Peter Hegarty & Sean Massey. Journal of Homosexuality. Volume 52, Issue 1-2. 2006.
This volume concerns tensions between queer theory and lesbian, gay, bisexual, and transgender studies, and the present article explores this tension with regard to social psychological studies of anti-homosexual prejudice. The early 1970s paradigm-shift away from disease models towards stigmatization models of homosexuality in psychology produced many attempts to measure individual differences in anti-homosexual prejudice among heterosexual-identified people (many of which were published in Journal of Homosexuality). This research drew conceptually on lesbian and gay liberationist thought and methodologically on the positivist tradition of attitude measurement in psychology. Since the 1970s, the study of anti-homosexual attitudes has become one of the most prolific areas of lesbian and gay research within psychology. However, studies of anti-bisexual and anti-transgender prejudice remain far rarer (although, see Spalding & Peplau, 1997; Tee, 2003). Both of us research anti-homosexual attitudes (see Hegarty, 2002; Hegarty & Pratto, 2001; Massey, 2004) and this paper aims to open the question of how the writings that have become known as queer theory lead to a reinterpretation of this area of social psychological research.
Anti-Homosexual Attitudes and Social Constructionism
Although attitudes have long been dignified as a central concern of social psychology (Allport, 1935), interest in attitudes has waxed and waned over the decades (Eagly, 1992). An attitude is typically defined as an internally located value judgment; as “a tendency or state that is internal to the person” (Eagly, 1992, p. 694) that “is expressed by evaluating a particular entity with some degree of favor or disfavor” (Eagly & Chaiken, 1993, p. 1). Psychologists commonly measure individuals’ attitudes by eliciting their level of agreement or disagreement with various opinion statements, and psychologists increasingly measure implicit attitudes via participants’ speed of reaction to attitude relevant stimuli (Greenwald et al., 2002).
Social constructionist scholars have repeatedly critiqued the essentialist assumption that internal psychological constructs such as “attitudes” can be known by available scientific methods (see Potter and Wetherell, 1987 on explicit attitudes and Steele and Morawski, 2002 on implicit attitudes). Within social constructionism, Kitzinger’s (1987, 1989) radical feminist social constructionist analysis of measures of anti-homosexual prejudice is most germane to the current discussion. By attending to the content of the questionnaire items selected to measure anti-homosexual prejudice, Kitzinger showed how psychologists had conflated pro-lesbian/gay attitudes with liberal humanist ideologies. For example, opinions that minimized differences between lesbians and straight women were defined as pro-lesbian attitudes. Kitzinger noted that this framework neglected political differences among lesbians and might even paradoxically cast radical lesbians as homophobic by virtue of their disagreement with liberal humanist assumptions.
Like Kitzinger’s analyses, this paper unpacks the implicit sexual politics of anti-homosexual attitudes research. However, unlike Kitzinger (1987) we do not consider this research to be a misplaced rhetorical exercise devoid of epistemic value. Moreover, while Kitzinger forwarded a radical feminist alternative to liberal humanism, our suggestions are developed from the problems raised by queer theory. For Kitzinger, sexual politics remains a question of inter-group conflict between differentially powerful, clearly defined groups. For example, she writes:
Central to radical feminism is the belief that the patriarchy (not capitalism or sex roles or socialization or individual sexist men) is the root of all forms of oppression; that all men benefit from and maintain it and are, therefore, our political enemies. (Kitzinger, 1987, p. 64)
In contrast, we seek to trouble the assumption that sexual politics operates exclusively, or even predominantly, through intergroup power contests. Rather we start with the observation that social psychologists have studied attitudes on just about everything: individuals, social groups, the self, objects, activities, policies, and ideologies. To what, then, are the anti-homosexual attitudes which social psychologists have measured opposed? Is the homosexuality in question a minority group, a form of sexual practice, an identity performance, or a political movement? One of the central points of this paper is that in spite of over thirty years of published research on this topic, we do not yet have the data to answer this seemingly basic question.
Our analysis is framed by Sedgwick’s (1990) claim that modern epistemologies are founded on problematic definitions of the homo/heterosexual binary. Specifically Sedgwick (1990, p. 1) juxtaposed the assumptions that sexual definition is either “an issue of active importance primarily for a small, distinct, relatively fixed homosexual minority” (the minoritizing view) or “an issue of continuing, determinative importance in the lives of people across the spectrum of sexualities” (the universalizing view). Sedgwick proposed this binary to render essentialist/ constructivist debates about sexuality redundant, privileged neither assumption over the other, and called for their use to further anti-homophobic inquiry rather than to resolve epistemological tensions between them.
In developing this analysis, we have leaned closer to the universalizing view. This is because current social psychological approaches to anti-homosexual attitudes are framed within an implicit minoritizing approach that is heavily focused on sexual identity rather than sexual practice, and which understands anti-homosexual prejudice as an exemplar of minority group oppression (see Kitzinger & Coyle, 2002; Herek, 2000; Sanford, 2000). Below we trace how and when attitude measurement techniques began to assume the minoritizing position, and show how the results of experiments on behavioral effects of attitudes can be re-interpreted from within the universalizing view. Although we look at attitude measurement in close detail in this paper, our broader aim is to stimulate a conversation between social psychology and queer theory. In the conclusion of the paper we suggest some reasons for the absence of such a conversation, and offer some possible starting points for discussion.
Measuring Anti-Homosexual Prejudice I: Gay Liberation and Universalizing Models
One of the greatest successes of the gay and lesbian liberationist movements was the removal of homosexuality from its minoritizing psychiatric definition as a mental illness in 1973 (Bayer, 1981). Since World War II, American mental health professionals had increasingly defined homosexuality in psychoanalytic terms, while simultaneously overwriting Freud’s (1923) universalizing-albeit heteronormative (Butler, 1990)-theory of primary bisexuality (Abelove, 1993; Lewes, 1989; Ryan & O Connor, 1998). This paradigm shift bore an uncanny resemblance to the earlier shift away from Eugenic models of race toward the study of anti-Black prejudice in post World War II American psychology (Richards, 1997; Samelson, 1978). The depathologizing of homosexuality occasioned new interest in the empirical study of anti-homosexual prejudice. Liberationist writings relied on both universalizing and minoritizing approaches to sexuality. Calls for rights for lesbians and gay men were central, but heterosexuality itself was also understood as a limited erotic choice and one that was socially enforced (see Altman, 1972; Hocquenghem, 1972, 1978; Radicalesbians, 1969; Rich, 1980; Whitman, 1969-70, 1997; and, the later writings of Ellis, 1976).
The term homophobia is most commonly attributed to clinical psychologist George Weinberg (Oxford English Dictionary), who described it as an irrational fear or “dread of being in close quarters with homosexuals” (Weinberg, 1972, p. 4). However, Weinberg’s work was deeply indebted to sexual liberationism. Weinberg acknowledged the influence of many gay and lesbian liberationists including Arthur Evans of the Gay Activist Alliance of New York, Randy Wicker, the founder of the Homosexual League of New York, Kay Tobin and Franklin Kameny, (Weinberg, 1972 and Ayyar, 2002). Chapters of Weinberg’s book Society and the Healthy Homosexual were first published in Gay, a magazine edited by gay liberation pioneers Jack Nichols and Lige Clarke. Weinberg also served as a spokesperson for some of the homophile organizations during their challenge to the APA’s inclusion of homosexuality as a diagnostic category in the DSM (Lewes, 1988).
Weinberg’s concept aimed to shift attention away from questions of homosexuals’ mental (ill)health towards questioning the mental health of homophobes. Weinberg (1972) argued that he “would never consider a patient healthy unless he had overcome his prejudice of homosexuality” (p. 1). In a universalizing move, he suggested that homophobia not only limited self-identified heterosexuals’ social relationships, it also inhibited their erotic potentiality-a point also made, but somewhat ambivalently, by psychologist and sexual liberationist Albert Ellis (1976). Articulating a “new sexual culture,” Weinberg (1972) wrote, “Gay liberation implies freedom from having to align oneself in sexual preference with dictates from anywhere” (p. 127). Positioning the sexual outsider as exemplary, he later stated that, “all love is conspiratorial and deviant and magical” (Ayyar, 2002). Other theorists, such as psychoanalyst Wainwright Churchill (1967) offered similar conceptualizations of anti-homosexual prejudice, locating its origins in erotophobic society.
Many of the social psychologists who developed technologies for measuring anti-homosexual attitudes favorably cited Churchill’s and Weinberg’s theories (e.g., McDonald & Moore, 1978; Larsen, Reed & Hoffman, 1980; Hudson & Ricketts, 1980). Even Smith’s (1971) Homophobia Scale, which was published prior to Society and the Healthy Homosexual and before the depathologizing of homosexuality, was informed by Weinberg’s writings (see Weinberg, 1972, p. 133; personal communication, June 25, 2003). Attempts to measure anti-homosexual prejudice explicitly located the source of discomfort around homosexuality in societal prejudice rather than in homosexuals themselves. Thus, Smith (1971) called for the study of the “social milieu” (p. 1089) in which the homosexual person lives and MacDonald, Huggins Young, and Swanson (1973, p. 10) argued for movement from an “organism deficiency” to a “social deficiency” explanation for discrimination. Such arguments occasionally drew on minoritizing frameworks by making analogies between anti-homosexual prejudice and anti-Black racism. For example, MacDonald and Games (1974) suggested that approaches to psychological inquiry that were “almost exclusively restricted to blacks [sic] and the poor … should be extended to include other discriminated against groups” (p. 10). In other words, anti-racist work provided an available model for understanding homophobia as a form of intergroup prejudice.
The “society” which was understood to be the origin of homophobia was to be assessed via correlations between measures of anti-homosexual prejudice and the endorsement of relevant ideologies. Weinberg (1972 )had theorized that homophobia was derived from such motives as religion, fear of being homosexual, repressed envy, a threat to values, and fear of death. Attitude researchers correspondingly gathered data showing correlational relationships that were germane to these claims. Smith (1971) found evidence of correlations between anti-gay prejudice and the need for conventionality, conformity, and the need to maintain traditional sex and gender roles. Dunbar, Brown and Amoroso (1973) found that anti-homosexual participants were significantly more conservative and negative regarding sexual practices in general, expressed significantly more sex guilt, and were significantly more rigid in terms of their view of “appropriate” sex-roles. Yet, even as anti-homosexual prejudice was becoming understood as partially comprised of sex-role rigidity, feminists were transforming social psychologists’ definitions of sex roles. Until the 1970s psychologists had assumed masculinity and femininity to be logical opposites, and often conflating masculinity-femininity with psychological differences between straight and gay men (Constantinople, 1973; Hegarty, 2003a; Lewin, 1982a, b). Feminist researchers re-theorized masculinity and femininity as constraining roles, and posited psychologically healthy androgynous individuals who could flexibly adopt either role as the situation dictated (e.g., Bem, 1974; Bem & Lenney, 1978). Research on androgyny and on antihomosexual attitudes similarly presumed that individuals who could transgress overly rigid sexual and gender roles were psychologically healthy. Several researchers worked to correlate new measures of gender roles and new measures of anti-homosexual prejudice (e.g., Minnigerode, 1976; Storms, 1978).
Societal homophobia, in line with liberationists’ universalizing notions and psychodynamic theory, was understood to shape individuals’ emotions. Both Weinberg’s and Churchill’s works were frequently cited as the theoretical justifications for these hypotheses (e.g., Smith, 1971, p. 1091; Dunbar, Brown & Amoroso, 1973, pp. 271-272; Hudson & Ricketts, 1980, p. 357). Attitude researchers predicted affective responses within the parlance of ego defense. Anti-homosexual prejudice was thought to derive from personal anxiety (Millham, San Miguel, & Kellogg, 1976), sex guilt (Dunbar, Brown, & Amoroso, 1973), and fear and denial of personal homosexual tendencies (Mosher & O’Grady, 1979).
Sociological theories that cast sexual orientations as learned or socially constructed provided a further theoretical basis for universalizing approaches to the measurement of anti-homosexual prejudice. Citing constructionist theories which located male sex socialization in the context of sex guilt (i.e., Gagnon & Simon 1973; Simon & Gagnon, 1969), Mosher and O’Grady’s (1979) measure of homosexual threat included heterosexuals’ relationship to their own erotic responses in their empirical measure of anti-homosexual prejudice. Fear and denial of homoerotic tendencies became a component of their Homosexual Threat Inventory, originally conceptualized as agreement or disagreement with items referring to same-sex eroticism and desire, such as “to love another man is to know the heights of the human soul” and “I could never bring myself to suck another man’s cock.”
In short, there was heterogeneity of measures of anti-homosexual prejudice in the 1970s variously indebted to psychoanalysis, liberationist thinking, social constructionism, second wave feminism and the civil rights movement. The new scales typically included items that measured attitudes towards lesbians or gay men as a distinct minority group. For example, Smith’s (1971) H-Scale included the item “Homosexuals should be locked up to protect society” and the item “A homosexual could be a good president of the United States” (p. 1094). Millham, San Miguel, and Kellogg’s Homosexuality Attitude Scale included the item “Most male homosexuals dislike women” (p. 5), and Mosher and O’Grady’s Homosexual Threat Inventory included the item “Homosexuals should stay in their own gay bars and not flaunt their deviance” (p. 864). However, universalizing ideas about universal bisexual potentiality were also commonly cited. For example, Smith’s scale included the item “If laws against homosexuality were eliminated, the proportion of homosexuals in the population would probably remain the same” (suggesting, when reversed, the need for social control of a universal homosexual potential). Millham, San Miguel and Kellogg’s scale included the item “Male homosexuality is a choice of lifestyles.” Finally, Mosher and O’Grady’s scale included the item “I am frightened that I might have homosexual tendencies.” Indeed, to the extent that it drew on such universalizing themes, the work of the 1970s represented measurement of biphobic attitudes as much as the measurement of homophobic attitudes.
Measuring Anti-Homosexual Prejudice II: Social Reform and Minoritizing Models
Towards the end of the 1970s, gay and lesbian politics became organized less around the goal of sexual liberation and more around the assimilationist goal of achieving civil rights (see D’Emilio, 1983; Kitzinger, 1987; Weeks, 1985). The utopian call for sexual revolution was supplanted by the more moderate (and arguably more achievable) goal of social reform. Reactionary figures such as Anita Bryant who cast gay men and lesbians as a threat to the American family played a lethally important part in defining the terms of this cultural conflict. Radical theories of sexuality, perceived by some as fulfilling stereotypes of the promiscuous, predatory and anti-family homosexual (and perhaps aggravating unreconciled sex-guilt within gay people themselves) became more difficult to articulate. The increasing engagement of mainstream institutions with lesbian and gay right agendas occasioned new sexuality hierarchies within lesbian and gay subcultures. For example, in 1980 the National Organization of Women (NOW) codified its support for lesbian rights while signaling its condemnation of pederasty, sadomasochism, and pornography (National Organization of Women, 1980). As Rubin (1984) put it, several seemingly liberal institutions such as NOW still attempted to draw an “imaginary line between good and bad sex.” Such sexual politics were motivated as much by fear of the unknown as by concern with liberation or fairness, as such a line was imagined to:
… stand between sexual order and chaos. It expresses the fear that if anything is permitted to cross this erotic DMZ, the barrier against scary sex will crumble and something unspeakable will skitter across. (p. 14)
Psychologists’ definitions of anti-homosexual prejudice also became increasingly focused on minoritizing concerns during the early 1980s (e.g., Herek, 1984, Kite & Deaux, 1986; Larsen, Reed, & Hoffman, 1980). Earlier scales were vulnerable to several criticisms; those who scored high on homophobia scales did not manifest the typical physiological reactions that accompany a clinical “phobia” (see Millham & Weinberger, 1977; Shields & Harriman, 1984). Also, the predicted link between anti-homosexual prejudice and gender roles failed to materialize (see Whitley, 2001). Finally, the earlier scales had also been implicitly androcentric; by asking questions about homosexuals they probably elicited attitudes towards male homosexuality only (see Black & Stevenson, 1984). Thus the development of new scales for attitude measurement was easily positioned as a scientific advance rather than as a shift in the implicit politics that informed the concept of anti-homosexual prejudice.
However, consider Herek’s (1984) Attitudes Toward Lesbians and Gay Men (ATLG) scale, which is now a standard in the field. Herek defined heterosexism in explicitly minoritizing terms as “a term analogous to sexism and racism, describing an ideological system that denies, denigrates, and stigmatizes any non-heterosexual form of behavior, identity relationship, or community” (Herek, 1990). Several of the earlier scales that drew on universalizing views were multi-factorial, and measured several distinct components of anti-homosexual prejudice allowing liberal notions of tolerance and civil rights to coexist with ego-defensive aspects of anti-homosexual prejudice. Herek (1984) questioned the necessity of multifactorial scales, and ultimately concluded that only a single factor was required to conceptualize heterosexism.
Herek (1984) reported four studies which drew on both earlier attitude measures (Levitt & Klassen, 1974; MacDonald et al., 1973; Millham et al., 1976; Smith, 1971) and some original items. Notably, Herek excluded items from Mosher and O’Grady (1979) arguing that the “locker room” language of these behavioral items would offend participants instead of tapping into feelings of personal threat. This empirically unsupported supposition was ironically made during the early days of the HIV/AIDS crisis when anti-homosexual prejudice operated most forcefully through heterosexuals’ feelings of personal threat. The omission of Mosher and O’Grady’s items demonstrates how social psychologists technologies are shaped as much by implicit sexual politics as much as the empirical results they gather.
Exploratory factor analysis revealed a variety of factors, including: condemnation/repression, personal revulsion/threat and the desire to avoid contact, desire to keep away from children, beliefs about homosexuals, denial of similarities between heterosexuals and homosexuals, and comparing of heterosexual and homosexual relationships. In each study, however, the condemnation/tolerance factor explained a much larger portion of the variance among items than the other factors, allowing Herek to argue that a unidimensional model was most “appropriate” and that all other factors were of trivial importance. Herek’s resulting Attitudes Toward Lesbians and Gay Men (ATLG) scale examines attitudes along a single continuum from tolerance to condemnation via items which assess emotional reactions to lesbians and gay men and support for civil rights issues. Although Herek’s definition refers to stigmatization of non-heterosexual behavior, questions concerning respondents own construction of their same-sex desires or behaviors are absent from the ATLG.
As Sedgwick notes, however, minoritizing and universalizing views of sexuality are never completely stable or separate in modern epistemologies. While developing the ATLG, Herek (1987) also began to assess the functions of anti-homosexual attitudes (and later, in articles related to AIDS, Herek, 2000). The study of attitude functions is indebted to psychoanalytic theorizing (see Katz, 1960), and had been critiqued on the grounds that attitude functions could not be empirically detected or measured. Herek’s (1987) work argued that heterosexism served a variety of functions for different individuals including the expression of core values, reflecting cognitive beliefs, and defending against threats to the ego. In this last regard, Herek has explicitly argued that for a minority of heterosexual-identified persons, heterosexism is a result of displaced fear of their own homoerotic potential. Thus, Herek has operationalized universalizing ideas in his attitudes work, even though they have been excluded from his operational definition of anti-homosexual prejudice.
Experimenting with Identity: Linking Attitudes to Behaviors
Implicit in the study of attitudes is the promise that these internal constructs will in some way affect social behavior. However, the attitude-behavior link has often been shown to be absent (LaPiere, 1934) and during the 1970s social psychological research increasingly examined factors that might affect the strength of this link (Azjen & Fishbein, 1980). Thus, demonstrations of behavioral effects of anti-homosexual attitudes also became a topic of research, and such demonstrations were typically made with experimental studies. As in most other areas of social psychology, the experimental participants were typically college undergraduates, (Sears, 1986). These participants made a judgment about, or interacted with, a target individual, whose perceived sexual orientation was experimentally manipulated. Such experimental studies have varied in their realism. Some have involved the presentation of written information about individuals identified as gay, lesbian, or straight (see Laner & Laner, 1979, 1980; Sigelman, Howell, Cornell, Cutright, & Dewey, 1990; Snyder & Uranowitz, 1978; Storms, Stivers, Lambers, & Hill, 1981; Weissbach & Zagon, 1975). In others, participants encountered flesh-and-blood individuals who enacted particular sexual orientations (see Clark & Maass, 1988; Kite, 1992; Kite & Deaux, 1986, Study 2; Kruwelitz & Nash, 1980; San Miguel & Milham, 1976). In both cases, differences between the actions of participants with high and low levels of heterosexism to lesbian/gay and straight targets have been used to ground inferences about the behavioral consequences of anti-homosexual attitudes. For example, Kite and Deaux (1986) report that intolerant heterosexual males and tolerant heterosexual males differed in their liking of gay male targets, the information that they elicited from him, the information they presented about themselves and their memories of the interaction.
While these experiments have presumed to assess behavioral reactions to lesbians and gay men as a distinct minority group, they owe an unacknowledged debt to universalizing theories of homo/heterosexual definition. The experiments require that target individuals manipulate the presentation of their sexual identities, and typically, the same individuals have played the roles of both lesbian (or gay) target and of straight target in these experiments. Thus, targets have “done” straightness by direct disclosure (Clark & Maass, 1988), mention of an intention to marry (Kruwelitz & Nash, 1980), or, most frequently, by saying nothing about sexuality at all and allowing assumptions of compulsory heterosexuality (Rich, 1980) to go untroubled (see Cuenot & Fugita, 1982; Kite, 1992; Kite & Deaux, 1986; Gurwitz & Markus; San Miguel & Milham). Lesbian and gay identities have been “done” in the laboratory through direct verbal (Clark & Maass, 1988) or written (Kruwelitz & Nash, 1980) disclosure, expressions of interest in ongoing experiments about homosexuality (San Miguel & Millham, 1976), suggestions of homosexuality as a topic for discussion (Kite, 1992; Kite & Deaux, 1986), buttons that says “gay and proud” (Cuenot & Fugita, 1982), and mention of involvement in gay student organizations (Gurwitz & Marcus, 1978).
As Sedgwick (1990, p. 4) notes the wearing of a gay-positive t-shirt not only reports an identity, it also constitutes it. Participants’ differential reactions to the targets in these experiments have been theorized as discriminatory reactions to lesbian/gay and straight individuals. However, the actions through which gay/lesbian identities are enacted are scripted to be highly volitional, and people who do not engage in such actions are not necessarily always straight. These experiments may just as readily be understood as assessing differential responses to out and passing lesbian/gay individuals, as to straight and (out) lesbian/gay individuals. In other words, the experiments may be assessing differential reactions to ways of enacting minority sexual identities, rather than differential reactions to members of separate discrete social groups.
Following Sedgwick and Butler (1990,1991; Osborne & Segal, 1994) we might understand these targets identity enactments as performative of identity. If the targets’ methods of performing identity are understood as constituting identity in different ways rather than simply reporting the same underlying identity, then it is less clear that these experiments are all examining the same social psychological processes. Explicit declaration of one’s homosexuality, mention of involvement in a gay student group, and the wearing of a gay-positive button are not equivalent speech acts, and each accomplishes something more than the revelation of a presumed underlying identity (as readers who have negotiated disclosure dynamics will probably recognize). Butler argues against presumed unity among those represented by political signifiers (i.e., women, lesbians and gay men) and argues for a mode of politics that intervenes in the moments when such identities are thrown into question. This leads us to think differently about the empirical claims that are made from experimental results, and the need to theorize and examine social psychological experimental practice. On the first point, future experiments that acknowledged the performativity of identity could begin to examine if different performances of sexual identities moderates the relationship between research participants’ anti-homosexual attitudes and their anti-homosexual behavior. On the second, the doing of experiments represents a particular kind of constructivist work; ironically, the positivist space of the psychological laboratory is perhaps the location where the postmodern fantasy that identities can be put on and taken off as easily as clothing has some legitimacy. An ethnographic study of psychological laboratories could greatly inform the politics of universalizing theories by illuminating a concrete set of practices where social identities are designed to be highly mutable (Hegarty, 2001).
Queer Theory and Attitudes Research: Working the Tensions
Compared to the early 1970s heterogeneous alliances between liberationist, psychoanalytic, and social constructionist thought and positivist technologies of attitude measurement, the intellectual work now known as queer theory (e.g., Abelove, Barale, & Halperin, 1993; Butler, 1990; DeLauretis, 1991; Fuss, 1991; Sedgwick, 1990; Warner, 1993) has, for better or worse, been considerably less engaged with academic psychology. Contrast, for example, the plethora of empirical articles on anti-homosexual prejudice published in the early volumes of Journal of Homosexuality with the almost complete lack of quantitative social science research in the more recent GLQ: A Journal of Gay and Lesbian Studies. Clearly, queer theory emerged in a way that was less bound up with the quantitative social sciences than did its liberationist forerunner. Below we examine why this might be so, and theorize where productive intersections might emerge in future work.
Of course, queer theory differs radically from liberationist thought in its conception of the individual subject. Liberationist work proposed that societal repression of individualized sexual natures was the primary mode of sexual politics. Foucault’s History of Sexuality: Volume 1 (1978), arguably the most influential book within queer theory, theorized power as operating through the production of sexuality and sexual categories as much as by their repression. Foucault called for heightened critique of ostensibly liberating breaks with the repressive past and seemingly rational advancements in the human sciences. As in much of Foucault’s writing, his theory of sexuality describes the sciences of the human subject as originating in political attempts to control and order human bodies (see also Foucault, 1977). As such Foucault’s work provides grounds for extreme skepticism about any project focused on individual differences or the measurement of mental states such as attitudes research.
Second, when queer theorists did turn to a theory of the individual subject, they tended to alight on psychoanalysis. Warner (1993, p. xi), for one, described psychoanalysis as “the most rigorous and sophisticated language about sexuality.” As Kitzinger (1998) notes, imperatives to think the psychological subject as a psychoanalytic subject within sexuality and gender studies substantially limit engagement with academic psychology. Relationships between quantitative psychologists and psychoanalysts have never been easy in American psychology (Hornstein, 1992), and by the late 1980s, cognitive psychology, and not psychoanalysis, was the dominant truth regime in psychology (Friman, Allen, Kirwin, & Larzelere, 1992). By comparison with queer engagements with psychoanalytic assumptions, critiques of the hetero- normativity of cognitive theories of the subject were slow to emerge (although see Curtain, 1997 and Helmreich, 1998 for notable exceptions). Thus queer theory became interested mostly in a theory of the subject which was peripheral to the interests of most social psychologists by the early 1990s.
Finally, lesbian and gay psychology has been a more cautious disciplinary project than queer theory. North American lesbian and gay psychologists positioned their work as empirical disciplinary work, often leading to a dismissal of the counter-disciplinary projects ongoing in lesbian, gay and queer studies. Most obviously, psychologist John Gonsierek (1993, viii-ix) in the foreword to a historical volume described lesbian and gay studies as an “inward looking,” “self-absorbed” domain where “dogma is substituted for critical thinking.” Lesbian and gay studies, for Gonsierek, was both “intellectually rigid and irrelevant both to the lives of gay and lesbian citizens and to honest intellectual inquiry.” This caution has also limited the degree to which lesbian and gay psychologists discuss sex. Lesbian and gay psychologists, however, have typically been less willing to challenge the stigmatization of gay and lesbian sex than have their colleagues in the humanities. Sex research has always been stigmatized, and an explicit affirmative study of gay and lesbian sexualities has been read as constituting an infraction against implicit norms for scientific objectivity (Hegarty, 2003b; Irvine, 1990). While many queer theorists focused their approach on sex, rather than sexual identity (Sedgwick, 1990; Warner, 1993), lesbian and gay psychology has tended to privilege identity over behavior as the object of inquiry (Kitzinger & Coyle, 2002).
In spite of these differences, we agree with the editors of this volume that useful theory can be made from tensions between lesbian, gay, bisexual, and transgender studies and queer theory. Unlike Kitzinger (1987) we are not content to dismiss social psychology as purely rhetorical practice. Rather, we believe there are clear grounds for re-thinking the empirical record on anti-homosexual prejudice in light of the conceptual developments made by queer theorists.
Consider first the politics of knowledge production. Within social psychology, researchers have historically understood themselves to be accumulating empirical facts and developing an ever-growing body of knowledge about social behavior (Farr, 1996). However, even as gay liberationist ideas were being translated into attitude scales, Gergen (1973) suggested that both the findings of social psychological studies and the methods used to achieve those findings were bound by history. As Kitzinger and Coyle (2002) note, contemporary lesbian and gay psychologists have tended to either adopt traditional quantitative methods (Herek, 1998) or to embrace qualitative methods that eschew the reality of internal psychological processes (Gough, 2002; Speers & Potter, 2000). Queer theory might suggest how social psychologists could have their attitude technologies and deconstruct them, too. In the context of HIV/AIDS from which queer theory largely emerged, it became necessary to critically read the biomedical discourse through which “facts” about AIDS were being produced and to develop strategies for living with the virus. As Paula Treichler noted, this involves a kind of double consciousness:
Of course, where AIDS is concerned, science can usefully perform its interpretive part: we can learn to live-indeed, must learn to live-as though there are such things as viruses. The virus-a constructed scientific object-is also a historical subject, a “human immunodeficiency virus,” a real source of illness and death that can be passed from one person to another under certain conditions that we can apparently-individually and collectively-influence. The trick is to learn to live with this disjunction, but the lesson is imperative. (1991, p. 69)
We would like to encourage the learning of such a trick in regard to the psychological category of anti-homosexual prejudice. We need to theorize by acknowledging the interpretive work that attitude scales can perform with regard to social prejudices, while also recognizing their historicity, contingency, and limitations. In concert with Sedgwick’s (1990) approach this may involve nurturing anti-homophobic inquiry within mutually incompatible epistemologies and a suspension of the Platonic desire to immediately resolve their inconsistencies.
One benefit of learning the trick to which Treichler (1991) refers would be the ability to see earlier modes of knowledge production as different from, but not necessarily inferior to, current paradigms. We might revisit the liberationist work of the 1970s as grounds for re-theorizing the complexity of heterosexism. Ironically, even work within the minoritizing perspective consistently shows relationships between antihomosexual prejudice and authoritarianism, traditional gender role beliefs, and affective reactions (Pettigrew & Tropp, 2000; Whitley, 1999). Herek’s shift away from homophobia to terms such as sexual prejudice and heterosexism may have obscured some persistent ways that anti-homosexual prejudice is bound up with affect and ego-defense. Consideration of the historicity of the psychological knowledge, coupled with a willingness to continue to work with that knowledge, might create new lines of inquiry within social psychology (see Massey, 2004).
The second productive “tension” concerns the position of identity categories in lesbian and gay psychology. Anti-heterosexual attitudes have been increasingly researched from within a minoritizing frame, and this frame limits our theorizing of the ways that different kinds of identity performances elicit different kinds of prejudicial responses. Queer theorists argued that the available political categories for thinking about sexual politics were insufficient, and prioritized a kind of social reflection in the doing of queerness (Warner, 1993). There were many reasons for this skepticism in addition to Foucault’s skepticism about the ontology of the human subject. Jagose (1993) describes queer as originating within gay and lesbian communities’ responses to the AIDS epidemic. Initially considered a gay disease, AIDS affected many who did not identify with this group, forcing researchers and activists to admit the obvious inconsistency between homosexual identity and homosexual behavior. Queer identity was then based on affinity rather than some essential quality, and attempts to analogize gay communities and gay and lesbian communities along ethnic lines (Epstein, 1985) were attacked as queer theory emerged (Cohen, 1991).
A queer approach would lead us to question the assumption that homosexual behavior and identity can be collapsed, or that one can serve as a valid ontology for the other. Butler (1993, pp. 1-23) has been careful to distinguish her account of performativity from the notion that identities can be constructed whole cloth, volitionally, as social psychologists report doing with their targets in their experiments. Her work then provides grounds for a critique, not only of lesbian and gay psychology, but of experimental psychologists’ more general nature/nurture debates which nominate learned or social behaviors as those which can be changed in laboratory settings, and natural or hardwired behavior as those which can not. Butler’s work would lead us to question the particulars through which social psychological variables are operationalized, rather than to assume that all operationalizations are equivalent (see also Cherry, 1995). What does the wearing of a button accomplish? Does it have the same performative effect as signaling a desire to take part in a psychology experiment pertaining to homosexuality?
Experiments using multiple performances of gay/lesbian or straight identities suggest that such an approach might produce useful theory. For example, Sigelman, Howell, Cornell, Cutright, and Dewey’s (1990) participants read about a straight male target who acquired a gay male roommate either by choice or by accident, and assigned the former target more homosexual tendencies, “gay-stereotyped” traits, lower mental health, and rated him to be less likeable. Experiments have increasingly shown that gay male targets are derogated only when they enact their identities in particular ways. Moreno and Bodenhausen’s (2001) participants only discriminated against an openly gay target (making a pro-gay speech) when his presentation was riddled with errors. Hegarty, Pratto, and Lemieux (2004) have since used vignette studies to show that gay male targets who express discomfort about straight environments are derogated in ways that equivalent targets who stifle their discomfort are not. Such experiments suggest that not all performances of sexual identities are equivalent, leading us to question whether or not experimental studies of gay identity represent a unified literature at all.
Finally, queer theory points to the need for a re-thinking of the relationship between racism and anti-homosexual prejudice in social psychology. Several of the key texts of queer theory linked their anti- essentialist claims to anti-racist scholarship (Duggan, 1990; Warner, 1993). As we have noted above, anti-homosexual inquiry has often looked to anti-racist work for ways of understanding both minority identities and majority prejudices. However, for queer theorists such analogies run the risk of raising one dimension of difference at the risk of obscuring others (Lorde, 1984), overlooking the effects of hetero- normativity on the lives of ethnic minority heterosexuals (Cohen, 1997), and ignoring the co-construction of racial and sexual categories (Sommerville, 2000). Such analogies confuse two very different forms of prejudice which social psychologists also argue are organized by different ideologies and motivations (Biernat, Vescio, & Theno, 1996; Whitley, 1999). At worst, the metaphor can imply that the two forms of prejudice are mutually exclusive, obscuring the ways that some lives are shaped by both. Such political commitments require that social psychologists do not simply position heterosexism as analogous to racism, but begin to use traditional technologies such as surveys and experiments to understand their intersections (see Battle et al., 2002; Clausell & Hegarty, 2003a).
Conclusion
Many of the now canonical queer theory writings cautiously position themselves as supplementary to, rather than in opposition to, scientific epistemologies (Treichler, 1991) and civil rights politics based on identity categories (Butler, 1990; de Lauretis, 1991; Warner, 1993). Similarly, our goal here is not to rehash social constructionist critiques of empirical social psychology that describes such work as meaningless, or simply as the operations of power (Gergen, 1973; Kitzinger, 1987). Rather it seems to us that empirical research on heterosexism constitutes a relatively new mode of creating knowledge, and that those of us who do that work would be well to attend to the sexual politics involved in carrying it out (Hegarty, 2001). Empiricist narratives lead us away from this kind of theorizing, diminish the contributions of our liberationist forerunners, and must be supplemented by the kind of reflection that queer theory engenders. Queer theory does not require the abolition of scientific epistemology. Rather, as Triechler (1991) notes, the trick is to have science play its interpretive part, while also remaining conscious of the distinction between the part and the whole of anti-prejudicial inquiry and action.