On the Science of Changing Sex

Autogynephilic vs. Even More Autogynephilic

Posted in Science Criticism by Kay Brown on September 12, 2014

critical-thinkingThe $64K question about the Freund/Blanchard two type taxonomy is, “Is it, in fact, taxonic?”  That is to say, “Are there in fact two types of MTF transsexuals with each having a different etiology?”  The evidence pointing to it is formidable;  If one is to try to prove this to be incorrect the evidence to counter it would have to be even more formidable and convincing.  A recent paper by Veale attempts and claims to do just that, but fails utterly upon examination, as we shall see.  What she has done is to demonstrate that the difference between exclusively gynephilic and bisexual/pseudo androphilic transsexuals is purely dimensional and not taxonic.

But first, we need to define what is meant by taxonic, and what is not taxonic.  As Gangestad explains it,

“Meehl defined a taxon as “a nonarbitrary class whose existence is conjectured as an empirical question, not a mere semantic convenience”. A domain containing taxa is taxonic. Examples include biological sex, biological species, some disease entities (e.g., measles), and some ideological systems in politics or religion (Meehl, 1992). Many taxa are characterized by their causal simplicity. Taxonic domains are more likely than dimensional ones to have specific etiologies, including dichotomous necessary causal factors. For example, infectious diseases are taxa, and their causes consist of specific microbes. (More complex causal processes, such as thresholds and polarization effects, “may also underlie taxa” .) The existence of taxa can be supported either by the demonstration of requisite causal processes or by formal mathematical taxometric methods, which decide whether latent taxa underlie a set of candidate indicators of a conjectured taxon based on numerical relations between them. If so, the formal-numerical taxa that are thereby defined are empirical. Their causal basis must be discovered through additional research, and, thereby, taxometric findings can guide future inquiry into the causes of variation in the domain.”

Turning to the Freund/Blanchard taxonomy, as Veale explains it,

“Blanchard (1989) proposed that there are two distinct types (taxa) of male-to-female (MF) transsexuals and these distinctions are characterized by their sexuality: ‘‘autogynephilic’’ or ‘‘homosexual.’’ According to Blanchard, autogynephilic MF transsexuals are sexually attracted to females (gynephilic), both sexes (bisexual), or neither sex (analloerotic); they are not unusually feminine in childhood; and prior to transitioning often live outwardly successful lives as males, frequently marrying and having children. These MF transsexuals also experience autogynephilia—a term which Blanchard used to refer to ‘‘a male’s propensity to be sexually aroused by the thought of himself as a female’’. Homosexual MF transsexuals are exclusively sexually attracted to males (androphilic), do not experience autogynephilia, are highly feminine in their childhood, do not generally have success with attempts to live in the male role, and tend to present for treatment of their gender dysphoria at a younger age. By splitting MF transsexuals into these two groups based on their sexual orientation, Blanchard 1988,1989) and others (Freund, Steiner, & Chan,1982; Johnson &Hunt, 1990; Nuttbrock et al., 2011a; Smith, van Goozen, Kuiper, & Cohen-Kettenis, 2005) have found evidence for the average differences between these groups that Blanchard proposed. However, these differences do not necessarily imply a typology exists. These results could also have been reached if there was a nontaxonic latent structure with a correlation between the sexual orientation of MF transsexuals and these other differences. Although Blanchard (1985a) found some evidence for a taxonic latent structure of the sexuality of MF transsexuals by obtaining cut-off scores for classifying MF transsexuals into two groups using an earlier version of the taxometric procedures described in this article, he did not specifically test whether the sexuality of MF transsexuals was taxonic or dimensional.”

Veale acknowledges that there are other data that suggest, strongly I would argue, that there are clearly two taxa within the MTF transsexual population.  (See FAQ for list of data supporting the two type taxonic structure in MTF transwomen.  Also see my blog category “Supporting the Two Type Taxonomy” for essays.)  She only directly lists the fraternal birth order effect, which exclusive androphilic transsexuals share with conventional gay men, but NOT autogynephilc transsexuals.  This is important, in that Gangestad demonstrated that sexual orientation roughly correlated with a latent taxa in the general population.  Thus, there IS a taxonic structure to be considered here… and if so, we need to ask what would a non-taxonic structure among the transsexual population look like?

What is not acknowledged in Veale’s paper, but should be, is that clinical experience shows that autogynephilia also exists in male individuals who do not exhibit gender dysphoria sufficient to motivate them to transition.  We call these men, “Cross-Dressers” (CD) or “transvestites” (TV).  From long clinical observation, it has been noted that there does not appear to be a sharp line between such autogynephilic individuals, that indeed most autogynephilic transsexuals had careers that exactly match those of CD/TV men before transition.  That is also to say, that there appears to be a smooth, dimensional (non-taxonic) spectrum from mildly autogynephilic men to autogynephilic transwomen.  Also from societal and clinical observation, we note that exclusively androphilic transwomen population appears to smoothly blend toward feminine gay male “drag queens” to “effeminate” gay men, with no obvious gaps in the populations.  That is also to say, that there appears to be a smooth, dimensional (non-taxonic) spectrum from mildly feminine gay men to exclusively androphilic transwomen.  If Veale is right, than that would imply that heterosexual cross-dressers are members of the same taxon as feminine gay men.  Although I don’t believe this to be correct, it has a certain superficial theoretical appeal in that cross-dressers and autogynephilic transwomen both claim that their behavior is related to a ‘feminine essence’.  Hypothesis were meant to be tested.  I will come back to this after further examination of Veale’s study below.

In order to test whether Blanchard’s hypothesis is correct, she used statistical procedures proposed by Meehl; MAXCOV and MAMBAC.  I’m not going to even try to explain how these work.  But there are a few points to using these techniques that should be noted as they will figure prominently in the reasons why Veale failed to demonstrate that Blanchard’s hypothesis is incorrect.  First, the bare minimum number of subjects in the study needs to be 300.  The second is that the minimum number of subjects in the smaller of the two suspected taxons must be at least 30, and that the measures used to differentiate the suspected taxons must be valid.  Veale collected the needed 300 transwomen, but she failed to reach the needed minimum of 30 exclusive androphilic transswomen; worse, she failed to differentiate between bisexual/psuedo-androphilic and exclusely androphilic transwomen, using an invalid measure, as Lawrence showed in her critique of the Veale study,

In this commentary on Veale’s article, which incorporates a reanalysis of some of her data, I will argue that her critique of Blanchard’s typology was invalid, because:
1. Veale’s measures of sexual orientation and autogynephilia were not well constructed, which probably interfered with the accurate identification of participants’ sexual orientations and artificially lowered estimates of their autogynephilic arousal.
2. In her taxometric analysis,Veale employed several measures that were unrelated to the defining features of Blanchard’s typology. This rendered her analysis invalid as a confirmation or disconfirmation of Blanchard’s typology: Whatever typology Veale was examining, it was not Blanchard’s typology.
3. Even if Veale had used well-constructed measures of sexual orientation and autogynephilia and had conducted her taxometric analysis utilizing the defining features of Blanchard’s typology, the number of genuinely androphilic participants she was able to recruit—probably 18 (5.8 %) at most, and possibly even fewer—was too small to be reliably detectable through taxometric analysis. If the structure of Veale’s data appears to be dimensional rather than taxonic, this is because her participants consisted almost exclusively of only one of the two MtF transsexual types: nonhomosexuals (or nonandrophiles).

I’m far less concerned about the construction of Veale’s novel measures of autogynephilia, as almost any measures that meet the minimum required correlations with Blanchard’s Core Autogynephilic Scale would do for a taxonometric analysis.  But the real deal killer to her study was that she incorrectly included so many clearly non-exclusively androphilic transwomen in the “exclusively androphilic” category.  How do we know this?  My suspicions were first raised by the low effect size (d=0.37) regarding core autogynephilia and sexual orientation.  From Veale:

Sexual Orientation:   Androphilic               Nonandrophilic
.                                         (n=36)                           (n=272)
Core                                 Value (SD)
autogynephilia          14.08 (14.66)              19.00  (12.21)

The values are nearly the same, especially considering the very large range of scores (standard deviation).  These score differences strike me as being very much like the differences in reported autogynephilia found between exclusive gynephiles and bisexuals and asexuals in previous studies, rather than the larger differences between non-androphilies and androphiles.  If a large number of the putative androphiles were in fact bisexual and asexual, I would expect the scores to look exactly like this.  Lawrence also suspected this to be the case as she corresponded with Veale to obtain the raw data used to classify the subjects into sexual orientation categories.  From this, she demonstrated that at least half (and likely more) of the 36 putatively ‘androphilic’ subjects were decidedly bisexual instead,

“I have reorganized these supposedly androphilic informants into three groups, based on their self-reported gynephilia before age 16 (‘‘degree to which, until the age of 16, you felt sexually attracted to females’’; Veale, 2005, p. 129) and within each of these groups, based on their self-reported androphilia before age 16 (‘‘degree to which, until the age of 16, you felt sexually attracted to males’’; Veale, 2005, p. 129). My reasoning is that some of Veale’s purported androphiles were probably actually pseudo-androphiles or bisexuals—that is, their fundamental sexual attraction was towards females, but at some point they developed a secondary sexual interest in males as a consequence of their autogynephilic sexuality—and that the easiest way to detect these persons would be to look for informants who reported moderate or greater levels of gynephilia or very low levels of androphilia or both during adolescence.

First, consider the 12 informants with Reference numbers 1–12 in Table 1: All reported sexual attraction to females at least ‘‘occasionally’’ before age 16, even though their total Androphilia scores were high and their total Gynephilia scores were low. Not surprisingly, however, their mean Core Autogynephilia score—24.0—was also high, compared to a mean of only 19.0 in the 272 informants whom Veale categorized as nonandrophilic. Moreover, the mean Autogynephilic Interpersonal Fantasy score of these 12 informants—10.9—was also high, compared to a mean of only 8.2 in Veale’s 272 nominally nonandrophilic informants. Based on their early history of gynephilia and their high mean Core Autogynephilia and Autogynephilic Interpersonal Fantasy scores, I believe there is good reason to conclude that these 12 informants were actually pseudo-androphilic or bisexual, rather than genuinely androphilic as Veale alleged.

Next, consider the six informants with Reference numbers 13–16 and 24–25 in Table 1. Although these informants reported little sexual attraction to females before age 16, they also reported ‘‘rarely’’ or ‘‘never’’ experiencing sexual attraction to males before age 16; this suggests that they developed significant sexual attraction to males rather late in life. The mean total Androphilia score of these six informants—11.8—was also somewhat lower than that of the rest of the nominally androphilic group that reported little gynephilia (Reference numbers 17–23 and 26–36; n=18), with a mean total Androphilia score of 16.4. Again not surprisingly, the mean Core Autogynephilia score of these 6 informants was 18.3, nearly equal to that of the 272 nominally nonandrophilic informants, 19.0, and substantially higher than that of the rest of the nominally androphilic group that reported little gynephilia (n=18), 6.1. However, the mean Autogynephilic Interpersonal Fantasy score of these six informants, 6.5, was similar to that of the rest of the nominally androphilic group that reported little gynephilia (n=18), 6.9. Although the case here is not quite as strong as for the 12 informants who reported significant gynephilia before age 16, I believe that most or all of these 6 informants who rarely or never experienced sexual attraction to males before age 16 were also pseudo-androphilic or bisexual, rather than genuinely androphilic as Veale alleged.

There are also a few other specific supposed androphiles whose androphilia seems questionable. For example, consider the informant with Reference number 30,whose total Gynephilia score of 8 reflected her being aroused by females in her current sexual fantasies ‘‘almost all the time,’’ currently feeling sexually attracted to females ‘‘occasionally,’’ and having been conscious of sexual arousal to females while in physical contact with them ‘‘occasionally’’: Given her Core Autogynephilia score of 37 and Autogynephilic Interpersonal Fantasy score of 13, one might suspect that she was actually pseudo-androphilic or bisexual. Or consider the informants with Reference numbers 18 and 26, whose total Androphilia scores of 9 both reflected the identical pattern of only ‘‘occasional’’ attraction to males before age 16 and currently, sexual arousal to males in their current sexual fantasies only ‘‘sometimes,’’ and having been conscious of sexual arousal to males while in physical contact with them only ‘‘occasionally’’ : Here one might suspect mild pseudo-androphilic or bisexual ideation in fundamentally analloerotic individuals.”

From this analysis, Lawrence contents (and I whole heartedly concur) that 18 of these individuals were not exclusive androphiles, and three more were very likely not.  Personally, I would say that these additional three are definitely not.   Given that this would leave, at best, only 18, and likely less, androphilic individuals, the number is definitely too low to be detected as a taxon using the techniques in Veale’s study.  Remember also that for taxonometric analysis, that the scales must be valid.  Failing to differentiate bisexual from exclusively androphilic demonstrated that the sexual orientation scale Veale employed was most definately not valid.  Further, given that most of the putatively androphilic individuals were in fact non-androphilic, Veale’s study does successfully demonstrate that the various non-androphilic categories are dimentional and non-taxonic, supporting Blanchard’s hypothesis.

Lawrence pointed out that Veale’s method of obtaining her subjects was prone to ascertainment bias.  Simply put, one does not find androphilic transwomen at autogynephilically dominated venues such as online forums and support groups.  Veale herself recommended that a “more representitive” sample would be needed for future studies.

Suggestions for future studies.

Hypothesis were meant to be tested.

Lawrence has criticized Veale’s use of her novel “attraction to feminine males” and to “attraction to transgender fiction” scales.  Veale has previously shown an interest in developing these as a means of exploring autogynephilic sexuality, as the former is based on the observation that many autogynephiles are gynandromorphophilic and that many autogynephiles like to write and share autogynephilically inspired erotica.  While I think these are worthy areas to explore, I have to concur with Lawrence that these are not the appropriate scales to use for future taxonometric analysis tests in MTF transwomen.  Instead I would highly recommend using Blanchard’s Autogynephilia Scale and the Childhood Gender Non-conformity (CGN) Scale from Gangestad, as I believe that these will show that the Freund/Blanchard categorization of MTF transsexuals is taxonic with the two types being autogynephilic and “homosexual”.  My prediction is that there will be a very high negative correlation between CA and CGN scales that will show a taxonic structure over sexual orientation.  Sexual orientation should be based on pre-SRS behavior, not on self-report, which has been shown repeatedly, to be unreliable and subject to social desirability bias in autogynephilic transwomen.

For a dichotomous separation of exclusively androphilic and non-androphilic transwomen, I recommend a multi-factorial sieve.  First, if they self report being gynephilic, bisexual, or asexual: believe them, they are non-androphilic.  Next, of the remainder, ask these two questions, “Are you now, or have you ever been, married to a woman?” and “Have you ever fathered a child?”; if yes, then they are non-androphilic.  (Lawrence previously used marital status to exclude bisexuals from androphiles.)  Finally, of the remainder, survey their actual sexual history with these two items, “Estimate how many times you have had sexual intercourse with a woman” and “Estimate how many times you had receptive sexual contact with a man’s penis before SRS (excluding  cross-dressed/feminized males)”.  A validation study may be needed to determine appropriate cut-offs for the first question.  I would be tempted to use “zero” (0), but that may be overly aggressive… but I would be very surprised if genuinely exclusive androphiles had more than a nominal experimental number, say three at most.  The second question supports the first, and also allows discrimination of celibate analloerotic individuals.  I expect that for androphilic transwomen, older than say 2o, the number of sexual contacts with men will be greatly higher than the number of vaginal intercourse experiences.

For a continuous androphilic scale, (moderately) free from social desirability bias, take the number of vaginal intercourse experiences and subtract it from the number of sexual encounters with men.  This scale will be negative for sexually experienced gynephiles and positive for androphiles.


Gangestad, Steven W.; Bailey, J. Michael; Martin, Nicholas G., “Taxometric analyses of sexual orientation and gender identity”
Journal of Personality and Social Psychology, Vol 78(6), Jun 2000, 1109-1121. doi: 10.1037/0022-3514.78.6.1109

Norris, et al, “Homosexuality as a Descrete Class”
Psychological Science, Oct 2015
doi: 10.1177/0956797615598617

Jaimie F. Veale, “Evidence Against a Typology: A Taxometric Analysis of the Sexuality of Male-to-Female Transsexuals” Archives Sexual Behavior
DOI 10.1007/s10508-014-0275-5

Anne A. Lawrence, “Veale’s (2014) Critique of Blanchard’s Typology Was Invalid” Archive Sexual Behavior
DOI 10.1007/s10508-014-0383-2


Comments Off on Autogynephilic vs. Even More Autogynephilic

%d bloggers like this: