Published as Hegarty, P. (1997).Materializing the hypothalamus: A performative account of the ‘gay brain.’ Feminism & Psychology, 7, 355-372.
Abstract
Simon LeVay's research on neuroscience and sexuality has been reiterated in popular media, scientific communities and legal debates. A close reading of this work, drawing on performativity theory (Butler, 1990, 1993 ), reveals that this popular success is the result of citing and reiterating a number of heterosexist, sexist and culturally imperialist norms. LeVay's work excludes women and ethnic minorities and denies the political, cultural and historical nature of sexuality. Performativity theory suggests the limits of empiricism for feminists, and the importance of postmodern readings of the subject of psychology and neuroscience.
In late August and early September of 1991 one could easily have thought that a major breakthrough had been made in the field of sexology. On 30 August of that year, the journal Science published a report by Salk Institute neuroscientist Simon LeVay titled 'A Difference in Hypothalamic Structure Between Heterosexual and Homosexual Men'. While the hypothalamus is rarely front-page news, this 'dis covery' was a media event. LeVay's study was reported almost immediately in several national newspapers (for example, Angiers, 1991a, 1991b; Maugh and Zamichow, 1991; Suplee, 1991; Winslow, 1991; Zamichow, 1991) and popular magazines (for example, Begley and Gelman, 1991; Crabb, 1991; Gorman,1991). The discovery, and LeVay himself, have enjoyed quite a celebrity status. The discovery was subsequently re-reported in both mainstream (for example, Grady, 1992; Kohn, 1992; Nimmons, 1994) and gay-oriented (for example, Dolce, 1993) publications in the early 1990s. LeVay also published three commercially successful books for the popular audience; two on the biology of sex difference and sexuality (1993, 1996), and one on lesbian and gay history (LeVay and Nonas, 1995) and has appeared on a number of popular American talk shows.
LeVay's work is only a part of a longer ongoing attempt to inscribe sexual desire within the discipline of biology by employing the concept of 'sexual orientation' (see Foucault, 1976n8; Irvine, 1990; Suppe, 1994; Weeks, 1985). Not surprisingly, LeVay's research has been taken up within scientific com munities. In addition to the original article in Science the prestigious journals Nature and Scientific American have seen fit to publish commentaries on LeVay's claims (LeVay and Hamer, 1994; Maddox, 1991). LeVay's finding has been cited to support accounts of the genetic basis of homosexuality (Hamer,1994) and of sex differences in cognitive style (Kimura, 1992).
Of particular relevance to feminist psychologists is the influence of the bio logical approach to sexuality on developmental psychology. In a recent special issue of Developmental Psychology devoted to sexual orientation and human development, several biologically-oriented articles used LeVay's finding to support their claims, although some regarded it more tentatively than others (Patterson, 1995). A recent universalistic attempt to account for the development of human sexuality also takes LeVay's result as unproblematic (Bern, 1996). Feminist psychologists have continually pointed to the political and epistemo logical problems of individualism and universalism in developmental psychology (for example, Burman, 1994; Walkerdine, 1984). These critiques apply also to the use of biological data to support universalistic accounts of the development of human sexuality.
Finally this work has had an impact in US law. The possibility that the difference in brain structure observed by LeVay constitutes a neurological determinant of sexual orientation, a claim which LeVay has alternately endorsed and refuted, was strongly endorsed by gay spokespersons in this first burst of media attention. It was argued that if homosexuality is immutable then gay and lesbian citizens constitute a suspect class under the Equal Protection Clause of US law. This legal status would render anti-gay discrimination illegal.
This initial public enthusiasm for the effects of biological research on civilrights may have been more the product of optimism than sound legal analysis, however. Halley (1994) has analyzed pro-gay legal arguments based on immutability and suggests that such claims -relating immutability to equal pro tection -may have been overblown. For the Equal Protection Clause to operate the relevance of the trait (for example, sexual orientation) to the purpose in hand must be demonstrated. Immutability is not a sufficient criterion to secure this protection.
Also, sexual minorities do not constitute an insular or distinct group. Rather,members of such minorities are often diffused throughout the population and anonymous. Gay men and lesbians, for example, are not identified as such at birth but rather are typically born into heterosexual contexts and differentiate their sexual identities in complex and different ways over the life span. In addition, many members of sexual minorities do not or cannot make their sexual identity known in contexts in which the Equal Protection Clause might operate (for example, the workplace) for a variety of practical and personal reasons. Consequently their status as immutable members of a discrete vulnerable group is legally questionable, regardless of the biological findings on the immutability of sexual orientation. In these regards, sexual minorities are very unlike the racial minorities whom the Equal Protection Clause does protect.
Finally, Halley has analyzed legal cases where the work of LeVay and otherbiologists have been cited by pro-gay plaintiffs as evidence of immutability. In such cases the immutability argument is usually rendered irrelevant or has complex and problematic consequences. In addition to failing when'given its day in court, the argument from immutability acts to divide the sexual minorities it attempts to protect. It typically negates or overlooks the existence of bisexual and queer identities, and often conflicts directly with pro-gay constructivist legal strategies. Halley suggests that immutability is neither necessarily pro-gay nor anti-gay, and that a pro-gay legal strategy needs to be based on common ground between essentialist and contructivist positions.
In short, LeVay's research has received notable attention in scientific, legal and public discourses. In this article, I critically account for this attention using performativity theory (Butler, 1990, 1993) claiming that it is by the implicit reference to, and restatement of, powerful sexist, heterosexist and imperialist norms that LeVay's work 'materializes' the gay brain. Performativity theory involves a critique of the account of the body in psychoanalytic discourse (Butler, 1990: 35-78, 1993), but has potential as a framework for reading other accounts of the body which pertain to psychology. Queer theory in general, including Butler's work, is opposed to essentialisms, arguing that a gay politics based on the primacy of an essential or unitary sexual identity will end up best representing the subject position 'twentieth-century, Western, white, gay male' to the exclusion of other gay subject positions (Duggan, 1995). Just as feminist psychologists argue that differences between the sexes are socially constructed in the dissemination and exchange of scientific information (for example, Hare Mustin and Marecek, 1990; Mednick, 1989; Unger, 1983), performativity theory helps us see the constructed nature of accounts of the gay male body in neuro science.
THE EMPIRICAL MATERIALITY OF THE GAY BRAIN
LeVay's work involves the posthumous dissection and comparison of the INAH nuclei in the hypothalami of 41 subjects; 19 homosexual men, 16 presumed heterosexual men, and 6 presumed heterosexual women. There are four INAH nuclei in the anterior hypothalamus; INAH 1, 2, 3 and 4. LeVay replicates previous findings that both INAH 2 and INAH 3 are sexually dimorphic, but he fails to replicate the finding that INAH 1 is sexually dimorphic. Finally, the INAH 3 of the homosexual men were smaller than those of the heterosexual men, and comparable in size to the INAH 3 of the women in LeVay's sample.
It could be argued that LeVay's work is popularly received because it is 'objective' or 'scientific'. By working within the recognizable paradigms of neuroscience, LeVay's research certainly brings with it a promise of certain answers to difficult questions about sexuality. However, scientific epistemological standards are characterized by debate as much as by consensus, and these standards are frequently in tension with each other and are not always resolved in rational ways (Knorr-Cetina and Mulkay, 1983; Woolgar, 1988). Biological studies of sexuality are no exception and LeVay's work has been extensively critiqued on empirical grounds (see Byne, 1994; Fausto-Sterling, 1992; Halley, 1994; Suppe,1994; and the essays in DeCecco and Parker, 1995). Although the title of LeVay's (1991) report boldly announces the discovery of 'A Difference in Hypothalamic Structure Between Heterosexual and Homosexual Men', this claim relies on highly problematic instantiations of each of the three central constructs; 'sexual orientation', 'hypothalamic structure' and 'difference'.
All of the 'homosexual' men, six of the 'heterosexual' men and one of the 'heterosexual' women studied had died of AIDS-related causes affording the categorization by sexual orientation of these subjects from the US Center for Disease Control (hereafter CDC) records. These records rely on doctors' reports such that this operational definition of 'sexual orientation' depends directly on whether a subject's closet had included the relationship with his or her doctor rather than on their sexual desires and practices. LeVay presumes that the other subjects are heterosexual 'based on the numerical preponderance of heterosexual men in the population' (LeVay, 1991: 1036). LeVay's justification of this assumption is problematic; he unhappily cites Kinsey et al. (1948) who reported that many American males have sex with men while identifying as heterosexual. The classification of the 'homosexual' subjects is also problematized by the existence of one bisexual-identified man in the sample who was included in this group. Discrete categorization of these subjects into 'homosexual' and'heterosexual' groups does not follow from their sexual practices, desires and experiences but rather is required by LeVay's analysis of variance paradigm. LeVay does acknowledge the possibility of misassignment, but in a highly circular way and only so far as it supports his claim to have found a natural difference in hypothalamic structure (see Fausto-Sterling, 1992: 252).
Just as LeVay's instantiation of the concept 'sexual orientation' is ambiguous, so too is his materialization of 'hypothalamic structure'. Although the 'size' of the INAH 3 nuclei has afforded a lot of humor in the media (for example, Dolce, 1993) LeVay's research uses only one of several possible instantiations of the concept of nucleus 'size' which might just as validly be considered a measure of cell density or distribution (see Suppe, 1994; Barinaga, 1991). LeVay's instantiation of 'size' is all the more problematic because the 'difference' he claims to have discovered is based on a particular interpretation of the concept of statistically significant difference.
LeVay examined four separate nuclei (INAH 1 through 4) and found a statistically significant difference by sexual orientation in only one case; INAH 3. LeVay correctly performed two analyses on his measurements of INAH 3 size. In the first instance he compared the INAH 3 sizes of all of the homosexual men with all of the heterosexual men and this difference was statistically significant (p = .0014). The second analysis compared the INAH 3 size of the 19 homosexual men, all of whom had died of AIDS-related causes, with the INAH 3 size of the six heterosexual men who had died of AIDS-related causes. Under this second analysis the statistical significance of the result dropped considerably (p = .028).
LeVayhad gathered measurements from four nuclei - all of which are possible neurological correlates of sexual orientation - when these analyses were performed. However this fact is not reflected in his statistical assumptions. Rather he analyzed differences in INAH 3 size as if this were the only possible locus of a 'difference'. It would have been more prudent to set a significance level in advance, such that the probability of concluding that a 'difference' exists by chance alone, in any of the four nuclei examined would be no greater than the specified significance level. The Kimball inequality is a statistical procedure that allows this. It determines the significance level for each individual test of a null hypothesis such that the 'group error rate', or possibility of falsely rejecting any of the null hypotheses among a set of related tests, is equal to the specified significance level (see Ott, 1988). Applying the Kimball inequality to LeVay's work, such that the group error rate is 0.05 would entail setting a significance level of .013 for each of the four individual tests. LeVay's 'difference' by sexual orientation in INAH 3 size, among persons who died from AIDS-related causes, is clearly statistically insignificant under this more stringent analysis (p = .028 > .013).
LeVay's work is not above critique on an empirical basis. The central concepts of this research (sexual orientation, hypothalamic structure and difference) can be instantiated in multiple ways, leading to different conclusions regarding the supposed neurological substrate of sexual orientation. Therefore, the response to LeVay's work cannot be attributed to any necessarily rational recognition by scientists and the public of irrefutable scientific work. Rather, the publication of this research by Science, without more stringent methodological requirements, suggests that some desire to materialize a difference between homosexual and heterosexual men is relevant not just to the popular reactions to this report but may also be constitutive of the report itself.
While empirical critique is important, the popular success of this work
suggests urgent political and psychological questions for feminism about the ways that heterosexism and sexism work through scientific accounts and become reified as scientific objectivity. By this logic, a feminist psychological analysis does not stop with casting an empirical judgment on a single study as 'good' or 'bad' science but also examines the ways in which that which is taken as 'good science' abjects women and sexual minorities. Such an analysis might appear, at first sight, counter to a psychological approach. Psychology has historically relied on endorsing rather than problematizing empirical inquiry and quantitative methods. However, psychology is increasingly being performed in ways that go beyond a positivist empiricist framework. In addition to the feminist contribu tions already mentioned discourse analysis (for example, Burman and Parker, 1993; Potter and Wetherell, 1987; Wetherell and Potter, 1992), rhetorical psychology (for example, Billig, 1987, 1991), and social constructionism (for example, Gergen, 1982, 1985; Henriques et al., 1984; Kitzinger, 1987; Sampson, 1983) are productive non-positivist directions within psychology.
A PERFORMATIVE THEORY OF THE HYPOTHALAMUS
Butler (1993) has argued that feminist and queer scholarship interrogate the ways in which the body is 'materialized' in discourse. She asks us to consider how accounts of the body come to 'matter', both in the sense of having import and of appearing as material. These two meanings of 'matter' are related, for the physical body is typically considered to be more important than any of its socially constructed meanings. Her project then goes beyond feminist accounts that attempt a distinction between an underlying, physical, solid, constant 'sex' and a constructed, fluid, variable 'gender'. Instead the drawing of such a line between sex and gender is understood as a discursive event; a speech act which determines what is taken as 'material' and what is taken as 'constructed'. The material body cannot be a pre-existent object that is later 'socially constructed'. Rather, 'the body' is also a social construction, an effect of discourse.
To 'concede' the undeniability of 'sex' or its 'materiality' is always to concede some version of 'sex,' some form of 'materiality.' Is the discourse through which that concession occurs - and, yes, that concession invariably does occur not itself formative of the very phenomenon that it concedes (Butler, 1993:10)?
The body is materialized through performativity, according to Butler. Austin (1962) defines a performative speech act as one that has the power to call into being or to enact that which it names. In Austin's account this performative power comes from the author's will. Butler contests this. Borrowing from Derrida (1988), she claims that performative speech acts have the power to call into being because they are recognizable as citations of norms. Performativity is then always derivative, a result of abiding by, and appealing to discursive laws, laws which are themselves reproduced in these citational practices.
For example, Austin's paradigmatic case of a performative speech act is the wedding ceremony. Here the minister calls into being the relation of marriage as he names it with the speech act 'I pronounce you man and wife'. However, wedding as a speech act can only be performed by citing the normativizing force ofheterosexualization. The power of 'I pronounce you man and wife' comes not from the will of the minister, but rather from the recognition by the community of the marriage of two opposite sexed persons. This reiteration is also an exten sion of that norm, marking the new couple as definitively heterosexual.
Austin's account relies on an author's will to explain performativity, but for Butler the author of a performative speech act is also an effect of discourse. Attributing performativity to a pre-discursive author often occurs; obscuring the derivative nature of performativity, such that performativity works through a process of 'dissimulated citationality' or unconscious recognition. Weddings do not proceed as a result of calling attention to the norm that only heterosexual unions can be authorized. This recognition is necessary, but this necessity is not acknowledged. It is covered under the illusion of authorship; the insistence that there is a pre-discursive subject, such as an authorized minister, who is doing the pronouncing.
Although Butler's analysis of the performativity of 'sex' has been most relevant to psychology as it comments on psychoanalysis, I would like to extend the approach to other discourses through which the body is materialized including neuroscience. Neuroscience calls into being the body -and the psychology - that it names. In LeVay's writing, it is the body and the sexual desires of the male homosexual that are being performed. As I have shown above, the performativity of LeVay's results are not due to abiding by empirical norms, although LeVay's status as a scientist has often been successfully cited to create such an illusion. The materialization that is performed in LeVay's writings (specifically LeVay, 1991, 1993) draws on and reiterates a set of norms pertain ing to 'sex' and its performance within scientific discourse. LeVay's work assumes and reinforces the norm that there are two types of person, homosexuals and heterosexuals, and that they can be considered biologically distinct. This norm is selectively applied to those who are considered male, such that women are absent in this account of the body. The male homosexual is materialized as a kind of woman, which then demands being constituted as a lack or failure. Finally, LeVay's work reiterates the claim that AIDS is a disease exclusive to gay men. In achieving this equation, the male homosexual is constituted by a reduc tion to an unconsciously promiscuous sexual nature.