Sexual-Orientation-Related Differences in Verbal Fluency Qazi Rahman, Sharon Abrahams, and Glenn D. Wilson Institute of Psychiatry, University of London This study examined the performance of 60 heterosexual men, 60 gay men, 60 heterosexual women, and 60 lesbians on 3 tests of verbal fluency known to show gender differences: letter, category, and synonym fluency. Gay men and lesbians showed opposite-sex shifts in their profile of scores. For letter fluency, gay men outperformed all other groups; lesbians showed the lowest scores. For category fluency, gay men and heterosexual women jointly outper- formed lesbians and heterosexual men. Finally, gay men outperformed all other groups on synonym fluency, whereas lesbians and heterosexual men performed similarly. A difference between heterosexual men and women was demonstrated on category and synonym fluency only. The findings implicate within-sex differences in the functioning of the prefrontal and temporal cortices. Sex differences in certain cognitive functions are well documented. Men are found to excel on certain tests of mental rotation, spatial perception, and mathematical prob- lem solving, whereas women perform better on certain tests of verbal fluency, perceptual speed, object location mem- ory, and facial emotion processing (e.g., Acevedo et al., 2000; Herlitz, Nilsson, & Backman, 1997; Kimura, 1999; McClure, 2000; Monsch et al., 1992; Voyer, Voyer, & Bryden, 1995). The widespread belief that women are better at “verbal abilities” arises from observations of differences between male and female children. Compared with boys, girls articulate and speak earlier, possess larger vocabular- ies, and are superior at spelling and reading (Kimura, 1999). These observations may, in turn, reinforce culturally depen- dent gender schemas—that is, preconceptions about stereo- typical sex differences, regarding overall female linguistic superiority. There are no substantial adult sex differences in verbal IQ or vocabulary (although standard IQ tests are usually designed to eliminate sex differences; Kimura, 1999). However, one female-favoring verbal function ex- tends through adolescence and to adulthood—verbal flu- ency. Verbal fluency tests require participants to produce the maximum number of words starting with a particular letter, or from a category of some kind, under a time constraint. Whereas the sex differences in favor of adult men on tests of spatial rotation are typically large (Cohen’s d = 0.9), those favoring women on letter fluency are small (d = 0.3), and those on category fluency are modest (d = 0.5; Acevedo et al., 2000). However, Hines (1990) reported a particularly large difference in favor of women on a measure of synonym fluency (d = 1.2), a form of category fluency (in which words are produced according to semantic constraints). There is also growing, yet conflicting, evidence for the influence of sexual orientation on gender differences in sexually dimorphic cognitive abilities. Initial reports indi- cated female-like performance by gay men on mental rota- tions, a typically male-favoring spatial task (Gladue, Beatty, Larson, & Staton, 1990; McCormick & Witelson, 1991; Sanders & Ross-Field, 1987), although other work has failed to replicate them (Gladue & Bailey, 1995). Three further studies have replicated the gender-atypical perfor- mance in mental rotation for gay men (Neave, Menaged, & Weightman, 1999; Rahman & Wilson, 2003; Wegesin, 1998). The most recent of these also found a large cross-sex shift in gay men’s performance on the Judgement of Line Orientation Test (Benton, Hamsher, Varney, & Spreen, 1983) compared with heterosexual men, whereas no differ- ences were found between lesbians and heterosexual women (Rahman & Wilson, 2003). Few studies have examined differences in the sexually dimorphic domain of verbal fluency per se (female favor- ing), and the results have been conflicting. In a widely cited study, McCormick and Witelson (1991) demonstrated fe- malelike performance on category fluency in gay men (whose scores did not differ from those of heterosexual women), whereas Neave et al. (1999) failed to find any gender-atypical performance on letter fluency. Both these studies used single-item fluency tests: The former used the category Animals; the latter used the letter L. Neave et al. did, however, report gender-atypical performance by gay men on a synonym fluency test, which these authors referred to as verbal associations (using four stimulus words— unspecified in the report). Gladue et al. (1990) used three letters (F, A, S) and also reported no sexual orientation effects. Two reports claimed to demonstrate gender-atypical performance in gay men in nonfluency verbal skills. Sanders and Wright (1997) reported that Verbal IQ (of the Wechsler Adult Intelligence Scale; Wechsler, 1958) scores of gay Qazi Rahman, Sharon Abrahams, and Glenn D. Wilson, Depart- ment of Psychology, Institute of Psychiatry, University of London, London, England. This work was supported by grants from the Medical Research Council of the United Kingdom to Qazi Rahman and from the British Academy to Glenn D. Wilson. Correspondence concerning this article should be addressed to Qazi Rahman, Department of Psychology, Institute of Psychiatry, University of London, De Crespigny Park, London SE5 8AF, England. E-mail: q.rahman@iop.kcl.ac.uk Neuropsychology Copyright 2003 by the American Psychological Association, Inc. 2003, Vol. 17, No. 2, 240 –246 0894-4105/03/$12.00 DOI: 10.1037/0894-4105.17.2.240 240