How queer surveillance has evolved from the Cold War to the Digital Age
If you’ve ever studied the Cold War in a history class, chances are that you’ve heard of either the Red Scare or McCarthyism. These two campaigns targeted (perceived) communists, but there was a third parallel operation that you might not have heard of. Known as the Lavender Scare, this investigation used similar monitoring tactics to uncover queer individuals who worked in government and national defence positions.
Surveillance of queer people was pervasive throughout Canada and the United States during the 1950s and 60s. Much like assumptions against communists, homosexuality was considered a weakness and security risk to the Canadian State, so political rhetoric often linked the two. Both were viewed to be morally weak, psychologically disturbed, undermining the traditional family, and untrustworthy.
Gender Roles in Government
While the Lavender Scare targeted the queer community, part of the fight was to enforce traditional practices of masculinity and femininity. Queerness was associated with deviant gender practices and character weakness. Queer people were considered unsuitable for work because they were thought to be physically and intellectually weak, while heterosexual men were emotionally stable and physically strong, because they performed socially acceptable gender practices.
Scholars Gary Kinsman and Patrizia Gentile detail the history of the Lavender Scare in their book, The Canadian War on Queers. While gay men and lesbians were both constructed as national security threats and targeted by the Royal Canadian Mounted Police (RCMP), gay men were the primary target because homosexuality, conceptualised as a loss of masculinity, was perceived as the larger social threat. Kinsman and Gentile narrate that women’s labour and sexual activities were regulated throughout the Cold War due to the obsessive need to reintroduce femininity to women who had growing access to work outside the role of wife and stay-at-home mother.
The Lavender Scare reflected increasing “gender insecurity,” as Janice Irvine coined, because women who subverted femininity—and, in turn, heterosexuality—defied the patriarchal family structure and threatened hegemonic societal organisation. The growing awareness of queer existence became an easy target to maintain traditional gender roles.
The underlying current of the Lavender Scare was a fear of anyone who was not ‘normal’—a discriminatory concept limited by race, class, and gender assumptions. Normality associates femininity with weakness, assumes whiteness to be the norm, and upholds a binary of gender and sexual experiences. By existing outside of the traditional binary of gender and sexuality, queerness threatened multiple hegemonic structures all centred around the idea of what ‘normal’ looks like. Queerness was constructed as a deviant expression of sexuality and gender that had to be policed within the national security apparatus.
Surveillance systems administer ‘risk’ and security—as determined by who is ‘normal’ or not—by delimiting which populations are socially illegible and thus, excluded from political belonging. Suspected homosexuals were monitored by the RCMP, both inside and outside of state workplaces. Officers frequented queer spaces (such as bars, parks, and homes), conducted multiple raids to arrest people, and interrogated queer individuals to collect names. They used extreme questioning devices, such as the Fruit Machine, to support their investigation of queer government and national defence workers.
The Fruit Machine
Developed by Dr. Frank Robert Wake in the 1960s, the Fruit Machine was a device designed for the RCMP to detect gay men. The machine’s extreme pseudoscientific and invasive questioning tactics supported a mass expulsion of Canadians from government and national defence work.
The RCMP used psychiatric interviews, medical examinations, and more to test for changes in emotional conditions. These included the plethysmograph (which measures blood volume in the finger by electronic or pneumatic means), the Palmer sweat test (measuring perspiration), word association tests, and the span of attention test (assessing the time spent paying attention to various images).
Yet, central to Dr. Wake’s work was the pupillary response test. This test projected pornographic and artistic images while simultaneously photographing the subject’s eye pupils to examine interest patterns. His test proposal named 60 subjects, divided into four groups of fifteen: ‘normal’ males, ‘normal’ females, homosexual males, and homosexual females.
The Fruit Machine was presumed to be an entirely objective provider of reliable scientific information, but the functional mechanism of the device was pseudoscientific and its results were inaccurate. The technology was based on flawed assumptions about the relation between stimulus and response, the power of visual images as simulators, and the notion that queer individuals and heterosexuals would respond to the stimuli with a frequency different enough to sort them. The pupillary response test failed to account for various heights, different-sized pupils, and variations between eyeball distances. The amount of light coming from the photographs changed with each slide, causing the subjects’ pupils to dilate unrelated to their interest in the picture. Additionally, pupil dilation was extremely difficult to measure, as the change was often smaller than one millimetre.
The Fruit Machine also included masculinity/femininity (M/F) tests, underscoring the role of normative gender and sexuality assumptions during the Lavender Scare. Subjects were often connected to a Palmer sweat test and a plethysmographic device to test anxiety levels. M/F tests explicitly coded homosexuality with effeminacy and lesbianism with masculinity by asking subjects to answer true or false to statements such as:
“I like mechanics magazines.”
“I think I would like the work of a librarian.”
Grounded in gender stereotypes, the M/F test emphasised that men and women should perform gender ‘correctly’ by answering each statement in accordance with patriarchal social norms. Men should like mechanics magazines while women should like working as a librarian; anyone who responds differently is weak and abnormal. By emphasising heterosexual practices of masculinity and femininity, the Fruit Machine worked under the notion that there are only two types of sexuality. The fluidity of sexual (and gender) identities that characterised many queer people undermined the government’s basic research assumptions.
The tests highlight how data is not neutral: it is a fundamentally biased output of unequal social, historical, and economic conditions. When individuals collect, analyse, and curate data, they subconsciously place their inherent beliefs and biases into it. For instance, Dr. Wake and his researchers had to constantly tinker with the technology and tests to get the results they hoped for, demonstrating that data upholds biased power structures that reflect systemic issues of power and privilege.
Those who failed the test were given the choice of either honourable or dishonourable discharge. However, losing jobs was often the least of the consequences for queer people, as Sarah Fodey describes in her documentary The Fruit Machine. Victims struggled with poverty, homelessness, having to go back in the closet, substance abuse, gay aversion therapy, sexual assault, and suicide.
Social Progress
Surveillance technologies and systems have historically targeted queer individuals, often as a means of social control and repression. The surveillance practices of the Lavender Scare reflect the time period, as same-sex activities were illegal throughout the height of the campaign and were only decriminalised in 1969.
In 2017, Prime Minister Justin Trudeau apologised on behalf of the Canadian government for historical harm done against the queer community. His apology acknowledged the witch hunt caused by the Lavender Scare, saying that “the Government of Canada exercised its authority in a cruel and unjust manner, undertaking a campaign of oppression against members and suspected members of the LGBTQ2 community.” Trudeau’s lengthy apology recognising the queer community’s pain, grief, and shame during the Lavender Scare indicates the strides made by the Canadian government—at the behest of queer activists, such as the 1971 We Demand march—towards LGBTQ+ rights.
Nowadays, Canada is generally thought of as a queer-friendly country to visit and live in. There has been growing right-wing hate and anti-trans legislation in the past few years, but queer state surveillance through methods like the Fruit Machine would not be seen as socially acceptable anymore. Yet, this reduced acceptability does not necessarily mean that surveillance is entirely extinct.
Modern Surveillance
A 2018 facial recognition technology from Stanford University suggested that artificial intelligence can identify one’s sexuality through photographs. Developed by Yilun Wang and Michal Kosinski, the algorithm can supposedly tell gay from straight as successfully as a person’s ‘gaydar’.
Drawing on historical physiognomy, Wang and Kosinski’s study found that prenatal hormone theory (PHT)—the idea that our sexuality is determined by hormone exposure in the womb—predicts links between facial appearance and sexual orientation. For instance, PHT anticipates gay men to have “smaller jaws and chins, slimmer eyebrows, longer noses, and larger foreheads.”
Their facial recognition technology used deep neural networks (DNNs)—a type of artificial intelligence that recognises patterns in large datasets to make predictions—tested on 35,326 extracted images of gay and heterosexual men and women on dating sites. Their DNN results appeared consistent with PHT, arguing that queer sexual orientations stem from “the underexposure of male fetuses and overexposure of female fetuses to prenatal androgens.”
Because DNNs can only see patterns based on the data it’s trained on, it reinforces preconceptions of gender appearance and performance. For instance, it learns to recognise the connection between masculinity and beards or femininity and long hair, and those who deviate from the pattern are similarly labelled as ‘abnormal’ or queer. It not only disregards fluid gender and sexual identities; it actively upholds hegemonic standards of what masculinity and femininity encompass.
Additionally, no images of racialised people were included in Wang and Kosinski’s study, raising more questions about the assumption of whiteness as ‘normal’ and the history of racism in physiognomy and science. This absence not only erases the existence and experience of queer people of colour, but also contributes to the historical association of racial and ethnic features with negative stereotypes. Due to data exclusion, the DNN solidified patterns where whiteness is synonymous with ‘normal’ and queer existence, employing scientific language and measurements to disavow flawed biases.
Wang and Kosinski’s work may not be an explicit tool for surveillance, but the underlying beliefs and power dynamics found in the Fruit Machine permeate their work. This is not to say that a person or an algorithm can’t accurately guess who is queer, but rather that sexual and gender identities are not static, fixed data points. Identities are broader and more complex than the three demographics (gay, lesbian, or heterosexual) recognised by the Fruit Machine and facial recognition technology.
Consequences of the Lavender Scare
The Fruit Machine and Wang and Kosinski’s facial recognition technology may seem worlds apart, but they share a common premise: both systems aim to classify and control individuals based on rigid and flawed assumptions about sexuality and appearance.
While the Fruit Machine was explicitly created to target queer people, Wang and Kosinski’s study similarly attempts to read sexuality through algorithmic inference. This modern iteration still evokes the same troubling narrative—that queerness can be detected and categorised based on visible markers.
In an interview with Brian Resnick, Kosinski justified the creation of their technology by asserting that “the beauty of a computer is it can go beyond the stereotype… the stereotype developed by a computer will be an accurate one [because] it’s a data-driven thing.”
Despite Kosinski’s words, the study only reinforces entrenched biases (such as gender insecurity from the Lavender Scare) instead of dismantling them. Wang and Kosinski’s work reflects how the inherent power dynamics in historical processes of state surveillance remain, and how technology has been slightly altered to fit digital surveillance through algorithmic tracking and hegemonic labelling.
The legacy of the Lavender Scare continues to resonate in our digital age, as queer identities are surveilled, albeit in subtler forms. The Fruit Machine may represent an overt form of state discrimination, but the algorithm developed by Wang and Kosinski unveils how technologies continue to mirror the same pervasive biases of the 1950s and 60s. By perpetuating rigid frameworks of normality, these systems echo the Lavender Scare’s drive to surveil, categorise, and control those who defy heteronormative expectations—this time through data patterns and algorithms, rather than through stimulus and response.
Examining the Lavender Scare allows us to question how modern technology reinforces hegemonic structures and reverberates historical surveillance tactics, in subtler ways and under the guise of objectivity. Technological advances may improve lives, but without critical approaches to data, they can just as easily become tools of exclusion, reinforcing outdated stereotypes and binary views of gender and sexuality. Just as the acceptability of the Fruit Machine changed, the ethics behind technology must be carefully scrutinised to protect the queer community’s digital privacy and safety.
