An even-handed debate? The sexed/gendered controversy over laterality genes in British psychology,1970s–1990s

Introduction

For over 150 years, researchers of various disciplinary origin have speculated about the association between handedness (manual laterality) and brain asymmetry (cerebral laterality). Manual preference has borne the promise of providing insight into brain development and language disorders since Paul Broca’s (1824–80) assertion that humans were right-handed because they were left-brained (that is, their left cerebral hemisphere dominated over the right one, analogous to handedness). Based on research on lesion-induced aphasia, the French anatomist claimed that a centre for articulate speech sat in the left half of the brain, and that this made the left hemisphere superior (Broca, 1865). In combination with the idea of a contralateral association between the brain and the rest of the body (that is, the right half of the brain directed the left half of the body and vice versa), Broca’s theory seemed to explain the common superiority of the right hand over the left (Harrington, 1987; Harris, 2016; Leblanc, 2017; Rutten, 2017).

This theory marked non-right-handers as abnormal, if not subhuman. Broca (1865) suggested that some left-handers had brains that mirrored the presumably normal brains of right-handers and that their right-sided speech centres provided superior strength and skill to the left half of their bodies. Ambidexters seemed even more abnormal. Broca and his followers believed that the absence of a clear manual preference derived from a virtually symmetrical brain and that this quality was associated with the brains of lower animals. According to Broca, brain asymmetry was the exceptional human quality, a consequence of articulate speech and other acquired higher cognitive functions (Harrington, 1987). To this day, scientists adhere to the idea that handedness, brain laterality, and cognition are inextricably intertwined (Marcori and Okazaki, 2020).

In the 1970s, 1980s, and 1990s, British psychologists Marian Annett (1931–2018) and Christopher I. McManus (b. 1951) reframed this association between hand, brain, and human intellect in probabilistic terms. Although their theories were similar, McManus and Annett were fierce competitors. They proposed hypothetical genes, whose existence had not been genetically confirmed, that were supposed to influence cerebral and manual laterality. Statistical calculations about the inheritance of laterality were at the core of their endeavours. The resulting models accounted for the probability of laterality types in certain populations, but they did not explain concrete biological processes. Furthermore, the models proposed genetic limits within which chance effects and environmental factors could influence laterality, thereby redefining the boundaries between nature and nurture.

Historian of medicine Howard Kushner (2017) has analysed Annett’s and McManus’s models and juxtaposed them with contemporaneous theories of handedness, but an analysis of the sexed/gendered dimensions of their theories and controversy is wanting. ¹ The present article also considers new sources, including unpublished documents and oral histories. In the following, first, I position the debate between McManus and Annett in the longer history of handedness research and the search for particular genes. Second, I provide short biographies of Annett and McManus. Third, I recount the core aspects of their theories. Subsequently, I analyse interlocking levels of sexed/gendered concerns that shaped this debate.

I begin my analysis of sexed/gendered aspects in the context of the controversies surrounding Annett’s and McManus’s competing theories. Sex/gender emerges as a crucial determinant of human identity, behaviour, and achievement when considering Annett’s career. For large parts of her academic life, Annett defended herself against attacks on her work in addition to compensating for a relative lack of professional resources. McManus mobilized his higher social capital in public critiques of Annett, which bore a sexed/gendered and ageist valence and called into question Annett’s credibility as a scientist. I do not wish to suggest that McManus purposefully sidelined Annett, nor do I want to create the impression that McManus did not work hard for his success. I am rather pointing to a structure of intersectional disadvantages that prevented Annett from attaining equal achievements with similar effort.

Afterwards, I juxtapose Annett’s marginalization as a woman psychologist with her complicity in stabilizing sociocultural hierarchies. Both psychologists published several iterations of their models, which increasingly relied on questions of sex/gender and reproduction. They combined stereotypically sexed/gendered ideas and notions of probabilistic inheritance to craft quasi-evolutionary accounts of the adaptive function of handedness and brainedness. These theories functioned as ‘origin stories’ in the sense developed by philosopher of science Donna Haraway (1984): they naturalized existing social inequities, including hierarchies relating to sex/gender and dis/ability, by providing teleological accounts of how the current world order has evolved.

Sex/gender also impacted genetic handedness research on the level of statistical and computerized technologies. Comparing Annett’s and McManus’s methods and professional success exemplifies the ways in which computers simultaneously empowered their users and stabilized existing power dynamics. There were notable asymmetries in the extent to which Annett and McManus had access to computers and the degree to which computers impacted the psychologists’ recognition in the field. Although Annett made significant epistemological compromises to digitize her calculations, her use of computers became a target for critique while McManus’s computerized analyses conferred credibility upon his work.

Although neither Annett nor McManus foregrounded concerns about sex/gender in their publications, our e-mail exchanges, or the interviews I conducted with them, I found sex/gender to be a pervasive factor in the psychologists’ theories, interactions, and professional recognition. In the early stages of my research, it became clear to me that the debate surrounding the genetic causes of handedness continued for several decades not because it took the scientific community so long to select one theory over the other, but because Annett refused to give in to a competitor with higher social capital. My interactions with Annett led me to suspect that she had subconsciously tried to be seen as more of an equal to her colleagues by formulating a theory that resonated with the pervasive chauvinism of 20th-century psychology. My review of the literature on (anti-)feminist currents in British psychology deepened this impression.

If Annett desired to win her colleagues’ recognition by formulating increasingly sexist and ableist theories, it was to no avail. Annett’s sidelined position in British psychology became clearly visible in 2019, when McManus and his colleague Alan Beaton (b. 1947) notified me that they had recently found out that Annett had died 18 months prior. ² Annett was not part of a professional network that would have allowed her death to be recognized among laterality researchers in a timely fashion.

This article exemplifies the inextricable entanglement of cognitive and social factors in scientific controversies (Longino, 2002), contributes to the scarce history of female psychologists in 20th-century Britain, and examines one instance of scientists’ intersectional situatedness, which has not yet acquired a status in the history of science that is of equal prominence to that in other fields (Berger and Guidroz, 2009; Carbado et al., 2013; Crenshaw, 2019). Additionally, this article echoes many observations from previous scholarship on the history of genetic inquiries into a variety of physical, psychological, and behavioural characteristics. Most importantly, the science of handedness provided categorical distinctions between humans based on their genotypes, and hence resonated with neo-eugenic agendas (Comfort, 2012; Duster, 2003).

The century of genes

Eugenicists and educators first formulated Mendelian models for the inheritance of handedness (Chamberlain, 1928; Jordan, 1911; Ramaley, 1913; Rife, 1950; Trankell, 1955). In the second half of the 20th century, psychologists and geneticists combined the Mendelian search for a handedness gene with the search for a brainedness gene (Annett, 1964; Levy and Nagylaki, 1972). None of these models could account for how handedness was passed on in families, but the studies provided McManus and Annett with crucial methods for measuring laterality as well as vast data sets that traced handedness across generations in Europe and the US. ³

Some peculiarities of handedness data have opened avenues for researchers to link theories about manual preference, brain anatomy, and cognitive abilities to sex/gender and reproduction. Generations of scientists have frequently found that: (a) men were more likely to be left-handed than women; (b) children were more likely to be left-handed if their mother was left-handed than if their father was left-handed (the so-called maternal effect); and (c) the rate of left-handers among twins was higher than that among singletons, and the manual preferences of identical twins frequently differed (Kushner, 2017; McManus, 2002; Porac, 2016). If handedness were strictly genetically determined, then only fraternal twins should be discordant for handedness, but the data suggested that this was not the case.

The search for a handedness gene stood out from contemporaneous inquiries into the genetic bases of intelligence (Gould, 1996), sexuality (Brookey, 2002; Rosario, 1997), and race/ethnicity (Bell and Figert, 2015; Koenig, Soo-Jin Lee, and Richardson, 2008; Wailoo, Nelson, and Lee, 2012; Yudell, 2014) in two important ways. First, since 1865, hereditary mechanisms of manual preference had been formulated with recourse to the brain. Whereas numerous 20th-century scientists regarded sexuality, race/ethnicity, and intelligence to be directly genetically determined, handedness researchers focused more on the brain than the hand. They formulated theories of genetically induced brain structure and function, and then explained the inheritance of handedness with recourse to these theories about the brain. Second, the proposed genetic connections between the hand and the brain resulted in a probabilistic concept of manual preference, whereas the quest for the ‘gay gene’ (Hamer and Copeland, 1994) or the hereditary mechanism of ‘general intelligence’ (g; Spearman, 1904) were decidedly deterministic. Instead of proposing a gene that strictly determined handedness, Annett and McManus developed models for genetic limits within which environmental influences and chance could contribute to the development of handedness and brainedness.

The main players in the late 20th century

Annett and McManus were the main players within a novel research community, and their decades-long debate often took a personal tone. ⁴ McManus and Annett met in person on several occasions but never worked together. They had one friendly encounter in 1974 (Annett, 2004). Subsequently, Annett’s and McManus’s exchanges were intensely competitive. Their age difference may partially account for this fact, but conflicting sociopolitical standpoints may have overshadowed their similarities in methods and theories even more. The difference in the psychologists’ social positioning—an older female researcher without a strong academic position against a younger, highly connected male professor—highlights crucial social dynamics of academic life. These aspects, which were at the same time structural and personal, are important to mention explicitly as the research into the seemingly neutral domain of handedness turned out to be centrally concerned with sex/gender differences and gendered intellectual characteristics.

Annett was criticized by several colleagues throughout her career, but I focus on her interactions with McManus for several reasons. First, McManus’s critiques were the most far-reaching and the most personal. They brightly illuminate underlying professional and personal tensions. Moreover, Annett took McManus’s critiques more seriously (and personally) than anybody else’s. During our interview and e-mail correspondence, Annett avoided mentioning McManus’s name and made clear that she perceived much of her professional life as a defence against his attacks. ⁵ Finally, McManus’s and Annett’s theories and methods were strikingly similar. Juxtaposing the psychologists’ professional experiences against the backdrop of these commonalities brings out most clearly the differences that the intersections of sex/gender, age, institutional affiliation, and use of technology could have on academic careers in late 20th-century Britain.

A first-generation college student, Marian Annett (née Drabble) received her undergraduate education at Bedford College (BA 1952), a women’s college later merged with Royal Holloway College. ⁶ She completed her PhD at the Institute of Psychiatry at the University of London in 1965. Early in her career as a clinical psychologist, Marian Drabble met experimental psychologist John Annett (b. 1930) during a campus visit to Bristol. That day, Marian and John were applying for positions as research assistants in the Burden Mental Research Department at the Stoke Park Colony, an institution for individuals with mental, developmental, and learning disabilities. ⁷ Both were hired. After they got married in 1955, John’s career advancement took precedence. Marian finished her dissertation as an external candidate and never held a professorship. In her own words, she practiced psychology as ‘a hobby’, mainly at home and without much institutional support. ⁸ She worked in researcher positions at the University of Hull and the Open University in the 1960s and 1970s. Later, she was affiliated as a lecturer and reader with the psychology departments at the Lanchester Polytechnic (now Coventry University) and the University of Leicester. The latter conferred emerita status on her in 1994. Marian Annett passed away in April 2018. ⁹

McManus, too, was a first-generation college graduate. Apart from that, the psychologists’ biographies differ significantly. McManus provided me with the following biography:

Chris McManus received his psychology education at the University of Cambridge (BA, 1972, Ph.D. 1979) and his medical training in Cambridge and then Birmingham (MB ChB 1975) and London (M.D. 1985). He lectured in various parts of the University of London from 1979 onwards, and in 1993 was appointed Professor of Psychology and Medical Education jointly at Imperial College and University College London (UCL). Since 1997 he has worked in the psychology department and the Medical School at UCL and is currently employed part-time and continues to conduct research in various areas. ¹⁰

To make sense of the disparity between McManus’s and Annett’s professional achievements and public recognition, it is important to keep in mind that Annett continued an unfortunate tradition of sidelined female scientists. Historical scholarship on women psychologists in Britain, and on British post-war psychology more generally (Collins, 2012: 205), is rare, but US-centred histories can provide some hints about the situation of 20th-century female psychologists (Fancher and Rutherford, 2012; Lawson, Anderson, and Cepeda-Benito, 2017; O’Connell and Russo, 1983; Rossiter, 1982, 1995; Scarborough and Furumoto, 1987; Stevens and Gardner, 1982b). In the US, the number of women psychologists grew rapidly from the 1960s onwards, and more than one-third of them worked in clinical psychology (Stevens and Gardner, 1982a). Similarly, in Britain, clinical psychology saw ‘the most vigorous growth’ after World War II (Rachman, 1983: xi). The large number of female clinical psychologists did not necessarily improve women’s standing in psychology. Influential male psychologists perceived clinical psychology and its concrete applications, not to mention the high number of women in the field, as a threat to masculine-coded ‘pure experimentalism’, and hence to psychology as a profession (Rutherford, 2015: 269). From the 1940s onwards, North American women psychologists formed organizations to counteract sexism in psychological practice and ideas (Johnson and Johnston, 2010; Katz, 1991; Pyke, 2001; Radtke, 2011; Russo and Denmark, 1987; Unger, 1998).

Annett, however, did not directly benefit from these developments. Members of the British Psychological Society did not form a Psychology of Women Section until 1987 (Wilkinson, 1990), and the situation for British women psychologists remained less ‘hopeful’ than in North America (Squire, 1989: 11). Moreover, Annett was living in a relatively traditional family, which did not make pursuing a professorship a viable option. ¹² Annett’s epistemologies remained similarly traditional, with a few exceptions noted below. When feminist psychologists inside and outside of Britain became more vocal, Annett did not endorse her colleagues’ ideas that sex and gender were intertwined and that psychology should stop essentializing the social construct of gender (Burman, 1990; Crawford and Marecek, 1989; Hare-Mustin and Marecek, 1990; Morawski, 1994; Rutherford and Pettit, 2015; Ussher and Nicolson, 1992).

Annett’s theory

Annett dedicated virtually her entire professional life to laterality and set out to find evidence for her conviction that that there were three handedness phenotypes: right-, left-, and mixed-handedness (Annett, 1967). Annett (1985a: xi) emphasized that her handedness model was not ‘received all at once in some sudden illumination from on high’. She published overhauled versions of her theory in four stages, starting with a publication from 1972. With this paper, Annett introduced an element of chance into the inheritance of handedness. This concept revolutionized the aetiology of handedness and, compared to earlier Mendelian models, offered a new understanding of the interplay of nature and nurture in the context of handedness and speech laterality.

Based on statistical calculations and theoretical considerations, Annett (1972: 356) propositioned a so-called ‘“right shift” factor’, a genetic influence that biased human handedness to the right and speech lateralization to the left. Foundational to this theory was the definition of handedness as ‘a by-product’ of left-hemispheric speech lateralization (Annett, 1985a: 260). In a self-published pamphlet that Annett rushed to print in 1978 (with the intention to publish it before McManus defended his dissertation), Annett called her new model the Right Shift (RS) theory. The RS theory assumed that an unspecified and biochemically unconfirmed gene either carried the RS factor as an allele or not. In carriers of the allele (approximately 81% of the population), Annett (1978) argued, the RS factor caused the growth of the speech centre in the left hemisphere. This left-lateralized speech centre had a secondary influence on (but did not strictly determine) the individual’s handedness type. Annett further suggested that carriers of the RS factor were strongly predisposed to right-handedness as a consequence of their genetically induced left-brainedness. In non-carriers of the RS factor (approximately 19% of the population), the speech centre developed in the right or left brain hemisphere based on chance. Consequently, 50% of those individuals were right-brained and 50% left-brained, and handedness also depended on chance.

Annett did not explain the biological mechanisms of her proposed hand–brain connection, although it was integral to establishing the credibility of her heritability rates. In this sense, Annett’s model, although it provided a new explanation of the origins of handedness, can be described as a probabilistic version of Broca’s early theory of laterality. Annett (1978) tried to escape the conservatism of her theory about the connection between brainedness and handedness by explaining that the two lateralities were statistically correlated, but not causally linked within individual organisms as Broca (1865) had suggested. Right-handedness was most likely concomitant with left-lateralized speech, Annett (1978) suggested, but any combination of speech laterality and manual preference could occur.

The reasons why Annett called her model the Right Shift theory provide insight into the ways in which 19th-century concerns about human exceptionalism were sustained by 20th-century computational statistics in handedness research. Annett (1972) assumed that the hand-skill distribution among non-carriers of the RS factor corresponded to a bell curve peaking at zero, the point of absolute ambidexterity. Per the RS theory, carriers of the RS factor were also normally distributed, but the correlating bell curve peaked somewhere to the right of zero. This shift to the right of the normal distribution of hand skill in carriers of the RS factor was what gave the theory its name. According to Annett’s model, the Right Shift marked a crucial difference between humans and non-human animals. Annett (1972) believed that non-human primates either did not have any consistent lateral preferences, or preferred their left or right hands (or paws or feet) at a rate of 25% and were ambidextrous in 50% of cases. Consequently, the left–right skill distribution for non-human animals peaked at zero, but the human population as a whole displayed a Right Shift.

Annett invested significant conceptual effort in making the RS model solve the three core problems of handedness research: (a) the excess of left-handed men; (b) the maternal effect; and (c) the twin data. To explain the higher rate of left-handedness in men than in women, only one a posteriori specification was necessary: Annett (1972) claimed that the RS factor was expressed more strongly in females than in males. This meant that female carriers of the RS factor had, on average, a more strongly left-lateralized speech centre than male carriers, which contradicted the contemporaneous idea that men’s language centres were strongly left-lateralized while women’s language abilities were bilaterally distributed (Hirnstein, Hugdahl, and Hausmann, 2019; McGlone, 1980). This difference in the degree of Right Shift, according to Annett, also caused cognitive sex/gender differences. For instance, she suggested that it explained why girls and women had better verbal skills than their male counterparts, and why, in turn, men and boys excelled at visuospatial tasks (Annett, 1972).

This tweak also solved the problem of the maternal effect (Annett, 1985a). If the Right Shift were expressed less strongly in males than in females, then more male than female carriers of the RS factor would be left-handed. Consequently, Annett claimed, left-handed fathers were more likely to pass on an RS allele to their offspring than left-handed mothers, and hence the former would have more right-handed offspring than the latter. Annett (1999a) revised this idea towards the end of her career by suggesting that maternal inheritance was an artefact in the survey data: research subjects tended to under-report their mothers’ left-handedness, but not their fathers’.

The RS theory also confronted the problems posed by twin data. The chance factor in the first version of the RS theory effectively explained why the handedness of identical twins could differ, but it left open why singletons were more likely to be right-handed than twins. A similar modification to the one that accounted for the excess of male left-handers seemed appropriate in this context as well: Annett (1978) propositioned that the RS factor was expressed less strongly in twins than in singletons. She did not offer empirical evidence to back this hypothesis.

In the following years, Annett carried the statistical rhetoric one step farther, aligning her analysis with an emerging logic of risk factors in the biological and psychological sciences. While the concept of chance had been integral to earlier versions of the RS theory, the notion of being at risk was a novelty. In the late 1980s, Annett suggested that both homozygous groups—the carriers of two RS alleles and the non-carriers—were ‘at risk’ of lower-level brain function, which most prominently manifested in the form of dyslexia (Annett, 1987: 46–7). Annett explained that the heterozygous population would on average perform on a higher intellectual level, providing them with a reproductive advantage over other genotypes (ibid.; Annett, 1993, 1995b; Annett and Manning, 1989, 1990).

In the late 1990s, Annett (1996c) presented the final version of her theory. It included the suggestion that homozygous carriers of the RS factor (if the allele showed a slight mutation) were also ‘at risk’ of developing schizophrenia and autism (quoted from Annett, 1999b: 177). Although claims of the genetic underpinnings of psychiatric illnesses and developmental disabilities have a long history, Annett’s association did not have much grounding in clinical psychology. It was based on her assumptions about the connection between brain anatomy and cognitive abilities, as well as her comparison of frequencies of the genetically unconfirmed RS factor and the incidence of dyslexia, autism, and schizophrenia. These associations seemed biologically impossible (Corballis, 1997) but resonated with contemporaneous claims that left-handedness was associated with various pathologies (Kushner, 2017; Porac, 2016).

Annett recommended population screenings for the RS factor, grounded in these theoretical considerations and in her sympathy for the hardships that supposedly accompanied certain kinds of laterality, including her left-handed son’s marginalization in school because his verbal development was ‘a little bit slower’. ¹³ In Annett’s view, the early identification of homozygous carriers and non-carriers of the RS factor would allow for targeted pedagogical or psychiatric interventions (Annett, 1996b, 2011; Annett and Kilshaw, 1984). In advocating this, Annett reinforced the RS model’s tendency to create a genetic typology of personalities and cognitive types.

McManus’s theory

Despite his close engagement with Annett’s work in his PhD thesis from 1979, ‘The Determinants of Laterality in Man’, McManus maintained that Annett’s model and his own had been ‘independently developed’ (McManus and Bryden, 1992: 126). Annett disagreed. She believed that ‘[McManus’s] theory was to rival and supplant the RS theory’ (Annett, 2004: 147). Suffice it to say that the two models were very similar. Most importantly, both psychologists insisted on the importance of chance, and both assumed a genetic connection between brain asymmetry and manual preference. In doing so, both connected handedness with the allegedly exceptionally human quality of speech, although McManus was less explicit about differences between humans and non-humans than Annett.

McManus’s (1979) theory was based on two hypothetical alleles, as opposed to Annett’s single RS factor. McManus formalized the chance element in handedness inheritance by introducing an allele that would determine an organism’s susceptibility to environmental influences. He proposed that the dextral allele D determined handedness and that the chance allele C determined the limits of environmental influences on handedness. According to the theory, these two alleles could combine to the homozygotes DD (100% right-handed individuals) and CC (50% left- and 50% right-handed individuals, because chance was the only influence on the direction of handedness). In the heterozygote (DC), McManus suggested, D was partially penetrant; in other words, more than 50% but less than 100% of the DC population were right-handed. Initially, McManus (ibid.) offered two potential values for the proportion of right-handers among the DC population: 87.5% and 75%. In the second version of his model, he settled on 75% (McManus, 1984), based on computerized statistical analyses of survey data (McManus, 1985a).

Extending the reach of his genetic model from the hand to the brain, McManus (1979) proposed that the C and D alleles regulated not only handedness but also speech lateralization directly. In contrast to Annett’s RS theory, handedness was thus supposed to have the same genetic cause as—and not be a by-product of—brain asymmetry. According to McManus’s (ibid.) theory, D advanced left-hemispheric speech and right-handedness independently but with the same probability. DD genotypes would always be right-handed and left-brained. Heterozygotes could exhibit any combination of left- or right-handedness and brainedness, but right-handedness and left-brainedness would be most prevalent. In the CC genotype, handedness and speech laterality would vary randomly.

McManus turned chance itself into a genetically determined factor by formalizing serendipity into an allele that presumably regulated handedness and speech lateralization. With C, the gene for responsiveness to the environment, McManus brought nurture into the realm of nature in a way that conceptually resembles contemporary epigenetics (Richardson and Stevens, 2015). In doing so, McManus biologized certain aspects of nurture more rigidly than Annett did. Per the RS theory, chance would affect brain lateralization to an unspecified degree in the absence of the RS factor. Under McManus’s model, the limit for all chance variations was precisely defined for each genotype.

Like Annett, McManus (2002) eventually suggested that the DC population benefitted from a heterozygote advantage. He claimed that the DC population, no matter the handedness phenotype, would be fitter than DD or CC individuals. McManus (ibid.: 228) seemed less confident than Annett in phrasing his hypothesis and called it ‘speculation’. Most importantly, in contrast to Annett, he did not offer concrete thoughts on the mental and cognitive inferiority of homozygotes. Instead, he emphasized the superiority of heterozygotes by proposing that DC individuals might have the most favourable degree of randomness in the distribution of lateralized brain centres, allowing them to survive longer and/or reproduce at a higher rate.

Like previous handedness researchers, McManus struggled with the problem of an excess of male left-handers. He initially considered that there might be a variable expression of the D allele in different sexes (McManus, 1979, 1985a). In the third and final version of his theory, he linked sex/gender and handedness much more closely (McManus and Bryden, 1992). Together with his Canadian colleague Philip Bryden (1934–96), McManus proposed that a hypothetical sex-linked factor could inhibit the expression of the D allele. This modifying factor, they claimed, was a recessive allele on the X chromosome and hence affected males more frequently than females.

This modification of McManus’s original theory seems equally ad hoc as Annett’s amendments regarding the variability of the Right Shift in men and women or twins and singletons. Likewise, it constituted a potential solution to the problem of the maternal effect. According to McManus and Bryden (1992), a significant number of left-handed fathers were concealed DD homozygotes who carried the inhibiting factor. The psychologists claimed that this special class of left-handers could pass on only D alleles, not C alleles. If the offspring of concealed DD fathers inherited an X chromosome from their mother that lacked the inhibiting factor, they would be equipped with at least one D allele (from their father) and at least one X chromosome with the dominant non-inhibiting allele (from their mother). In these children, directional asymmetry would not be suppressed, and they would be right-handed with a probability of significantly more than 50%. This way, left-handed fathers would end up with right-handed children.

What about left-handed mothers? McManus and Bryden (1992) assumed that fewer left-handed women than men had DD genotypes (because the inhibiting factor was recessive and only a double dose of X chromosomes with the inhibiting factor allowed for left-handedness to develop in female DD genotypes). Consequently, most left-handed mothers would pass on at least one C allele to their offspring, which increased the likelihood of left-handedness. Like Annett, McManus (2016) proposed a different solution to the puzzle towards the end of his career. ¹⁴ Apparently, the maternal effect disappeared if one assumed that a certain percentage of left-handed social fathers were not biologically related to their right-handed children, who had right-handed biological fathers. In the end, Annett and McManus agreed that the maternal effect was a sexed/gendered artefact, be it for reasons of subjects’ under-reporting maternal left-handedness or scientists’ erroneous faith in monogamous heterosexuality.

Controversy and clashes

The differences in content between McManus’s and Annett’s models were not substantial enough to explain why Annett was the main target of critique, against which she defended herself fiercely (Annett, 1996a; Caverni, 1995; Della Sala and Grafman, 2004; McManus, Shergill, and Bryden, 1993). As the debate surrounding handedness genes was gearing up, historians of science had begun to endorse the idea that scientific controversies were not primarily intellectual disputes but social negotiations. Some scholars compared scientific debates to court proceedings: ‘A controversy is like a trial, where the accuser and the defendant are engaged in a dispute in front of a judge or jury’ (Machamer, Pera, and Baltas, 2000: 11). Integral to the rhetoric of such a trial before the scientific community were ‘attacking a view and defending it’ (ibid.: 12), and the merit of a scientific theory seldom determined the success of such attacks or defences. More often than not, scientists settled conflicts according to implicit criteria and hidden assumptions (Baltas, 2000) as well as political, social, cultural, moral, and economic power structures (Freudenthal, 1998; Mamiani, 2000; Mendelsohn, 1987; Shapin, 1975).

In line with the idea of a scientific controversy as a trial in which sociocultural factors hold at least as much weight as intellectual aspects, the success of Annett’s and McManus’s theories depended on their respective credibility as scientists. Being a woman was only the most obvious characteristic that put Annett at a disadvantage here. Squire (1989) argued that women psychologists remained stigmatized in the late 20th century, even if they acted in line with ‘traditional psychological priorities’ (ibid.: 14), which at the time were shaped by ethnocentrism, androcentrism, sexism, elitism, and positivism (Sherif, 1992). According to Squire (1989: 14), male psychologists saw ambitious female psychologists as ‘double deviant[s]’: women (hence ‘deviant’) who refused to accept the professional limitations of their sex/gender (hence ‘double deviant’).

The conflict between Annett and McManus was also heavily influenced by differences in age, education, and institutional affiliation. Annett made this view explicit in our interview. When asked if she believed that some of the reproaches against her theory had something to do with gender roles in academia, she responded affirmatively and highlighted the intersecting dimensions of age and education: ‘Well, I do! I think it’s implicit. Hardly anybody would admit to it exactly, but if you say: “This is this woman who has got this great theory, but here’s this young man who’s got a medical degree and he’s got a theory”—you got to believe him!’ Annett was referring to McManus in this statement, although she refused to mention his name. ‘You know whom I’m talking about, really’, she said. ¹⁵

McManus, too, seemed keenly aware that more than professional disagreement stood between the two psychologists: ‘She and I are great enemies, traditionally.…Our personal relationship is, as far as she’s concerned: I’m her enemy. As far as I’m concerned: I don’t do that sort of stuff. But she really does not like me’. ¹⁶ After Annett’s passing, McManus reiterated his point that ‘that sort of stuff’ was not for him. He drew a clear line between professional disagreements and one’s personal opinion of another, and he emphasized that, according to his memory, there had been no personal hostility between him and Annett. ¹⁷

The mildness with which McManus retrospectively judged his dispute with Annett may have been due to Annett’s passing and the time that had elapsed since their fiercest exchanges. It may just as well testify to McManus’s significant surplus of social capital. It is conceivable that Annett saw an enemy in McManus because his theory and persona endangered her professional recognition, and that McManus did not share this perspective because ‘women are not a threat’ (Morawski and Steele, 1991: 120). In McManus’s view, the ‘heat’ of the argument had intellectual, not social, results and contributed to ‘forcing clearer theoretical analyses and statements of the models’ (McManus, 2019: 3).

The following anecdote provides evidence for the reading that Annett was no danger to McManus, which is why he could afford to not see an adversary in her. ¹⁸ When McManus found out that Annett had died, he was clearing out his office and came across a letter that Annett had written in 1983. The letter was addressed to the editor of a journal to which McManus had submitted a paper for publication. It contained Annett’s negative assessment of McManus’s work, which Annett had apparently deemed necessary to bring to the editor’s attention. Although Annett had indicated to the editor that the content of her letter was not to be shared with McManus, the editor forwarded the entire letter to McManus.

This instance of peers’ loyalty to McManus at Annett’s expense resonates with subsequent antagonistic exchanges between Annett and her colleagues in the published literature as well as with Annett’s impression that her work and intentions were ‘always, always’ misunderstood. ¹⁹ Moreover, in support of the trial metaphor, the anecdote illustrates that ‘the winning party’ in a scientific controversy is not necessarily the one with the better theory, but ‘the one that has resisted most if not all attacks’ and thereby convinced fellow scientists of its pre-eminence (Machamer, Pera, and Baltas, 2000: 13).

The two psychologists focused their published debates on methodological and conceptual issues, but the personal animosity with its sexed/gendered as well as age-, education-, and status-related dimensions was clearly noticeable in the choice of words and metaphors. For example, McManus repeatedly criticized Annett on mathematical grounds. Throughout his career, he claimed authority when it came to statistical questions. Mostly ‘self-taught’ in the matter, he taught statistics courses and published short notes correcting various misunderstandings of statistical concepts (McManus, 1987, 1998). ²⁰ In one of his publications, he criticized Annett for her choice of statistical methods to determine the RS theory’s level of fit with the data (McManus, 1985b). In the same paper, he reproached Annett for her critique that he should have derived his parameters algorithmically rather than heuristically. McManus (1985b: 32) ‘f[ou]nd that difference obscure and irrelevant’ and suggested that Annett unnecessarily distinguished between ‘statistical nit-picking’ (algorithmically) and pragmatically settling on a ‘method [that] happens to work’ (heuristically). Although McManus did not make this explicit, his emphasis on heuristic over algorithmic methods was also a nod to recently formulated best practices for psychological computing (Apter, 1973). Mobilizing his up-to-date education, McManus (1985b) attempted to close the argument by suggesting that his methods were in line with current statistical practice.

Annett did not claim to be using the latest statistical tests, and she was unwilling to subordinate her work to statistical methods. She tried to fend off McManus’s critique with the following quote from a statistical textbook (Edwards, 1972): ‘The scientist must be the judge of his own hypotheses, not the statistician’ (Annett, 1985b: 28; original emphasis). She defended her calculations by insisting that she and her colleague had followed ‘the traditional’ way of using statistics within psychology (ibid.; emphasis added). McManus would not let that count. He demanded repeatedly that Annett explain precisely what kind of data would allow for the rejection of her model and thereby prove the RS theory’s falsifiability. Otherwise, McManus (1995: 573) wrote, ‘there is a nagging worry that any data can somehow be explained’—a charge that (un)surprisingly nobody levied against McManus when he played down the lack of statistical evidence for the superiority of his model over Trankell’s (1955), as I lay out below.

McManus also questioned the empirical and statistical rigidity of Annett’s work. He claimed that Annett’s theory was speculative and that ‘the waving of arms(,) and the imprecisions of words are no substitute for formal quantitative analysis and precise calculation’ (McManus, 1995: 565). Although both scientists depended on hypothetical alleles as their chief explanatory mechanism, not on biochemically confirmed genes, McManus (1995, 1997) mobilized intersectional stereotypes in his critiques of Annett, drawing a picture of her as an old(-fashioned), emotional, second-rate scientist. This was in line with the stereotype of women as ‘irrational’ (Squire, 1989: 13) and ‘intuitive’ (Weisstein, 1993: 207) as well as the contemporaneous psychological view that ‘the woman’s life goes into decline after approximately age 40’ (Gergen, 1990: 471).

In their last public exchange, McManus countered the final version of Annett’s theory with patronizing impertinence. Setting himself up as a heroic, tireless (and male) Hercules against Annett’s (female) onslaught, he referred to the RS theory and its modifications as heads of the ‘Hydra’ that he had been fighting for decades (McManus, 2004). Annett (2004: 147) responded that she ‘admire[d] the nerve and the mathematical games’ of McManus, but she questioned if his work was ‘sensible’. She closed the debate with the following line: ‘Words suitable for academic discourse fail me’ (ibid.: 149).

Hierarchical orders and gendered origin stories

Annett’s consecutive amendments to her theory reinforced traditional hierarchies between women and men, humans and non-humans, and individuals of different abilities. This inadvertent hierarchization resonates with similar cases of 20th-century female scientists who actively participated in stabilizing systems that served to disenfranchise themselves and other minorities (Rossiter, 1982). Social scientist Corinne Squire provided potential reasons why. She suggested that female psychologists often ‘stay[ed] close to the male-oriented tradition’ in order to establish their scientific ‘rationality’ and avoid the contemporaneous stereotype of being ‘irrational’ or ‘intuitive’ (Squire, 1989: 13, 27). Consequently, numerous women psychologists subjugated themselves to existing standards and frequently found it necessary to ‘“sell out” to methodologies that [they] criticize[d] in other contexts’ (Unger, 1998: 42). Annett never gave the impression that she felt pushed to use methods that disagreed with her understanding of acceptable research, but she used binary classifications that were part of a system that discriminated against her. ²¹ Additionally, she struggled to be taken seriously as a scientist even though she drew on both well-accepted statistical and relatively novel computerized methods.

Furthermore, Squire (1989: 13) suggested that female psychologists’ adoption of the traditional psychological culture led them to reproduce biased concepts and discriminatory structures. Psychologist Rhoda Unger and colleagues (Unger, Sheese, and Main, 2010) have attributed 20th-century women psychologists’ adherence to the status quo to internalized sexism (and the lack of reflection thereon): a learned behaviour and way of thinking through which oppressed women perpetuate their own oppression and that of other marginalized individuals. The theory of internalized oppression, of which internalized sexism is only one aspect (David, 2014), explains the discriminatory aspects of Annett’s theory as a habitual reproduction of the pervasive hierarchies between male/female and able/disabled that were integral to British psychology at the time.

Annett’s proposal of genetic risks resonated well with the intellectual environment at the time. According to sociologist Nikolas Rose (1996), the risk discourse was a characteristic of late 20th-century ‘“advanced” liberal democracies’. Rose (2007) further argued that the contemporaneous molecular view of the human body and mind changed the interaction between authorities and the people. The perfection of one’s health and abilities in the interest of productivity were the markers of the new biomedicalized politics. Rose and historian of science Joelle Abi-Rached did not condemn these developments. They suggest that ‘brains at risk’ (of antisocial behaviour or psychiatric disorders) might be controlled in the interest of a peaceful society (Rose and Abi-Rached, 2013: 197).

Other scholars have viewed the risk discourse with less optimism than Rose and Abi-Rached. Sociologist Dorothy Nelkin, for instance, critically assessed the testing culture in genetics and neuroscience of the late 20th century. She argued that genetic testing had created a group of ‘pre-symptomatic ill’ people, and that these individuals were in danger of becoming a ‘genetic underclass’ (Nelkin, 1992: 189–90). Sociologist Troy Duster (2003) similarly argued that even benevolent recourses to risk and susceptibility opened the ‘backdoor’ to neo-eugenic practices. Annett’s suggestion to screen for the RS factor defied these cautious warnings (Annett, 1996b, 2011; Annett and Kilshaw, 1984).

McManus’s and Annett’s theories implied an evolutionary hierarchy among different genotypes or phenotypes, even though McManus (2016) did not support the idea that certain risks and pathologies were associated with particular handedness types. ²² For example, the concept of heterozygote advantage in both models indicated that mixed genotypes were fitter than the homozygous population. Furthermore, both models contained mechanisms to modulate the strength of the laterality gene. Annett claimed that the Right Shift was stronger in singletons and women than in twins and men, and McManus hypothesized that a recessive factor inhibited the expression of the dextral allele D more often in males than in females. These adjustments allowed the psychologists to distinguish between different types of humans along various intertwined categorical lines, including sex/gender, circumstances of birth, and handedness.

Haraway’s article ‘Primatology Is Politics by Other Means’ elucidates the entanglement and hierarchization of classifications. According to Haraway, the narratives of human society and primatology both rest heavily on the delineations between constructed binaries: nature/culture, primitive/civilized, dark/white, female/male. Haraway unraveled the political and cultural importance of primatology for origin stories of so-called Western human societies and the genesis of ‘Man’ (Haraway, 1984: 492). She argued that primatological theories reframe—in an allegedly objective way—the metaphysical tales and moral myths that make up the history of modern civilization.

The science of left and right brains and hands played into the logic of origin stories. The RS theory was a conceptual intervention that separated in hierarchical orders carriers of the RS factor from non-carriers, twins from singletons, and females from males. If only carriers of the RS factor were responsible for the characteristically human shift to the right as laid out above, then non-carriers would show the same distribution as non-human primates, a bell curve that peaks at zero. Even though Annett’s proposal was not necessarily a proclamation of the pathological nature of left-handedness, it certainly hinted at a dehumanization of non-carriers of the RS factor. Moreover, if the RS factor were expressed less strongly in twins than in singletons, then twins would be, however gently, placed on a lower level of an imagined mammalian hierarchy than singletons. Likewise, if the shift were stronger in women than in men, then the latter would be closer to non-human animals on a metaphorical evolutionary tree.

Similarly, the McManus–Bryden model suggested that D caused fewer males than females to be right-handed, although both populations had the same ratio of D carriers. The directive force of the D allele would be overwritten more often in males than in females because males were purportedly more susceptible to the inhibiting allele. Hence, handedness in the male population would be less often genetically determined and more susceptible to environmental influences. The logical consequence would be that women are less malleable than men, at least regarding laterality.

McManus’s implication that women were less malleable than men was as much a reversal of certain classical gender stereotypes (Malane, 2005; Matlock, 1994; Russett, 1991) as Annett’s implication that men, by virtue of experiencing a smaller Right Shift, were closer to nature than women (Jordanova, 1989; Ortner, 1972; Park, 2004; Richards, 2004). Annett’s suggestion that women’s speech centres were more strongly left-lateralized than men’s also defied common belief (Hirnstein, Hugdahl, and Hausmann, 2019; McGlone, 1980). Neither McManus nor Annett blindly followed stereotypical prescriptions for Male and Female, and both theories had notable progressive implications. Nonetheless, Annett’s suggestion of the smaller Right Shift in men was not as revolutionary as it might seem, since it perfectly aligned with the notion of Male as the standard, the norm, or the unmarked category against which Female was seen to be deficient. Furthermore, Annett assumed that women’s strong left-lateralization for speech resulted in superior verbal skills, an idea that resonated with the stereotype of the chatty woman. Similarly, McManus’s idea that female handedness was less malleable implied that women’s hand characteristics were more strongly genetically determined than men’s. This view is reminiscent of the old doctrine that males as a group are particularly variable, encompassing the good and bad extremes, while females are a more homogenous group, or average (Shields, 1982).

Even the partial history of genetic handedness concepts presented here resists a clear mapping of right/left onto male/female, culture/nature, or any other binary classification construct. As Haraway (1984) observed, binary classifications do not erect static hierarchies. They can be combined in many ways to create fresh iterations of traditional origin stories. The categories and their hierarchical connotations are flexible in their application, but crucial as load bearers in the construction of a natural world order.

Computerized statistics as a technology of (dis)empowerment

The decision to propose genes for handedness and brainedness based on statistics alone is well in line with the overwhelming retreat to numbers that has been common practice in the human and life sciences for more than a century. Philosopher of science Ian Hacking (1990, 2006) has traced the unprecedented epistemic supremacy of statistics in the human sciences since the 19th century, and historian of science Theodore Porter (2003: 248) has suggested that psychology was ‘perhaps the most enthusiastically statistical of all the social or human disciplines’. What set Annett and McManus apart from their predecessors was the employment of electronic computers.

Annett’s push into the (at the time) male domain of electronic computers seemingly did not confer trustworthiness upon her research. Entire journal issues were devoted to critiques of her work (Caverni, 1995; Della Sala and Grafman, 2004), whereas the success of McManus’s utilization of computerized statistics became apparent in the overwhelming absence of published pushbacks against his work. One reason why the use of computers did not endow Annett’s work with more credibility lies in ‘software’s and computing’s gendered, military history’ (Chun, 2011: 29). Electronic computers were a matter of national security and economic productivity, and computer programming, which had been considered women’s work before, became an increasingly male domain from the 1960s onwards (Abbate, 2012; Hicks, 2017; Misa, 2010; Morley and McDonnell, 2015). Hand in hand with the masculinization of the programmer, digital technologies and code itself became endowed with an aura of rationality and detachment—that is, masculinity (Adam, 1998; Ensmenger, 2010, 2015).

Media theorist Wendy Chun’s (2011: xii, 2) analysis of computers as ‘mediums of power’ offers an explanation for why the critiques of Annett’s computerized methods further weakened her precarious position in the field instead of transferring the authority of alleged masculine rationality to Annett’s model. Chun (2011: xii, 2, 20) suggested that computers empower users to envision, create, and ‘navigate [their] increasingly complex world’ while simultaneously imposing on them the demands of ‘neoliberal governmental power’; in other words, users are both ‘empower[ed]’ and ‘estrange[d]’. While Chun’s analysis focuses on 21st-century computer users, the idea that computers confer power on users and stabilize existing power dynamics held equally true for Annett and McManus. In Annett’s case, the latter characteristic outweighed the former. She was already marginalized as an older female clinical psychologist without a professorship. Moreover, she had never been educated in the use of computers and did not have immediate access to them, and her colleagues rejected her expertise in the use of masculine-coded machines. In contrast, McManus began using computers while on ‘a long vacation course in pharmacology in Cambridge in 1970’ and honed his skills in using them during his free time, and his credibility as a male MD/PhD benefitted significantly from seemingly rational and detached computational analyses. ²³

Annett was not particularly computer-savvy, but computers promised to facilitate her statistical calculations. ²⁴ Towards the beginning of her career, she carried out calculations by hand, at home, often over the course of several days. For the RS theory, Annett used computers to operationalize the most laborious calculations with help from computer scientists and her son. For her most-cited paper, Annett (1970) used a computerized association analysis to hone the distinction between the three presumed handedness phenotypes and give them empirical weight. Like other human scientists in the 1960s (November, 2012), Annett made significant epistemological compromises in order to employ a computer. The association analysis was meant ‘to demonstrate that hand preference is distributed continuously and not discretely’ (Annett, 1970: 303). Nevertheless, in the process of translating her concepts into computer logic, Annett disregarded her conviction that handedness was a non-binary category (left-, mixed-, and right-handedness) and organized her data in two groups. M. J. Norman of the University Sub-Department of Computation at the University of Hull collaborated with Annett on the association analysis, but ‘he made no decisions’ regarding the variables or interpretation of results. ²⁵ Annett defined the analytical categories and Norman wrote the code. At first, Annett experimented with the possible handedness distinctions of left versus not left, right versus not right, and either versus right or left. She eventually settled for the division between left and not left because it provided her with the most statistically significant associations (Annett, 1970).

Inconsistent classifications of hand preference permeated Annett’s work for years to come. Annett did not see a problem in using a binary categorization to advance her argument of the existence of a mixed-handed phenotype, let alone the continuous distribution of manual preference. However, several of Annett’s colleagues suggested that her switching between dichotomous and trifold handedness classifications undermined the credibility of her work, and that she should decide on either of the two frameworks (Caverni, 1995). Annett changed neither her theory nor her code. She replied to her critics: ‘Oh yes I can [have it both ways], if that is what the empirical findings demand’ (Annett, 1995a: 636).

The complexity of McManus’s calculations required digital methods from the start, and McManus did not stop at verifying his model through the data but successively adapted his model to fit the data. Whereas many psychologists hesitated to embrace computers in the 1970s and 1980s (Elwork and Gutkin, 1985), McManus started using punched cards during his undergraduate studies (McManus, 1979). To formulate his genetic model of handedness, McManus (ibid.: 249–54, 273–85) depended on ‘corrections’ of the data (for instance, to offset response bias or measurement error) and on successively ‘fitting’ the model to the data. In non-technical terms, he smoothed the data set, then guessed the parameters of his model, and finally tested them against the preprocessed data. McManus relied on current statistical conventions for calculating the model’s goodness of fit with the data. If the discrepancy between the predictions and the actual data was too large, he modified his model and tested again.

McManus’s trust in computer-generated numbers had its limits. For the first version of his model, McManus (1979) used chi-squared tests to illustrate the superiority of his model over other theories of laterality, but the statistical tests did not single out McManus’s formula as the superior one. A much older Mendelian model (Trankell, 1955) remained in the competition for the best fit. McManus ascribed this ambivalence to a lack of data, not to the inappropriateness of his theory. Only after the statistical methods had failed to provide sufficient evidence for the superiority of his own theory, McManus (1979) claimed that his theory was superior because of its more realistic content: it contained a chance element and left room for environmental influences, which the competing model did not.

Considering McManus’s willingness to weight theoretical considerations more heavily than statistical tests, it may seem surprising that McManus (1979) charged Annett with defining her parameters based on theoretical predictions instead of empirical observations. Anticipating critiques of his own procedure, McManus emphasized that his modifications were not arbitrary. He maintained that some models, including immature versions of his own theory, did not fit the corrected data. Hence, McManus argued, his measures for goodness of fit were in fact decisive. Moreover, when he tested his model on faulty data sets, the statistical measures indicated poor fit. McManus (1985a) took this as additional support for his model, which he had designed to fit supposedly correct, not faulty, data sets.

While McManus did not entirely rely on computers to identify the best laterality model, he believed that these machines could reveal true hand preference, which might be hidden underneath the surface of cultural pressures for left-handers to use their right hands. McManus (1979) used computerized corrections to account for this right-bias in the data. Several years later, he adjusted the original parameters and determined 7.75% as ‘the true incidence of left-handedness’, as opposed to the—supposedly smaller—incidence of ‘manifest left-handedness’ (McManus, 1985a: 7, 16).

McManus effectively combined traditional and cutting-edge technologies to gain trust for his work on two different levels. By employing well-accepted statistical tests and operationalizing them on a computer, McManus united the authoritative potential of the traditional social-mathematical and the novel formalized-mechanized proving cultures (MacKenzie, 2001). He personified the well-educated white male scientist who knew his statistics, while the computer seemed to innovatively fit the model for him, allegedly objective, visualized on colourful printouts, and empirically testable. McManus’s data-driven approach was the epitome of psychologists’ using the computer to ‘do(ing) new things, perhaps some of those very things that “classic” science either could not or would not do’ (Ashby, 1973: viii). According to this definition, Annett was a ‘classic’ and old(-fashioned) scientist who could or would, for lack of education and resources, not keep up.

Conclusion

Annett did not give up hope that her theory would be experimentally confirmed. ²⁶ It seems unlikely that anyone will try to do so, not least because Annett’s professional network deteriorated long before she stopped publishing in 2011, and she did not keep any records of her research out of concern that her children would have to ‘clear after [her]’. ²⁷ McManus may remain a formative force in laterality research for a while. He is slowly winding down his successful career while still teaching, publishing, editing, and archiving his most important papers. ²⁸

Despite his professional accomplishments, McManus did not emerge as the unreserved victor from the controversy surrounding handedness genes. One century after the popularization of the first well-known Mendelian model of handedness (Jordan, 1911), the computer, the very tool that made Annett’s and McManus’s theories possible, shattered them. McManus and colleagues’ large-scale computer-based analysis of thousands of genetic twin data sets determined that there was no handedness gene (Armour, Davison, and McManus, 2014; McManus, Davison, and Armour, 2013). Handedness could be caused by a stable polymorphism, or various handedness genes may run in specific families, but apparently no single gene can explain the data.

Likewise, other researchers have abandoned the idea that a simple genetic mechanism can explain the inheritance of manual preference (de Kovel and Francks, 2019; Ocklenburg et al., 2017; Wiberg et al., 2018). Recently, neuroscientists have suggested that manual preference may rely on epigenetic mechanisms (Schmitz et al., 2017). This shift towards epigenetics, a field that only superficially breaks with genetic determinism and biological reductionism (Richardson, 2015; Wastell and White, 2017), may solidify the alleged innocuousness of handedness as a category while opening new avenues for the association of manual preference with questions of sex/gender, sexuality, and dis/ability.

Although most industrialized societies have abandoned forced right-hand usage, the intersection between handedness and social hierarchies extends far beyond Annett’s and McManus’s research (Bertrand, 2001; Cornel, 2019; Kushner, 2017). The history of handedness research serves as a reminder that scientific classification systems mirror the social experience of researchers, and that the reach of these classifications extends far beyond the scientific laboratory or (home) office. Most likely, new scientific frameworks and more computerized data analyses will neither do away with the idea of the deviant left-hander nor make science less biased. In other scientific fields, such persistent hierarchization and stereotyping in the absence of a reliable aetiological theory would elicit massive controversy. Only the seeming neutrality of manual preference has prevented a widespread awareness of the political clout of handedness research and the acceptance of the fact that all classification systems can reproduce hierarchical social structures and impede social justice.