What is autogynephilia? According to the Erotic Target Location Error model, it is in some sense the application of gynephilic attraction to yourself; that is, a sexual interest in being what you find attractive (erotic target identity inversion), plus an attraction to women (gynephilia). This hypothesis predicts that other sexual interests (e.g. attraction to amputees or children) have corresponding autophilias (e.g. apotemnophilia or autopedophilia), which might act analogously to autogynephilia (e.g. causing body integrity identity disorder or age dysphoria).
There’s some evidence of this; for example, apotemnophilia and autopedophilia exist and tend to cooccur with acrotomophilia and pedophilia. However, there’s a lack of research in the validity of the concept within the normal range of attraction. For example, do autogynephiles who like redheads also want to have red hair? Some time ago, I did a survey awkwardly named Survey on traits you’re attracted to or would like to have where I examined exactly that. I decided on a number of traits that one might find attractive and/or want, asked people about those traits plus a number of other things (e.g. AGP), and got these results:
The correlation coefficient between attractiveness of and desire to have various traits, split by various groups. The “(Average)” spot at the end shows the average correlation coefficient over all of the traits.
First of all, let’s think about what we would a priori expect. It’s perfectly possible that something other than ETLE might cause a correspondence, so we can’t assume that non-ETLEs would have zero correlation. However, surely ETLE should lead to a greater correspondence than there would otherwise be, perhaps in proportion to the intensity of the ETLE. So a pattern of [Non-AGP cis men < AGP cis men < Transfems] would be appropriate. It would also be great to have [Cis women < Transfems], since otherwise there would be a lot of worry about confounding, but it’s not as clear that we can assume this to hold.
Now, in practice I don’t really have a big enough sample to talk about the transfems, so it’s probably better to focus on the non-AGP and AGP men. As you can see, there was quite a big correspondence between what traits they found attractive and what traits they wanted to have, but this applied equally well to both AGP and non-AGP men. I don’t think this is conclusive evidence, as I can see a number of problems with the approach, but it’s certainly something that’s worth taking into account.
Let’s talk problems with my approach:
- A lot of the traits I used are a bit weird. I had a hard time picking good traits, and I didn’t make it easier for myself by trying to pick traits that followed certain rules (not too sexually dimorphic, not too universally seen as good/bad, equally attractive in men and women, not signalling membership of specific subcultures). I didn’t even manage to follow the rules to a degree that I felt was satisfactory either. Perhaps a better selection of traits (and rules?) would improve things.
- It’s not clear that the traits picked could truly be said to be part of the participant’s sexual orientation. It’s possible that the main thing both of the questions measured was how positive an emotional valence people had around the traits, with essentially no signal coming through in terms of sexual orientation. In that case, it’s not surprising that I found no effect.
- It’s unclear whether the desire to have those traits would apply if having them didn’t come with being female. Even if we assume an AGP man who likes redheads would also desire to be a woman with red hair, it’s not clear that this would lead to especially much interest in being a man with red hair. On the other hand, many apotemnophiles are interested in losing limbs without changing sex, and many autopedophiles are interested in being children without changing sex, so it seems that maybe the same should apply to other traits.
- It’s possible that the men I involved weren’t AGP enough. It would be relevant to repeat this test with some men who are extremely AGP.