INTRODUCTION
Brain Organization and Sexuality
The complex interaction of the social environment and the biological substrate in the sexual development of the individual has provoked intense debate. In recent times, the focus of the debate has moved closer to the biased interaction theory (BIT) of psychosexual development.
Essentially, this theory proposes that adult “behavior is a composite of prenatal and postnatal influences with the postnatal factors superimposed on a definite inherently biased sexuality” (Diamond, 1965, p.169; Diamond, 2006).
This view of human sexuality rests on the supporting data from animal studies, ranging from the classic research on the administration of testosterone to pregnant female guinea pigs (Phoenix, Goy, Gerall, & Young, 1959), to the work of Beach (1976) on the induced masculinization or feminization of animals, to the work of Goy, Bercovitch, and McBrair (1988) on the masculinization of female primates and Gorski’s (1991) research on sexuality in rats. Much of this research has been evaluated by Morris, Jordan, and Breedlove (2004) who conclude that the interaction between hormones and genes should now be the focus of research.
Research into the sexing of human brains has also been carried out and comes from diverse sources. The best known studies are probably those of females with congenital adrenal hyperplasia (CAH) by Berenbaum (1990) and Hines (2004), the anatomical studies of brains of androphilic males (LeVay, 1991), and those of the brains of transsexual people (Kruijver, 2000; Swaab, 2004). Such studies have begun to clarify some of the complex biological underpinnings of human sexuality and have been complemented by many reports of cognitive function that have teased out some of the psychological differences between the sexes (Halpern, 2000; Kimura, 1999).
Potentially, one of the most fruitful sources of insight into sexual development is the phenomenon known as transsexuality, whereby the individual “knows” he or she is better suited to life in the gender opposite to that of birth assignment and usually transitions gender to effect such a life. The etiology of the transsexual condition is a much disputed issue and has, in the past, often been treated as resulting from some unidentified process in the individual’s personal social development. But current refinements of BIT have brought this phenomenon into the fold of empirical research. Understanding such apparently intractable conditions may provide general insights into sexual development. One possible way forward is to identify known sex-differentiated brain functions such as verbal processing to try to determine whether they co-vary with gender or body-sex.
Dichotic Listening and Brain Organization
Kimura (1961) published the results of two studies of dichotic listening and language lateralization. Wada and Rasmussen (1960), using sodium amytal, a fast-acting anesthetic, had earlier determined the hemisphere to which language appeared to be lateralized for a group of epileptic patients about to undergo surgery. Kimura showed that the dichotic pattern produced by these clinical patients was related to their known pattern of language lateralization. If language was lateralized to the left hemisphere, the patient reported more stimuli being presented to the right ear, producing the so-called right-ear advantage (REA). If language was lateralized to the right hemisphere, however, more stimuli were reported from the left ear, producing a left-ear advantage (LEA).
Kimura confirmed this pattern using a non-clinical group and later published the seminal model upon which most of the dichotic research over the following 40 years has been based (Kimura, 1967). The model suggests that the contralateral pathways connecting the ears with the cerebral hemispheres dominate the ipsilateral pathways in dichotic stimulation. This produces the characteristic right-ear advantage because the stimuli presented to the right ear go more directly to the left hemisphere in which language is usually represented. Stimuli presented to the left ear arrive at the left hemisphere via the right hemisphere following a longer neural route and, thus, suffer a time penalty since an increased number of synapses are involved.
Sex differences were reported early in the history of dichotic research (Bryden, 1973; Lake & Bryden, 1976) and were robust if consonant-vowel stimuli were used in a divided attention paradigm (Hugdahl, 1988). The early studies indicated that males were more likely to produce REAs, and females, a more balanced dichotic pattern or even a left-ear advantage. Later, it became apparent that other group characteristics modified the sex difference such that participants in statistically sex-atypical occupations (e.g., a female computer programmer or a male primary school teacher) produced opposite-sex-like patterns in dichotic tasks (Govier & Bobby, 1994). Thus, males in female occupations produced lower overall mean REAs and females in male occupations produced higher overall mean REAs. With these results, Govier (2003) argued that the pattern of dichotic responses was an indication of sex-differentiated brain organization.
The experimental methods necessary to sharpen the sex difference gradually evolved, and Table 1 below indicates the range of lateralization indices reported from four illustrative studies that used stimuli and methods similar to the study reported here.
TABLE 1. Lateralization Indices from Four Previous Dichotic Studies
| Lateralization Index | |||
Study |
Stimuli | Male | Female |
Lake & Bryden, 1976 |
60 pairs | .07 | .03 |
Munro & Govier, 1993 |
First 40 pairs | .17 | .06 |
Govier & Bobby, 1994 |
100 pairs | .35 | .10 |
Rahman et al., 2008 |
60 pairs | .28 | .11 |
The strong evidence that performance on dichotic listening (DL) tasks is based on brain organization and that tests of DL produce sex differences raises the possibility that examination of such auditory abilities may be useful in shedding light on brain organization in persons with the transsexual condition.
Cohen and Forget (1995) used a dichotic paradigm to investigate laterality in male-to-female (MtF) transpeople. For their verbal dichotic task, they did not find a difference between those participants who had transitioned from male to female and their typical female controls. However, as they also did not find a sex difference between their typical male and female controls, there is a strong possibility that their experimental method did not meet the required low sound-sensitivity levels. Moreover, all of their participants were right-handed, as measured by the Edinburgh Handedness Inventory (EHI; Oldfield, 1971). This also may have militated against finding a verbal stimulus effect. The researchers did, however, report a sex difference on their nonverbal melodies task that produced a left-ear advantage in the male control group but a right-ear advantage in the MtF group. There was also a difference between typical males and transpeople in the tones task. There were no significant differences between transpersons and typical women in the non-verbal tasks. Cohen and Forget interpreted the non-verbal data as support for the assumption that hormone treatment produced an activation effect that shifted the nonverbal dichotic pattern of their trans-participants in a more female direction. This may not have been the case, however, since the pattern of their data may have been the result of a more female organization in some brain structures before hormone treatment.
A more recent study of functional cerebral lateralization in MtF transsexuals by Wisniewski, Prendeville, and Dobs (2005) found no group differences on their tests of lateralization using tasks of a visual split-field. Their sex-differentiating cognitive tasks (identical pictures, mental rotation, and building memory tests) showed no significant group differences. Neither did they find any relationship between hormone profile and performance on these tasks. However, the researchers did not use a control group of females to demonstrate a sex difference on their visual split-field task. It may be that this task is not a sensitive indicator of sex differences in lateralization.
The study reported here used a well-tried dichotic listening task that, in the past, had produced large sex differences in verbal processing in the typical population. We considered it sensitive enough to investigate functional brain organization in male and female participants identifying as transmen and transwomen.
Handedness Effects
Handedness has been shown to be a moderating factor in dichotic listening. In general, left-handed participants are more likely to produce an LEA than right-handers. More than 80% of male right-handers but fewer than 65% of male left-handers produce an REA (Bryden, 1988). It is assumed that this is because handedness is related to speech lateralization (Segalowitz & Bryden, 1983). Handedness was also important in the present study because a raised incidence of left-handedness has been reported in MtF and female-to-male (FtM) individuals by Orlebeke, Boomsma, Gooren, Verschoor, and van den Bree (1992) and Cohen-Kettenis, van Goozen, Doorn, and Gooren (1998) and has been confirmed in MtF individuals (Green & Young, 2001; Watson & Coren, 1992). A more recent study by Wisniewski et al. (2005) has also provided supporting evidence for the relationship between nonexclusive right-handedness and the MtF condition, although their difference was “marginal.” However, a large study by Zucker, Beaulieu, Bradley, Grimshaw, and Wilcox (2001) of 205 boys with gender identity disorder found that they were significantly more likely to be left-handed than a control group of clinically referred boys who had been referred for other reasons. The group with gender identity disorder but not the control group also had a higher rate of left-handedness than three groups of non-referred boys. Thus, the relationship between non-right-handedness and the MtF condition has, in the main, been well established. There are also data to indicate that handedness may be influenced early in prenatal life. In women whose mothers had been treated with diethylstilboestrol (DES) before week 9 of gestation, hand preference shifted away from strong right-handedness (Smith & Hines, 2000). Thus, participants’ handedness can be regarded as a second important indicator of brain organization, although the mechanism relating handedness with gender identity is unknown. Further evidence that handedness is somehow related to sexuality stems from the findings of Lalumie`re, Blanchard, and Zucker (2000), who found a correlation between handedness and sexual orientation in men and women and the findings of Green and Young (2001) who found relationships between hand preference and both sexual preference and transsexualism.
Hypotheses
Several hypotheses were proposed:
1. There would be a significant difference in the lateralization scores between the transwomen (MtF) and the typical male comparison group with the participants who had transitioned from male to female producing a reduced REA as typically seen in females.
2. There would be a significant difference in the lateralization scores between the transmen and the typical female comparison group, with the FtM participants producing an increased REA.
3. There would be a higher incidence of non-right-handedness among the MtF participants compared to the typical male population.
4. There would be a higher incidence of non-right-handedness among the FtM participants compared to the typical female population.
METHOD
Participants
A total of 158 participants were tested, of whom 68 were transwomen (M age = 43 years; SD = 13.6 years; range, 19–79) and 26 were transmen (M age = 37 years; SD = 10.6 years; range, 19–59). The control groups of 36 typical males (M age = 37 years; SD = 12.5 years; range, 18–72) and 28 typical females (M age =33 years; SD = 15.2 years; range, 18–68) also completed the dichotic task.
The majority of trans participants were recruited and tested at the London Gender Clinic, a specialist sexual health facility in London. Participants were attending normal appointments at the clinic and were invited to take part in the study after their clinical consultation. The participation rate was in excess of 90% and non-participation was entirely due to time constraints associated with the participants transport arrangements. Others were recruited by and tested at the Pacific Centre for Sex and Society (PCSS), University of Hawaii. Some of the non-trans-participants were recruited in a leisure center in London and tested in their homes, while other participants were students at the University of Leeds. None of the participants were compensated for their participation. The study had institute review board approval and full ethical approval from both the University of Hawaii and the University of East London.
Table 2 below illustrates the profile of the sexual orientation of the transsexual participants.
TABLE 2. Sexual Orientation, Trans Status, and Handedness
| MtF* | FtM | |||||||
| Androphilic | Gynecophilic | Ambiphilic | Aphilic | Gynecophilic | Androphilic | Ambiphilic | Aphilic | |
| N | N | N | N | N | N | N | N | |
Right-handed |
5 | 14 | 15 | 7 | 9 | 4 | 3 | 2 |
Left-handed |
4 | 1 | 4 | 2 | 0 | 0 | 1 | 0 |
Mixed-handed |
2 | 3 | 6 | 0 | 2 | 4 | 1 | 0 |
| *Four MtF right-handed participants and 1 non-right-handed participant declined to provide data on their sexual orientation. | ||||||||
The mean number of years for which the trans-participants had taken gender appropriate hormones was for FtMs, M = 3.08 years, SD =5.18; and for MtFs, M = 3.44 years, SD = 3.55.
MATERIALS
Dichotic Stimuli
The dichotic stimuli comprised pairs of consonant-vowel-consonant (CVC) syllables. Each pair differed by the initial or final consonant (e.g., ged–ped or dod–dop). The stimuli were recorded on to a digital console for processing, such that the onset and finishing asynchronies were minimized. The stimuli were recorded in a female voice in a standard English accent on a compact disc. Playback was via RadioShack or Panasonic stereo earphones on either a Matsui compact disc player (model 103a) or an Emerson disc player (model HD 8116). The stimuli pairs were presented through 6-s intervals in two blocks of 30 pairs. The second block was presented through reversed headphones to obviate channel asymmetries. The stimuli were presented such that they were just clearly audible.
Handedness Questionnaire
Handedness was assessed using EHI (Oldfield, 1971), a 10-item inventory in which each activity elicits hand preference responses for a different activity (left hand, right hand, either hand), as well as the question, Do you ever use the other hand? This secondary part of the inventory was not used in our final analysis as its meaning was found ambiguous in a pilot study.
We slightly modified the procedure for labeling handedness used by Orlebeke et al. (1992). Participants were categorized as right-handed if they indicated that they preferred using their right hand exclusively for 7 or more of the 10 activities. The rest of the participants were classified as non-right-handed. This category included participants who indicated that they preferred using their left hand for 7 or more activities (conventionally left-handed) or who were classically ambidextrous and could use either hand for most tasks or mixed-handed—that is, they carried out some tasks predominately with the right hand and some with the left.
PROCEDURE
Before testing, the procedure was described to each participant and his or her right to withdraw from the testing at any time was emphasized. The participants were also informed that if, at the end of the testing session, they wished to withhold their performance data from the study, it would be destroyed in their presence.
The tasks were administered in a quiet room without interruption. The handedness inventory was presented first, and the procedure for completion was explained. The participant was asked if he or she understood the instructions and any questions were answered. When the participant was ready, the inventory was completed.
The dichotic listening task was then presented. The procedure was explained to each participant, and when they were ready, they were required to place the headphones in a comfortable position on their ears and to gently hold them in place for the duration of the test. The participants were asked to listen to the practice session, which comprised four stimulus pairs that were also presented to them visually on a typed sheet to aid their understanding of the task. They were asked to confirm that the stimuli were just audible and not loud and that they understood the task. The participants were asked to report what they heard: both syllables if possible or just one of the syllables. Participants who could not report anything were instructed to say, “Pass.” They were then asked to shut their eyes and to try to look forward during the stimulus presentation. The auditory stimuli were started and correct responses were recorded on a printed check sheet. Incorrect responses were also noted but not analyzed. A laterality index was computed as follows: (R − L)/(R + L) (Hugdhal, 1988; Rahman, Cockburn, & Govier, 2008; Studdert-Kennedy & Shankweiler, 1970). At the end of the testing session, participants were asked if they had any questions about the procedure and invited to comment on their experience. Finally, all subjects were thanked for their cooperation.
TABLE 3. Mean Lateralization Scores, Correct Left Ear and Right Ear Scores, and Standard Deviations (SD) as a Function of Trans Status
Trans Status |
N | Left Ear Mean | SD | Right Ear Mean | SD | Lateralization Index | SD |
MtF |
67 | 24.49 | 8.26 | 27.07 | 10.18 | 0.03 | 0.25 |
FtM |
26 | 27.38 | 6.91 | 28.27 | 9.82 | 0.00 | 0.22 |
Control males |
36 | 20.03 | 7.23 | 35.25 | 7.99 | 0.27 | 0.24 |
Control females |
28 | 28.39 | 6.56 | 31.36 | 6.73 | 0.05 | 0.19 |
| Note. The total possible score for each ear was 60. Lateralization Index is (R − L)/(R + L). | |||||||
RESULTS
Lateralization
The number of participants in each trans-category and the dichotic listening data are shown in Table 3. For logistical reasons, often relating to time constraints, it was not possible to collect all of the desired data from each participant. The number of participants for whom data were available is shown for each analysis.
Preliminary inspection of the participant sample showed that the MtF group was older than their male control group and that the difference was statistically significant, t = 2.17, df = 93, p = .03. Age data were not collected from 7 MtF and 2 control male participants. The FtM group did not differ significantly in age from their female controls, t = .87, df = 51, p = .39. Age information was not collected from 1 FtM participant. The data were analyzed using an ANOVA with the factor Trans Status at 4 levels: MtF, FtM, control males, control females. In order to account for possible confounding effects of age, a second factor, Age, was entered as a covariate.
This gave Trans, F = 10.45, (df = 3, 142), p <.001. For the covariate, Age, F = .35, (df =1, 142), p = .55. Thus, age did not significantly covary with laterality. The overall observed power of the test was .99. The effect size of the Trans Status factor was .184 (partial eta squared), accounting for approximately 18.4% of the variance in the data. Pair-wise tests were carried out with the Bonferroni adjustment for multiple comparisons. The effect size d was also computed for each of the comparisons. These analyses are shown in Table 4.
TABLE 4. Multiple Comparisons (Bonferroni Adjusted) and Effect Sizes
Comparison Group |
.05 Significance Level | Effect Size d |
Control males vs. Control females |
Sig | 0.88 |
MtF vs. Control males |
Sig | 0.89 |
FtM vs. Control females |
ns | 0.22 |
FtM vs. MtF |
ns | 0.10 |
| Note. The effect size d is calculated as (Mean 1 − Mean 2)/Pooled standard deviation. | ||
The overall analysis indicated that trans status was a moderately small but important variable affecting dichotic performance. The pair-wise comparison between the male and female control groups showed that the dichotic test produced a significant sex difference, with males being more lateralized. This comparison was associated with a large effect size. The MtF trans-group was significantly different from the male controls, also with a large effect size. The comparison between the FtM group and their female controls, however, was not significantly different and the effect size was small. The FtM group was not significantly different from the MtF group; the effect size was small.
In order to test whether these patterns of lateralization were a function of handedness rather than trans status, the analysis was repeated for right-handed participants only. The means and standard deviations for this sample are shown in Table 5.
TABLE 5. Mean Lateralization, Correct Left Ear Scores, Right Ear Scores, and Standard Deviations as a Function of Trans Status for Right-Handers
Trans Status |
N | Left Ear Mean | SD | Right Ear Mean | SD | Lat Mean | SD |
MtF |
45 | 23.69 | 7.41 | 26.13 | 9.33 | .03 | .27 |
FtM |
18 | 25.28 | 5.68 | 28.33 | 9.65 | .04 | .23 |
Control males |
34 | 19.76 | 7.31 | 35.50 | 8.12 | .28 | .25 |
Control females |
26 | 27.77 | 6.20 | 31.46 | 6.84 | .06 | .19 |
This analysis gave Trans, F = 7.94, df = 3, 109, p = <.001; partial eta squared = .179, power = .99; Age, F = .42, df = 1, 109, p =.52; partial eta squared = .004, power = .10 (see Table 6).
TABLE 6. Multiple Comparisons (Bonferroni Adjusted) and Effect Sizes
Comparison Group |
Significance Level .001 | Effect Size |
Control males vs. Control females |
Sig | .88 |
MtF vs. Control males |
Sig | .86 |
FtM vs. Control females |
ns | .10 |
FtM vs. MtF |
ns | .04 |
Focusing only on the right-handed participants has not altered the pattern of the results of the statistical analysis. Trans status was a significant factor in dichotic performance; there was a significant difference between the male and female control groups with males showing a larger REA. The MtF group was significantly less lateralized than the male controls, but the FtM and their female controls did not differ significantly. The lateralization of the MtF and FtM groups was not significantly different.
Hormone Treatment
To test the effect of years of hormone treatment on dichotic laterality, a Pearson correlation was carried out on each trans-group. For MtF participants, the mean number of years on gender-appropriate hormone treatment was 3.08 years (SD = 5.18; min, 0; max, 30). A Pearson r test of the correlation between number of years on hormones with lateralization yielded, r = −.04, p = .75 (two-tailed), N = 59. For FtM participants the mean number of years on gender-appropriate hormone treatment was 3.44 years (SD = 3.35; min, 0; max 11). A Pearson r test of the correlation between years on hormones with lateralization yielded, r = −.16, p = .46, (two-tailed), N = 24.
There was no evidence in these data that hormone treatment significantly affected dichotic laterality in these groups of transpeople.
Handedness
An odds ratio analysis for being non-right-handed between the MtFs and the male controls was equal to 0.11. Thus the chance of finding non-right-handedness in the MtF group was more than 10 times greater than finding it in the male control group (Table 7).
TABLE 7. Handedness (Percentage) as a Function of Trans Status
Trans Status |
N | Right-Handed | Non-Right-Handed |
MtF |
68 | 45 (66.2%) | 23 (33.8%) |
FtM |
26 | 18 (69.2%) | 8 (30.8%) |
Control males |
36 | 34 (94.4%) | 2 (5.6%) |
Control females |
28 | 26 (92.8%) | 2 (7.2%) |
A similar analysis for the FtM versus female controls gives an odds ratio of 0.17. Thus the chances of finding non-right-handedness in the FtM group was more than five times greater than finding it in the control female group.
DISCUSSION
The primary finding from this study was the significantly different dichotic lateralization pattern of the MtF participants from that of the typical male control group. The mean MtF pattern was very balanced, showing a marginal but significantly lower right-ear advantage compared to the marked REA of the male control group. In this respect the first experimental hypothesis was supported.
The second experimental hypothesis was that the FtM participants would produce a significantly more marked REA than the female reference group. The finding that the FtM participants produced a mean dichotic pattern that was balanced and in the opposite direction to the hypothesized pattern is puzzling. It was, in fact, not significantly different from the lateralization pattern for typical females.
These lateralization indices for the control groups were similar to the lateralization scores for groups of typical males and females obtained by Rahman et al. (2008) who used similar stimuli and a similar experimental procedure. The lack of significant correlations between the number of years taking gender appropriate hormone therapy and dichotic lateralization for both MtF and FtM participants implies that their atypical dichotic performance was not the product of their hormone treatment as adults.
The results of the dichotic test clearly indicate a link between gender and the organization of verbal processing capacity. However, this relationship cannot be simple. The transwomen produced results that were predicted on the basis that, at least in some respects, their brains were feminized and that this measure of verbal laterality would confirm their putative feminization. The transmen, on the other hand, did not produce a more masculinized dichotic pattern; in fact, they produced an unexpected pattern not significantly different from the pattern for the typical female group.
These data, taken with the results of the handedness analysis, imply that the trans-conditions of males and females have different underlying etiologies. With preliminary data, Diamond and Hawk (2004) have reported a significant difference in the concordance of transition among male monozygotic twins from that found among female monozygotic twins. It was about twice as frequent among male twins as female twins.
Trans Status and Handedness
The raised incidence of non-right-handedness among the transwomen compared to the typical males in this study was in line with much previous research and is well established. The similarly raised incidence of non-right-handedness in the transmen compared to the typical female participants has been reported in previous research (Orlebeke, et al., 1992) and may be regarded as now being confirmed. However, given that the percentages of right-handers in the general population is nearer 90% rather than the higher percentages found in the control groups in this study, the odds ratio analysis would seem to indicate a misleading overestimate of the occurrence of non-right-handedness in the trans-population. Also, the way in which all non-right-handers have been grouped together in this study has highlighted the trans- versus control-group differences.
In research in the areas of sexuality and gender, handedness has been measured in a variety of ways, but there is little agreement as to how this should best be done. Most studies have used the EHI (Oldfield, 1971). Orlebeke et al. (1992), in an early report of raised sinistrality among transpeople, used a modification of this handedness scale. More recently, Wisniewski, Prendeville and Dobs (2005) employed it in their study of lateralization among transpersons. The seminal study by Knecht et al. (2000) also used the Edinburgh Inventory but converted the range of scores, potentially −100 for strong left-handedness to +100 for strong right-handedness, into seven categories. These seven categories seemed to resolve into two at each extreme with three more central groups. Recently, Dragovic, and Hammond (2007), using data derived from the Annett Hand Preference Questionnaire (Annett, 1970) and a battery of sophisticated statistical instruments, have suggested that the handedness of their participants can be best characterized as falling into three categories: right, left, and mixed. However, they also make a strong case that the use of questionnaires to measure handedness is fraught with theoretical and practical difficulties. This rather pessimistic analysis has been weakened by Knecht et al. (2000), who have shown that handedness (as measured by a questionnaire) and hemispheric language dominance are closely related. Using functional transcranial Doppler sonography, they found that right-hemisphere language dominance increased linearly as handedness moved from strong right-handedness, through ambidextrality to strong left-handedness in their nonclinical participants. This result strongly supports the utility of handedness questionnaires and also supports the view that the performance on dichotic tests reflects the underlying verbal processing substrate. The relationship between non-right-handedness and reduced or reversed asymmetry in verbal dichotic tasks together with the relationship between reduced asymmetry in dichotic tasks and transsexuality raises the possibility that non-right-handedness, reduced dichotic asymmetry, and transsexuality have a related underlying etiology.
The recent reevaluation of the standard view of sexual differentiation of the brain sees the different expression of genes in male and female brains as playing a major role in sexing brains (Arnold, 1996; Vilain, 2000). In the current study the high rates of non-right-handedness together with the shift in language lateralization implicates two different but linked brain areas. Handedness seems to involve lateralization in the cortical and subcortical motor pathways. The subcortical structures include the striatum, the caudate nucleus, and the putamen. These structures, together with the other basal ganglia, modulate the movements initiated by the appropriate cortical areas. Mayer, Mosler, Just, Pilgrim, & Reisert (2000) have shown that the ventral striatum, especially the substantia nigra, in rodents is the site of the highly concentrated expression of the Y-linked SRY gene, which is important in the regulation of dopaminergic activity in that region (Dewing et al., 2006). In humans, SRY has been shown to be expressed in the hypothalamus and the temporal and frontal cerebrum (Mayer, Lahr, Swaab, Pilgrim, & Reisart, 1998) and its significant expression in the human ventral striatum is thought likely (Davies & Wilkinson, 2006). Thus, handedness may be a consequence of the dynamic interplay between the appropriate cortical areas and the motor elements of the basal ganglia, both of which are modified by sex-differentiating hormones or genes.
Evidence for sex-atypical activity in the hypothalamus has been reported in transsexual people by Bergland, Lindstro¨m, Dhejne-Helmy, and Savic (2008), who found that their group of MtF participants produced a pattern of hypothalamic activity in response to odorous steroids that was intermediate between that of typical males and females but with “predominantly female-like features.” Also the MtF condition is known to involve an anomaly in the basal nucleus of the stria terminalis (Swaab, 2007). And, after this article went to press, a paper appeared by Garcia-Falgueras and Swaab (2008) that presented further evidence of hypothalamic involvement in the MtF and FtM conditions.
The anomalous expression of Y-linked genes cannot be implicated in the etiology of the FtM condition, as such individuals, in common with non-trans-females, do not have a Y chromosome in their genome. Thus genes on other chromosomes might be involved in basal ganglia and hypothalamic activity in female populations. In the absence of more general gender-atypical characteristics that might indicate early exposure to inappropriate hormonal environments, atypical gene expression would seem to have explanatory power.
CONCLUSIONS
The marked difference in dichotic pattern between MtF transpeople and typical males in the less lateralized direction is a strong indication that the brains of MtF transpersons are, in some respects, feminized. The unexpected dichotic pattern seen in the FtM group indicates separate underlying etiological mechanisms in the MtF and FtM conditions, the one not being the mirror of the other. The marked predisposition to non-right-handedness in the MtF trans-group and a similar but less clear tendency in the FtM group may point to atypical gene expression in subcortical and cortical brain areas.
The basal-ganglia-mediated link between handedness and language lateralization would imply that handedness and language lateralization are developmentally linked.
It is very difficult to see how the atypical patterns of dichotic performance produced by the trans-groups in this study could have arisen through their interaction with the social environment. Clearly, these findings are a consequence of biological processes. The previously noted high incidence of non-right-handedness found among transpersons was again seen in this study. This also points strongly to an inherent physiological involvement. The present authors know of no societal influence that encourages non-right-handedness.