘hermaphroditism’ in this way, Victorian sensibilities could be preserved. The desire to preserve a rigid sex binary was linked to societal gender norms and fear of homosexuality. The existence of intersex people was a threat to the social order which needed to be contained, so narrow scientific definitions were sought after for reasons which were not always purely scientific.
Searching around the internet for alternative interpretations concerning intersex, I came across this tumblr blog which includes the most broad and liberal conception of ‘intersex’ that I could find. The authors of the blog insist that medical authorities should not have the sole right to determine who is and who is not intersex, and that people with PCOS are definitely intersex if they want to self define as intersex. Crucially, the question of whether someone is intersex or not has more to do with identification than it does with any sort of medical description. The ideological underpinning of this blog may well be indicative of a new historical era. Dreger’s phrase for the Victorian approach towards intersex people was ‘The Age of the Gonads’. This blog may well go down in history as representative of ‘The Age of Trans’.
Is sex really a ‘continuum’? Is sex ‘real’?
We can think of sex as a continuum, with male traits at one extreme and female traits at the other. All people will fall within the reach of two intersecting normal curves, the left curve representing mostly males and the right curve representing mostly females. The intersection in the middle is the group of intersex people. The problem for this very abstract model is the fact that biological sex can be conceptualised across a number of distinct axes. We could look at genetic factors such as chromosones, gonadal tissue, secondary characteristics (such as body hair), genital morphology, hormone levels or reproductive capacity. Every factor would produce a different sort of graph. The huge complexity of intersex conditions would defy any attempt to provide a realistic picture with such a simplistic model.
Professor Daphna Joel refers to a ‘3 G’ model of sex which defines sex on the basis of genetic, gonadal and genital factors. Using this model, 99% of all people fall into one of two categories, male and female. Males have all of these physical features: XY chromosones, testes, prostrate and seminal vesicles, penis and scrotum. Females have all of these features: XX chromosones, ovaries, womb and fallopian tubes, clitoris, vagina and labia. This model is described as ‘almost perfect dimorphism’. Intersex conditions mean that we cannot say that sex is absolutely dimorphic. Some conditions (such as those which involve ambiguous genitalia) are cases of unusual intermediate phenomenon. Some conditions (such as Complete Androgen Insensitivity) involve a set of mis-matching features (XY chromosones and testes combined with female genitalia). Joel emphasises how incredibly unusual this almost perfect dimorphism is: we can be 99% confident for example that a baby born with a penis will have all the other features ‘matching’ (testes and XY chromosones). There are very few natural phenomenon with this high degree of probabilistic uniformity.
I’m going to conclude by quoting a passage from an academic paper by Caroline New, who argues against the idea that intersex conditions support the notion that sex is not ‘real’:
Postmodern writers massively exaggerate intersexuality and misrepresent sexual attributes as continuous rather than as distributed dimorphically, despite the variations and overlaps on any one dimension. Do these variations mean that sexual difference is not real? Once again postmodern feminists have higher standards than anyone else for categorisation. Hawkesworth maintains that females and males are not ‘natural kinds’ because there is no set of properties possessed by every member of each of these groups (1997). From a realist point of view, ‘natural kinds’ are so called because they tell us something about the causal structures of the world. Causally important properties are contingently clustered, but in such a way that the presence of some properties renders the presence of others more likely – because there are common underlying properties that tend to maintain the clusters of features (Keil, 1989:43). Biological kinds can never meet the essentialist criteria postmodern thinkers implicitly require (Boyd, 1992). Biology is messy and complex, and its regularities take the form of tendencies rather than laws. In the case of sexual difference, these tendencies are strong, ‘the genotypic and phenotypic division of bodies into two sexes crosses species and millenia’ (Hull, forthcoming). Sexual difference, then, is a ‘good’ abstraction. Pace deconstructionists, it brings together characteristics that are internally connected, and the connections in question are substantial, not merely formal (Danermark et al 2002).
[i] Dreger, Alice D. ‘Hermaphrodites and the Medical Invention of Sex’. Harvard University Press,1998.
[ii] Ibid, p.40
[iii] Ibid, p.42
[iv] Ibid, p.42
[v] [Melanie Blackless, Anthony Charuvastra, Amanda Derryck, Anne Fausto-Sterling, Karl Lauzanne, Ellen Lee, 2000, How sexually dimorphic are we? Review and synthesis , in American Journal of Human Biology 04/2000; 12(2):151-166.]
[vii] Leonard Sax, 2002, How common is intersex? a response to Anne Fausto-Sterling, in Journal of Sex Research, 2002 Aug;39(3):174-8.