Rorschach Research on Male Homosexuality from 1921 to 1969

Peter Hegarty. Journal of the History of Sexuality. Volume 12, Issue 3. July 2003.

In February 2001 the validity of the Rorschach inkblot test became a matter of public concern-once again. During a phone-in debate on National Public Radio a research psychologist who had recently coauthored a lengthy critical review of Rorschach research stated that “this has been brewing for quite some time; the criticism of the Rorschach really began in the nineteen fifties and the nineteen sixties, and has really never quite died down.” The president of the International Rorschach Society replied that “the Rorschach is a wonderful old test; it’s been around for eighty years or so now, used by generations of psychologists, all around the world, who have found it pretty useful.” While psychologists are clearly ready to marshal the history of the Rorschach to their own ends, historians of science and medicine have rarely given the test sustained attention. Several recent and otherwise excellent histories of American mental health sciences pay scant if any attention to the test. But historians of sexuality do need to become familiar with the Rorschach if they are to understand the emergence of “lesbian and gay psychology,” a psychological approach that developed in the 1970s after the emergence of lesbian and gay liberation and the depathologizing of homosexuality. The terms lesbian and guy refer not to the identities of practitioners but to the psychological study of lesbian and gay (and, more recently, bisexual and transgender) persons in the context of a broader heterosexist and gender-restrictive society. Before the 1970s the Rorschach was employed repeatedly as a means of detecting gay men (and, to a lesser extent, lesbians). Yet as early as the 1950s one psychologist, Evelyn Hooker, began to critique this use of the test. The experiment that grounded Hooker’s critique is often understood as the origin of lesbian and gay psychology, and Hooker’s Rorschach work is most often remembered for using the scientific method to gay-affirmative ends.

Any history of the emergence of lesbian and gay psychology makes assumptions about the relationship between psychological science and sexual politics. Contemporary lesbian and gay psychologists often use empiricist ideas to describe their work as above and beyond political matters. For example, John Gonsierek claims that “we [lesbian and gay psychologists] successfully challenged the illness model of homosexuality and defeated it, primarily by critical thinking and arguments based on empirical information.” However, he finds lesbian and gay studies to be a “significant challenge from the politically correct left,” where “dogma is substituted for critical thinking.” According to Gonsierek, lesbian and gay studies are “inward looking,” are “self-absorbed,” and have become “intellectually rigid and irrelevant both to the lives of gay and lesbian citizens and to honest intellectual inquiry.”

In direct opposition to Gonsierek’s empiricist claims, psychologist Celia Kitzinger describes the emergence of lesbian and gay psychology as an extension of patriarchal power. Through rhetorical analysis of research and therapeutic literature, Kitzinger exposes how lesbian and gay psychologists construct lesbians as “well-adjusted” persons only when lesbians understand their sexuality in privatized psychological terms. Women who politicize their lesbianism are described by psychologists as “immature” individuals who have “internalized homophobia.” Kitzinger argues that, for some, lesbian (and gay) psychology is little better than the illness models that it superceded. Thus for Gonsierek, empiricism allows lesbian and gay psychology to transcend politics, but for Kitzinger, psychological science is little more than politics by other means.

I want to argue for a more contextual approach to the relationship between knowledge and power than either of these positions affords. Philosopher-historian Michel Foucault and anthropologist Bruno Latour each presume that relationships between power and scientific knowledge vary across (and partially shape) social contexts. Foucault and Latour have each constructed new objects of inquiry that collapse micro- and macrolevel analyses of social organization by focusing on assemblages of humans and technologies. As anthropologist David Hess notes, Latour achieves this with the concept of the “network” of scientists and machines much as Foucault does with the “apparatus of power.” Latour and Foucault both see language as performative, locate the meaning of discourse in its effects other than in its authors’ intentions, and assume that a statement that becomes cited as scientific has particular performative force. Both also argue that contextualized studies of scientific practices are the means of discovering the particularities of power in fact production. These points of agreement inform my history of the Rorschach presented below.

However, the history of the Rorschach also exposes important differences between Foucault’s analyses and Latour’s. Foucault’s writings on modern sexualities focus on marginalized persons whose locations in fields of power are negotiated through the construction of scientific facts. For Foucault, power comes “from below,” and persons who might appear to be subjected by power come to “play the role of adversary, target, support, or handle in power relations.” Latour, by contrast, often overlooks how scientific networks depend in critical ways on nonscientists’ participation, as feminist scholars of science and technology have often pointed out.

This difference informs Latour’s and Foucault’s theses about the ignorance that characterizes fact production in modern Western societies. Building explicitly on Foucault’s work on sexuality, literary critic Eve Sedgwick claims that ignorance is not “a single Manichaean, aboriginal maw of darkness”; rather, there exists “a plethora of ignorances, and we may begin to ask questions about the labor, erotics, and economics of their human production and distribution.” Foucault and Latour both highlight some examples of ignorance and reproduce others in their accounts of the modern condition. For Foucault, modern societies are peculiar not in that they “consigned sex to a shadowy existence, but [in] that they dedicated themselves to speaking of it ad infinitum, while exploiting it as the secret.” Thus, even as “sexuality” is turned into an object of scientific rationalism, its status as an object of mystery and curiosity is intensified. In contrast, Latour argues that modernity is sustained by an inability to perceive simultaneously the scientific translation practices that make new hybrids of culture and nature (e.g., holes in the ozone layer, the AIDS virus) and the purification practices by which we imagine our world to be made up of separate domains of “nature” and “culture.” For Latour, the secret that modern subjects are only beginning to glimpse is that scientists have always been reconstructing the epistemology of the world even as their work has been understood as simply revealing natural categories. Thus, for both Foucault and Latour, modern epistemologies are not only fundamentally flawed, they also effectively reproduce the problematics they are ostensibly aimed at expunging.

These two theses about modern ignorance are already implicitly related. Latour lists the AIDS virus as one of his first examples of hybridity but goes no further in tracing its relationship to sexual communities. Foucault notes that resistances to power are “furrowing across individuals themselves, cutting them up and remolding them,” but he continues to emphasize sexuality rather than hybridity as the site where modern power is exercised. By examining the Rorschach literature on homosexuality I hope to force a more explicit engagement between Latour and Foucault. Indeed, the history of the Rorschach suggests that Latour’s approach can enhance our understanding of the labor and even the economics of this scientific work but not its erotics. Foucault’s analysis, in assuming that sexual pleasure can be displaced into the practices of psychological research, can provide not only a better understanding of lesbian and gay psychology’s history but also a framework for conceptualizing the field’s current problematics.

The Emergence of Rorschach’s Test in the 1920s

The lifelong research of the Swiss psychiatrist Hermann Rorschach was published in 1921 in a volume entitled Psychodiagnostik. In this volume, which appeared just prior to his untimely death and is his only complete book, Rorschach presented his conclusions about the inkblot test of personality that he had developed. The test consists of ten cards depicting symmetrical inkblots, seven of which are monochrome, three of which are in color. These cards are usually presented in a fixed sequence to a subject who, in describing the inkblots, is understood to “project” his or her psychology onto them.

Rorschach categorized responses to the blots based on the “determinants” of the subject’s response. Determinant categories included physical factors such as the inkblot’s color, the perception of movement, responses to the whole or a part of the inkblot, the use of white spaces between the inkblots, and so on. Rorschach regarded the determinants as a means of capturing an individual’s mode of experiencing the world; as he remarked, from the test “we do not know what he [sic] experiences, but rather how he experiences.” While Rorschach explicitly privileged the “how” of experience over the “what,” some determinants had concrete referents. For example, references to humans and animals and “original responses” were all considered distinct determinant categories.

Most of Rorschach’s work involved clinical patients and psychiatric diagnoses, but some included social research. As Foucault suggests and Kurt Danziger, a historian of psychology, details, modern human sciences draw on contradictory epistemologies and incompatible research practices in their approach to the individual subject. Clinical diagnosis and social research, for example, involve radically different assumptions about the prior relationship between the persons involved, the embodied skill required for testing, and the care that is necessary in choreographing the testing event. Acknowledging this, Rorschach prescribed separate testing procedures for clinicians and for social researchers. Clinical testing practice required almost ritualistic precision: “[T]he length of the extended arm is the maximum permissible distance. Care must be taken that the subject does not catch a glimpse of the plate from a distance.” Social research required less caution. Discussing an explicitly racist use of the test, Rorschach noted that “the test itself is technically so simple-it can be done through an interpreter-that it may be done with the most primitive Negro as easily as with a cultured European.” The body of the researcher is considered to be part of the stimulus in a clinical setting but not in a social research setting.

The tension between clinical and social research returns in Rorschach’s discussion of test interpretation. While Rorschach often voiced caution about his interpretive scheme (which has in turn been repeatedly revised by subsequent generations of test users), he was ambivalent about the part that clinical expertise and statistical inference should play in revising the test. While “the experience and practice of the examiner using the same series of testblots counts heavily” (pointing toward embodied experience), “some statistical method” might “avoid false subjective conclusions based on analogies” (suggesting quantitative research). But he considered too much reliance on statistics as problematic, because “many correct subjective conclusions will be stifled at the start.” As contemporary debates about Rorschach validity attest, such tensions between embodied clinical expertise and statistical evidence are far from resolved in Rorschach psychology.

Finally, Rorschach also considered his test to be a means of psychologizing the psychologists who used it. In regard to kinesthetic determinants (i.e., perceptions of movement) Rorschach wrote that “if the observer himself has a personality too inclined to make anaesthetic interpretations or lies at the opposite extreme, it will be difficult for him to judge properly.” Using determinant categories as ground here, Rorschach prescribed norms for test users, contrasting the “experienced” or “practiced” interpreter favorably with the judgment sullied by the psychologist’s individual personality. As all psychologists necessarily have individual personalities, all are potentially vulnerable to being psychologized in this scheme. Objectivity then results from escaping the objectification of one’s personality as a set of determinants and the consequent reduction of one’s expert judgment to a matter of subjective opinion. Psychodiagnostik can even be read as one such successful escape; although he did not say so in his book, Rorschach believed that he himself had a rather extreme kinesthetic personality.

Developing Rorschach Networks in The 1930s

In the decade after Rorschach’s death, his test inspired little active research. However, by the mid-1930s researchers in the United States had begun to reexamine the test and form a network of practice around its usage. The epistemological tensions implicit in Rorschach’s work manifested as quantitative and qualitative research programs. For example, psychologist Marguerite Hertz gathered data from thousands of subjects and used statistical aggregation to develop increasingly detailed mappings between particular responses and underlying psychological types. In contrast, psychologist Bruno Klopfer conceived of the test subject in phenomenological and holistic terms.

Klopfer became the central figure in the American Rorschach network. In 1933 he had emigrated from Germany to Switzerland, where he first practiced Rorschach testing. In 1934 he acquired his first American position as a research assistant to the anthropologist Franz Boas and began to cultivate an interest in the test among psychologists, psychiatrists, and anthropologists. In 1936 he founded the journal Rorschach Research Exchange, which he edited until his death in 1971. He also taught Rorschach techniques through private seminars in the late 1930s and early 1940s. Through his journal, his seminars, and later his textbooks, Klopfer managed to create alliances between mental health professionals and social researchers who had common interests in understanding the symbolic thinking of psychiatric and anthropological others. Klopfer’s own interpretive abilities were much talked about and became a fulcrum about which Rorschach knowledge would often pivot.

World War II and the Use of the Rorschach to Detect Male Homosexuality

The nascent Rorschach network developed further when it became implicated in military work during World War II. Historian Ellen Herman notes that the war allowed many psychologists to simultaneously advance individual careers, display patriotism, and implicate their practices into the workings of government. Rorschach researchers were no exception. In 1942 Bruno Klopfer organized a “volunteer Rorschach unit” to facilitate “systematic cooperation between the fellows and members of the Rorschach institute who are now serving in the armed forces and those who want to do their part while they continue their routine work.”

As there were many recruits and the time for testing was short, the constitution of the American military was shaped by psychiatric expertise to a singular degree. Quick tests that could be administered by physicians with little training in psychiatry were required, and the Rorschach was adapted accordingly. In “group Rorschach” testing, inkblots were presented to groups of subjects (typically by slide projector) who wrote rather than spoke aloud their responses. Group testing also allowed researchers to amass large amounts of data, burdened the subjects with the transcription of their responses, and allowed test interpretation to occur off-site. Group Rorschach’s rendered the examination of the troops more efficient by speeding up the determination of who was fit for military discipline, extending psychological surveillance, and allowing the extraction of new knowledge about individual psychologies.

Klopfer’s volunteer Rorschach unit made contact with thirteen Rorschach workers in the armed forces, and the military and civilian Rorschach researchers convened in 1944 to develop this cooperation. For military researchers, the Rorschach was primarily of interest as a means of detecting malingerers who might feign an illness to get out of particular duty or to get out of the service altogether.

Homosexuality became grounds for psychiatric exclusion from the military for the first time during World War II. In the 1930s psychoanalysts had developed accounts of male homosexuality as both a transitory neurosis and a deep-rooted permanent psychosis. Both theories were cited as truths in self-help materials given to soldiers. Servicemen were strongly encouraged to confess their personal concerns about homosexual desires to psychiatrists, who in turn were required to report to military officials information that raised suspicions about a serviceman being a “sexual psychopath.” As historian Allan Berube puts it, this system left psychiatrists in the double bind of being both therapist and informant. Some psychiatrists privileged the former goal and misdiagnosed individual gay men who would otherwise have received dishonorable discharges. Others privileged the latter to the point of employing biological and psychological tests to distinguish “genuine” sexual psychopaths from “malingerers.”

Among this latter group of psychiatrists, at least three research teams tried to develop the Rorschach as a means of detecting homosexual men among the troops. One researcher explained the need for such a test as follows: “When a soldier comes to the psychiatrist claiming to be a homosexual, the interviewer must determine whether he is dealing with a genuine, chronic homosexual, and as occasionally, with a man simulating homosexuality in order to detect a discharge from the Army.” As the categories of “serviceman” and “civilian” were mutually exclusive and the goal of testing was to insert men into one of these two social categories, sexual orientation was necessarily constructed as a binary trait. Rorschach testing involved a classic Foucauldian “incorporation of perversion.” The “chronic case” was defined as a type and separated out from the “simulator” based on a codification of discourse produced apropos of sex and within a confessional encounter. From being a temporary problem for military discipline, the Rorschach homosexual was becoming a genuine species.

Rorschach researchers insisted that the test could clearly discern a pattern of responses among genuine chronic cases that was not shown by simulators. Although one research team in a military prison utterly failed to make the Rorschach determinants distinguish “sexual psychopaths” from regular prisoners, two other research studies augmented Rorschach’s interpretive schemes by attending to the symbolic content of servicemen’s responses. Their authors were more optimistic about the Rorschach’s potential. Simultaneously, Robert Lindner conducted several experiments on the male inmates at the Federal Penitentiary in Lewisburg, Pennsylvania, using the group Rorschach. Lindner also argued that “the ‘content’ aspect of Rorschach work has generally and unjustifiably been underworked,” and homosexuality figured heavily in his arguments.

Symbolic content analysis gave researchers an expanded vocabulary for describing the homosexual personality. These descriptions were often more colorful than the brightest splashes of ink on Rorschach’s cards. Bergmann interpreted the mention of “huge repulsive breasts covered with the most exquisite delicate brown silk” as evidence of “genuine” homosexuality. Such responses revealed “unconscious disgusts” such that “the revulsion against the feminine body, seeps through these images, despite the effort to ‘dress’ the response up as a normal one.” Lindner inferred that the paranoia of male homosexuals was revealed through “mysterious notions of electrical influence.” Responses that described “a projectile” or “a ship cleaving through water” became evidence of latent homosexuality with “that curious combination of fear of and wish for penetration.”

These interpretations gained credibility from references to psychoanalytic theory. Borrowing Freud’s logic, Lindner argued that content analysis showed how “the Rorschach response, like the dream, is also a ‘royal road to the unconscious.’” Due and Wright interpreted references to “feminized” behavior (such as cooking and cleaning) as evidence of “conscious and unconscious feminine identifications,” “strong fear of assuming the active male role,” “over-attachment and dependence to the maternal figure,” and “failure to develop adequate emotional ties with the father.” Psychoanalytic theories also occluded the vexed positionality of the researchers as therapist informants and justified continued surveillance by psychological testing. For example, Due and Wright described homosexuals as essentially paranoid because “[t]he homosexual, uncertain of his social status with others, [is] constantly exposed to disapproving social attitudes.” While pathologizing their subjects’ fears, they remained silent about how their own testing practices justified them. Content analysis may have illuminated a royal road to the homosexual unconscious, but researchers were careful to describe that path as unfamiliar. Homosexuality was an unconscious conflict involving mysterious notions and curious fears and wishes. When reporting knowledge about male homosexuality, Rorschach researchers also signaled their lack of understanding of gay men’s psychologies.

Postwar Testing: Redefining the Normal, 1945-1958

After the war Rorschach research speeded up enormously. While at first the test was used infrequently, by the late 1950s it had become the most frequently used diagnostic instrument in clinical psychology. Although there are no published Rorschach studies of lesbians from the 1940s, the test was used to study female sex workers associated with a military base in San Francisco during the war. Rorschach tests were also conducted on Nazi prisoners in the Nuremberg jails. In short, several new enemies of the wartime body politic were studied through their interpretations of inkblots. Reflecting a move toward the psychological mainstream, in 1950 the Rorschach Research Exchange was retitled the Journal of Protective Techniques and began to include research on other projective psychological tests.

The publication of William Marshall Wheeler’s Ph.D. dissertation in 1949 enhanced the legitimacy of using the Rorschach as a way to detect homosexuality in the postwar context. Later researchers would refer to the signs for homosexuality simply as “Wheeler signs.” Wheeler distributed questionnaires to therapists at the Los Angeles Veterans Administration Mental Hygiene Clinic asking them to categorize each patient as marked by an “absence of homosexual tendencies,” “repressed homosexuality,” “suppressed homosexuality,” or “overt homosexuality.” Therapists were to report on the “male” and “female” roles that the men assumed during gay sex, the men’s genders (as “essentially masculine” or “somewhat effeminate”), and the relationships in their families of origin (e.g., “in the marital relationship between the parents, who was the dominant parent [or parent surrogate]?”). In short, the questionnaire presumed and perhaps also proliferated the truth of current psychoanalytic theories of male homosexuality.

Each of the one hundred patients identified in the study also completed a Rorschach test. Wheeler compiled a list of twenty Rorschach indices of homosexuality, drawing on his own experience with “twenty-four overt male homosexuals outside the clinic” and the suggestions of Mortimer Meyers and Bruno Klopfer. Wheeler noted the lack of adequate controls and statistical analysis in wartime studies but employed these studies as Latourian allies to warrant his own optimism about content analysis: “The few studies of content, rather than determinants, as reported in the history, did seem to promise chance of both positive results and the formation of relevant hypotheses.”

Like the wartime researchers, Wheeler imposed binary categories of sexual orientation on data that did not readily accommodate them. he first collapsed all of the men categorized as “overt,” “suppressed,” or “repressed” homosexuals into a “homosexual” category. Since most of these men had been originally categorized as “repressed,” it is probable that in this study many more men were categorized as “homosexual” than had ever enjoyed sex with other men. Wheeler compared the Rorschach results of the resulting groups of “homosexuals” and “nonhomosexuals.” Since results varied from sign to sign, he was led to conclude that “further study of the records of homosexuals would be profitable” but also that “the data do not suggest that we should discard any of the twenty signs.” Wheeler also inferred that homosexual men were characterized by paranoia, derogatory attitudes toward women, feminine identification, “anal interests,” and “interest in physical relationships between like beings.”

Wheeler went beyond his predecessors in a crucial regard by conducting a statistical analysis of the clinic therapists’ diagnostic skills. Using correlation coefficients, Wheeler assessed the level of agreement between each of the therapists’ judgments of his or her patients’ sexualitics with the results of the Rorschach test. This required another binary classification of the sample based on the results of the Rorschach tests. However, eighty-five of the one hundred men evinced at least one of Wheeler’s twenty “homosexual signs.” Somewhat arbitrarily, Wheeler determined that the Rorschach indicated homosexuality only when three or more “homosexual signs” were detected. Wheeler concluded that therapists were better at detecting homosexuality if they were psychiatrists (rather than psychologists or social workers), if they had been in psychoanalysis themselves, or if they were rated by the chiefs of the clinic as good judges of patients’ homosexuality. Wheeler went beyond Rorschach’s comments about the effects of clinicians’ personalities on their judgment; he used statistics to turn diagnosis into an explicit object of knowledge. In the process his work forged a tentative alliance between statistical inference and the more holistic interpretive practice that had been in conflict in the Rorschach network in the 1930s. The utility of the hypothetical signs suggested by Klopfer and Meyer were confirmed by the most modern statistics. Not only did symbolic content analysis have validity, but clinicians exposed to psychoanalysis were found to be the best diagnosticians.

Reversing the Rorschach: Evelyn Hooker and the 1950s

The postwar period has been described by many historians as one of both intense repression of homosexuality and productive community organizing. The military ban on homosexuality informed the later inclusion of homosexuality in the Diagnostic and Statistical Manual of the American Psychiatric Association in 1952. Executive order 10450, signed by President Eisenhower in 1953, extended the ban on military service to all government employment. In addition, police entrapment, harassment in gay bars, and interference with mail were common in the postwar period.

Ironically, however, the war had allowed many men and women opportunities to discover and recognize gay and lesbian desires, identities, and subcultures. New freedoms were experienced by many women through wage work, by many servicemen through travel beyond U.S. borders, and by all through living outside of the nuclear family. Many gay men and lesbians were neither willing nor able to return to the heterosexual lifestyles that were increasingly presented as normative. As Alfred Kinsey and his colleagues were to observe, homosexual acts among American men and women were far more ordinary than either public opinion or repressive state practices would suggest. Within the large port cities of the United States, gay men and lesbians rejected the definition of their behavior as deviant, sinful, and sick and achieved new levels of political consciousness of their oppression as a minority group.

One consequence of increased gay and lesbian consciousness in the postwar era was the formation of homophile societies aimed at improving the social status of homosexuals. The earliest and largest men’s society—the Mattachine society—was founded in Los Angeles in 1950. After a change of leadership in the early 1950s, the Mattachine society, as one of its goals, encouraged the recruitment of participants for research studies in the hope that science would end public ignorance about homosexuality. The society began publication of the monthly Mattachine Review in 1955, which frequently ran articles debating the nature, origins, and health of homosexuality.

Historian John D’Emilio notes that the Mattachines’ strategy of courting liberal scientists had its greatest success with the psychologist Evelyn Hooker. Hooker was uniquely located within the discipline of psychology and the dynamics of cold war hysteria. Trained in methods of animal experimentation, Hooker had no professional stake in assisting the clinical diagnosis of homosexuality. Indeed, she had a personal stake in resisting it, as many of her friends were gay men. She was also a liminal figure in the academy; she had been discriminated against because of her gender, and her second husband had been targeted by the House Committee on Un-American Activities. Moreover, her officemate at UCLA was Bruno Klopfer, who had encouraged her as early as 1945 to use the Rorschach to study gay men outside the clinical context. In 1953 Hooker secured a grant from the National Institute for Mental Health to study “non-patient homosexuals” and began the Rorschach study for which she is most frequently remembered.

Wheeler’s research set the norm for subsequent studies of homosexuality in the 1950s-studies that drew their research subjects from clinical and legal institutions. The Rorschach “homosexual signs” were even used to help diagnose conditions such as schizophrenia and the personality of peptic ulcer patients. Postwar Rorschach studies rarely acknowledged even the possibility of gay men faking responses to the test or otherwise subverting its logic.

For her study, in collaboration with the Mattachines and others, Hooker recruited thirty gay men whom she tested with three projective tests: the Rorschach, the Thematic Apperception Test (TAT), and the Make a Picture Story Test. She also recruited a “control group” of thirty heterosexual men who completed the same tests. Hooker often remarked later that the heterosexual men were difficult to recruit because of the stigma attached to participation in a “sex study.” Both Bruno Klopfer and Mortimer Meyer interpreted each Rorschach protocol, assigned each testee an “adjustment” rating, and guessed each man’s sexual orientation. Hooker found the overall adjustment level of the two groups to be similar. More strikingly, neither Klopfer nor Meyer could detect the sexual orientation of the testees from their Rorschach protocols alone. Like Wheeler, Hooker had made clinical diagnosis an object of psychological knowledge. Unlike Wheeler’s, her work threatened to delegitimize the entire search for “homosexual signs.” Once again, statistical inference and clinical expertise appeared to be divergent, and Hooker’s study positioned the former as a means of gaining epistemological leverage over the latter.

Like previous Rorschach research, Hooker’s experiment constructed sexual identity as a binary trait. Her work formed part of a shift in psychological theory toward reliance on the terms independent variable and dependent variable to describe psychological causes and effects. The causal independent variable in Hooker’s experiment was sexual orientation, and she required a clear categorization of the subjects as either “homosexual” or “heterosexual.” Like Wheeler, Hooker codified her subjects’ self-descriptive speech acts to stabilize this categorization. Each “heterosexual control” subject was informed that he was in a “sex study” only after he had agreed to participate, and each was then asked about homosexual experiences “in a matter of fact way and only after a good relationship of co-operation had been established.” Hooker then excluded some men from the control group on the basis of how they answered her question about homosexual experiences: “If the individual seemed to be severely disturbed by the question, or responded in a bland way, or denied it vehemently, I did not include him in the sample of 30. It is possible, though I doubt it, that there are some heterosexuals in my group who have strong latent or concealed overt homosexuality.” Par from a simple methodological caution (which Kinsey’s work had recently shown to be necessary), Hooker’s description of her practice here constitutes a performance of both scientific objectivity and heterosexuality. Although she was arguing that the Rorschach could not detect sexual orientation, she asked her readers to believe that she could perform such detection unaided. This may have been true, as Hooker was “wise” to the gay subculture of 1950s Los Angeles. However, she does not appeal to her participation in that subculture to bolster her authority. Instead, she leaves the reader ignorant of the interdependence of her independent variable and her liminal position in the community from which her subjects were drawn.

Hooker studied only gay men, both in her experiment and in later qualitative research. Later in life, she recalled of her Rorschach experiment:

Later, once the results were known, I was often asked by gay women “Why didn’t you do for us what you did for gay men?” First I said “You didn’t ask me, and the men did.” But there’s more to it than that. Suppose I had gone to NIMH in 1953 and said, “I want to do a study of lesbians.” The first thing that would have happened, I am convinced, is that they would have said, “We think this bears investigation.” They would have been thinking, “Perhaps she herself is a lesbian.” I think that may have been one question in John Eberhart’s [of the NIMH] mind when he came and spent the day with me. He knew that if I was going to go on in this field, I had to be above reproach. I would have to be pure as the driven snow.

In her 1957 paper, Hooker also uses the metaphor of purity, noting that she “attempted to secure homosexuals who would be pure for homosexuality; that is, without heterosexual experience.” Bruno Latour describes “purification” as the process by which the modern self is separated from others such as “madness, children, animals, popular culture, and women’s bodies.” In Latour’s work, any link between purification and sexuality remains implicit. By contrast, in Hooker’s writings, purity involves the separation of sexual categories (which are also categories of mental health and illness) and is described from Hooker’s unquestionable location as a mentally healthy heterosexual. In this case, what Latour calls “purification” is utterly bound up with a constructed ignorance about the possibility of Hooker being anything other than straight.

Latour also notes that the meanings of a scientist’s work are not to be found by taking that work at face value but by looking to the uses of that work in later publications. If lesbian and gay psychology has developed through the recognition of sound empirical science and Hooker’s experiment is foundational to that development, then that development clearly depended upon homophile groups far more than on the network of Rorschach users. Even after Hooker’s study was published, Rorschach researchers continued to investigate signs of male homosexuality. In later literature reviews Hooker’s work was interpreted as negative evidence against Wheeler’s claims but did not undermine the validity of searching for homosexual signs. In contrast, the homophile press rapidly constructed Hooker’s work as original and authoritative. A comment on her study appeared in the Mattachine Review even before her 1957 paper was published, and the Ladder printed a favorable summary of Hooker’s views. The entire text was published soon thereafter in the Mattachine Review. Although it was accompanied by the caveat that “[t]he following article is not easy to read,” readers were urged to work through it “whether they comprehend Dr. Hooker’s reference to ‘chi square’ and ‘grand medians’ or not.” Hooker’s experiment rooted her progay stance on authoritative scientific ground, but that authority also distanced the people it represented from engagement with her arguments.

In the 1960s Frank Kameny would challenge the economy of scientific credibility in which heterosexual “experts” were courted to speak on behalf of gay men and lesbians. Conceding that Hooker’s study was rhetorically useful, Kameny bemoaned that the Mattachines’ participation in her research involved “nothing more than supplying candidates for experimentation” and argued for more explicit rights of epistemological self-determination for lesbians and gay men. Contra Gonsierek, Hooker later acknowledged that it was Kameny’s radical challenge to the psychiatric orthodoxy that prompted the depathologizing of homosexuality in the early 1970s.

The Cognitive Reversal of the Rorschach, 1965-1969

By the late 1950s more psychologists were clinicians than were laboratory-based scientists, and Hooker’s experiment was only one of many projects that critiqued clinical practice at this time. Part of this general critique involved a more thorough construction of clinical judgment as an object of statistical knowledge linked to the national economy. As one author put it: “Suppose that only 5% of people who are ABEPP diplomats, who have attended Rorschach workshops and have administered 300 or more Rorschach under supervision can arrive at clinically useful conclusions with the instrument; the taxpayer is being somewhat defrauded if the other 95% draw part of their salary on this basis.”

Hooker’s engagement with this broader statistical attack on clinical expertise was minimal. Indeed, shortly after the publication of her Rorschach papers, she published a loving account of Klopfer’s diagnostic powers, and later she trained as a clinician herself. Ironically, Hooker’s experimental critique of Rorschach diagnosis occurred as she herself became less engaged in psychological science and more engaged in clinical practice.

However, other experimental psychologists found Rorschach diagnoses of homosexuality to be a useful target. Loren and Jean Chapman characterized the Rorschach as a partially valid test of homosexuality, noting that “clinicians usually show substantial consensus as to specific meanings of various test responses, although these same meanings may have been discredited by research evidence.” Loren Chapman had recently published experiments that, he claimed, demonstrated a general human bias called the “illusory correlation,” which compelled individuals to overestimate the degree to which stimuli that were conceptually related had been paired. Moving beyond Hooker’s critique, the Chapmans used this concept of “illusory correlation” not only to critique clinicians’ uses of the Rorschach but also to psychologize clinicians in cognitivist terms.

The Chapmans reviewed some of the Rorschach studies described above and concluded that only two of Wheeler’s signs could still be considered valid. The “valid signs” were the perception of “human or animal-contorted, monstrous, or threatening” in response to the seventh card and a “human or humanized animal” in response to the fifth card. In contrast, many of Wheeler’s signs were deemed “invalid,” although they had “associations” with homosexuality. These invalid associative signs included references to “anal content,” “feminine content,” “male or female genitalia,” “sexual confusion,” and “sexual uncertainty.” Typical of this period, the Chapmans’ literature review cited Hooker’s 1958 paper as a limited critique of Wheeler’s signs rather than as proof that Rorschach diagnosis of homosexuality was impossible. Indeed, they assumed that certain diagnostic signs of homosexuality were valid, even if they were not the signs that Rorschach clinicians typically used.

The Chapmans surveyed thirty-two clinicians who had used the Rorschach to analyze homosexual responses. The clinicians reported using the “valid signs” in their practice far less frequently than the “invalid associative signs.” Furthermore, a group of undergraduates were asked to rate on a six-point scale the tendency for each of the “signs” to call to mind homosexuality. All the “invalid” signs that the clinicians reported using commonly were rated more likely than any of the “valid signs” to call to mind homosexuality. This suggested to the Chapmans that the invalid signs were conceptually related to homosexuality and were likely to be misperceived as correlated with it.

In the actual illusory correlation experiments themselves, groups of undergraduates were presented with thirty cards, each of which was said to represent an individual patient. Fifteen of the thirty cards identified one of the patient’s “problems” as his “sexual feelings toward other men,” the other fifteen described other problems. Within each set of thirty cards an equal number of cards paired a homosexual man and a heterosexual man with each of the two valid signs and with a third invalid associative sign. After they had seen all thirty cards, the participants were asked if they had “notice[d] any general kind of thing that was seen most often by men with this problem [i.e., had sexual feelings for other men]?” Participants typically judged that the invalid associative sign had been most frequently paired with this problem, although there was no such correlation in the actual experimental stimuli.

Illusory correlation research substantiates the contested objectivity of experimental psychologists over that of clinicians in several ways. Hooker had demonstrated that particular eminent clinicians could not detect gay men with the Rorschach. The Chapmans’ concept of “illusory correlation” explained clinicians’ overconfidence in psychological terms. Moreover, it allowed clinical judgment to be studied as an abstracted phenomenon; no clinicians were necessary in the illusory correlation experiments; available undergraduates could conveniently stand in for them. Moreover, the notion of the illusory correlation positioned the statistical practices of anthropometry as an objectivist anchor against which clinicians’ subjectivities could be measured and psychologized. The psychology studied at the nexus of Rorschach research and research on gay men was no longer simply that of the gay men themselves but also that of their clinicians. The ontological primacy of Freudian theory was simultaneously replaced by a cognitivist theory of the human mind as a biased and errorprone information-processing device.

The abstraction of diagnosis into an experimental form away from the context of its occurrence and into the university-based laboratory made the job of constructing homosexuality as a binary trait a trivial matter. There were no flesh-and-blood homosexual and heterosexual men being diagnosed in these experiments, and categories of sexual identity could be created whole cloth in the stimulus materials. This process of abstraction remains one of the most vulnerable points in the logic of quantitative psychology; Rorschach defenders in our own time continue to highlight the distance between the assumptions of quantitative research and the practices of Rorschach diagnosis.

Chapman and Chapman’s illusory correlation concept was controversial. The validity of their stimuli was challenged, their study was replicated, and researchers tried to discover ways to undo the effects of illusory correlations on judgments of homosexuality. The paradigm was expanded to critique other clinical tests. To this day, illusory correlation experiments remain a central method of researching stereotyping in social psychology. Yet the centrality of Rorschach diagnosis of homosexuality to that history is frequently erased. When psychology students learn about the illusory correlation in their introductory textbooks, references to the Chapmans’ work are far rarer than references to later illusory correlation experiments that were even more abstract. Some social psychologists have critiqued laboratory-based cognitivist approaches to prejudice as overly technical constructions of social thought that overlook the ideological interests of cognitive theory. Similarly, the continued discussion of the illusory correlation without reference to its early use as a critique of the diagnosis of homosexuality socializes psychology students into a closeted disciplinary narrative. Such education ensures that objective psychologists remain ignorant about both past diagnoses of homosexuality and the necessarily unstable character of all claims about objective knowledge in modern psychology.

Rethinking the Epistemological Politics of Psychological Research

While I was researching this essay, a historian asked me if I was going to “use the Rorschach as a Rorschach” to critique the psychologists I was writing about. This question shows both that critiques of the Rorschach can reiterate the test’s assumptions and that the test is embedded in contemporary understandings of the psychological. Psychologists’ critical uses of “the Rorschach as a Rorschach” have been multiple; even Rorschach’s own writings configure the test’s users as its objects. However, the Rorschach literature also suggests important questions about the current limits of psychological thinking as a form of critique.

From 1945 to 1969 the Rorschach gradually reconfigured the clinicians that used the test as biased statisticians rather than clinical experts, and its male homosexual objects began to appear less like sexual psychopaths and more like targets of prejudice. This was more than an enlightened development within an established paradigm and less than an extension of a monolithic power dynamic into new forms of domination. Rather, precarious alliances between inferential tools, theories, institutions, and individuals’ clinical abilities were disassembled and reassembled to lay claims on the fragile category of objectivity. This history can be accommodated by Latour’s concepts of “translation” and “purification”-but only up to a point. It seems necessary to emphasize (as Latour has not) how psychologists’ objectivity performances require evasion of the position of the knowledge object. Rorschach psychologized clinicians with determinant categories but did not tell his readers that he was an extreme kinesthetic type. Wartime researchers produced knowledge about “genuine homosexuals” but simultaneously signaled their ignorance of homosexuals’ psychology. Wheeler used statistics to show that homosexuality was detectable and that the best clinicians for the job were those with psychoanalytic training. Hooker used links with gay subculture to ensure that her experimental variables were pure but shored up her own heterosexual purity by studying only male homosexuals. The Chapmans’ objectivity proceeded through inferential statistics and used those statistics as a normative benchmark from which clinical judgment could be psychologized as an aberration.

Psychology is commonly thought to be “less objective” than natural sciences. Perhaps this is more than a consequence of the relative complexity of the discipline’s subject matter. Modern Cartesian thinking splits body from mind and positions science as a mental disembodied activity. This ensures a more thorough ignorance about reflexivity in sciences of the flesh than sciences of the psyche. Psychologists can psychologize their opponents (and be seen to do “real science” in the process) more readily than biologists can “biologize” their opponents. As a result, Latourian alliances and “scientific facts” may always be unstable in psychology, and objectivity may appear particularly precarious.

Because of the dependence of psychological objectivity on the psychologists’ escape from the position of the knowledge object, a critical account of psychology must foreground the ways that psychologists evade the positions of their knowledge objects. Latour’s purification/translation couplet might be most helpful here if understood as itself a modernist binary that is “structured-indeed, fractured-by a chronic, now endemic crisis of homo/heterosexual definition, indicatively male.”106 The Rorschach test was repeatedly directed toward this crisis, but the psychologists who used it were vulnerable to being caught up in its impurities. More than an affectless concept of “translation,” a critical analysis of fact construction in psychology requires something more akin to Foucault’s “perpetual spirals of power and pleasure.” Hooker’s concerns about snowlike purity may well be only the visible tip of an iceberg of such spirals. The present study also shows how psychologists have implicitly presented their heterosexuality while studying “homosexual signs” through explicit homophobia, displays of ignorance about gay sex, erasure of links with gay organizations, the conduct of research on sexual identity in which no gay or lesbian persons are present at all, and even publishing with their spouses.

Given all of these strategies, which a single study of a single psychological test has shown, the traditional empirical psychologist might conclude by saying that “more research is necessary in this area.” Indeed, I want to urge further consideration of the history of the category of objectivity in psychology and its relationship to heteronormativity. The Rorschach provided psychology with a means of studying male homosexuality at a distance from gay sex (and ultimately from gay bodies). Contemporary psychologists affirm lesbian and gay identities more often now than in Hooker’s time. However, there is still very little in the growing field of lesbian and gay (and bisexual and transgender) psychology that matches the frankness with which Lindner spoke of the phenomenology of anal sex in 1946 as a “combination of fear of and wish for penetration.” Lesbian and gay psychologists might do well to recognize and explicate the current operations of heteronormativity that shape what counts as objective science in psychology. In so doing, they might, in a queer way, establish a new form of intellectual kinship with Hooker, who reflected on her Rorschach work toward the end of her life: “[T]o go against the prevailing viewpoint, I didn’t have any problem about that. I suppose I didn’t have any problem about that because very early on I learned that, um, the world of factual discovery, of empirical factual discovery is constantly being remade, changed. So what is called this year evil and whatever, next year may constitute the blessing of the human race.”