Why Blanchardians are scientifically unsound (and no, it’s not because activists hate them)

Every now and then, I pop over to ThirdWayTrans’ blog — someone whom I consider fascinating from several points of view, especially because even after detransitioning, he’s open-minded enough towards those who wish to transition for the right reasons; he just warns that a diagnosis of ‘gender dysphoria’ might not be that easy to figure out, as he painfully had to figure out for himself during the two decades living as a female, after having transitioned due to a misdiagnosis.

Unfortunately, as so often happens, such blogs tend to attract radicals of all kind, and while actual transphobic people have been kept somewhat away from TWT’s blog, a lot of Radical Blanchardians often populate the comments, trying to defend their point of views (or, rather, Blanchard’s); while Radical Anti-Blanchardians try to demove them from such views, which are demeaning to the trans community.

This is, of course, an endless battle of opinions which will never die as long as enough Americans are alive; it’s the same as Creationism vs. Evolution; Faith vs. Atheist Humanism; and so forth. When it’s about personal opinions, you will only be able to persuade others with your argumentation if those others are open-minded enough; but usually they aren’t and will cling to their opinions, no matter how illogical they might be. This is all about the emotional battleground; reason has little to do with it, and therefore, I usually keep out of those battles.

However, it irks me when the ‘scientific argument’ is invoked. Scientists obviously are human, have their bias, and in spite of everything, they can make mistakes. But there is a very particular meaning assigned to so-called ‘scientific truth’, and it’s not to be taken so lightly when flinging arguments around.

So I tried to reply to one comment there, made by Rod Fleming, claiming that Blanchard’s views and theories are fundamentally sound. In his own words, he wrote:

Many have tried to debunk Blanchard and all have failed. His theory is sound and is the basis of scientific understanding of transsexualism.

One must be careful of such wide generalisations! We have to explain the context in which such claims are made, or we would be just spreading a falsity.

If the above claim is made within the context of ideology/activism, what matters is the idea behind it, and one idea is just that, a thought, a mental construct. It’s not ‘true’ nor ‘false’ before enough people believe it (or disbelieve it). Thus, one may claim that the opinion of those who debunk Blanchard, from an ideological/activist perspective, is worthless, while Blanchard’s own ideas are solid and valid. Because in this particular context that particular opinion (or Blanchard’s, for that matter) is equally valid as the opinion of the anti-Blanchard activists, then such a claim would be acceptable, since it’s merely a question of how many people sustain one opinion against other. And because Blanchardians are still around defending his ideas, one can correctly argue that, in spite of everything, Blanchard’s ideas are still being spread around, they are still influential, they still persuade others to think like Blanchard did and accept his ideas — and, in that context, Blanchard’s ideas were not ‘debunked’, since they are still being accepted (as opposed to, say, believing that the Earth is flat and that the Moon is made of cheese).

But when we step into the scientific aspect of Blanchard’s theories, then the whole issue is completely different. Science obviously also starts with an idea — if that idea can be scientifically validated, it becomes a conjecture. If we actually do some tests or experiments to validate the idea, then it becomes a hypothesis. If we can prove beyond a shadow of doubt that the hypothesis is correct and that the data used and the results can be independently validated, then it becomes a theory (back in the 1700s and early 1800s it would be called a law of science; it’s same thing, it’s just that we changed terminology a little, when ‘science’ split from ‘natural philosophy’, with which it shared a common system of thought and similar methodologies to acquire knowledge, which we today call ‘the scientific method’).

So, Blanchard’s ‘theory’ began as a conjecture, and as a conjecture, it means making the question ‘Are transexuals split between two groups, effeminate male homosexuals and autogynophiliac males?’ Posing that question is scientifically correct, that is, it’s a valid scientific question to be posed, since it can be falsifiable: this is a requirement for scientific conjectures since the days of Karl Popper (a philosopher of science from the early 20th century). It means that such a question can, indeed, be put to an objective test, data can be collected and analysed, and either the hypothesis is confirmed or it is rejected. Posing any question which can be rejected through the analysis of data made through observation is a scientifically valid quesion; here, the ethics or morality of such a question is totally irrelevant: for science, all it matters is that you can answer a question by setting up an experiment to validate or reject the hypothesis; and this is certainly the case with Blanchard’s conjecture.

We can even argue further that it is an important question, because it may mean for a doctor that different treatments are needed for each case — so we are in the domain of establishing differentiated diagnosis, which crucial to medical science, and therefore the question makes sense: it is, indeed, a valid conjecture, it may have a practical application in treating patients (by correctly diagnosing them), and, furthermore, this question can be proved or disproved using the scientific method.

But here is where problems arise. The original data that Blanchard used to ‘prove’ his conjecture — thus turning it into a hypothesis, and inviting others to replicate his results so that he could claim that it was more than a hypothesis, it was a theory (that’s as close we can get to call it ‘scientific truth’) — had flaws in it. Blanchard, as a serious scientists, published all his data, and it was relatively easy to show how he had made some trivial statistical errors here and there. I’m not an expert in statistics, but basically Blanchard’s own published data disproved his theory completely. And this is what ‘debunking’ means in science: a scientist’s peers go through the same data, and find that some calculation errors are present, which induced the original author to reach false conclusions. Note that it is not unusual for such things to happen in science: the actual questions that scientists pose come mostly from inspiration, and sometimes they seem to be logically following from the scientists’ point of view, and this certainly was the case of Blanchard: after reviewing so many patients at his clinic, it seemed to him that they were of only two kinds. All he needed to publish his theory was to get the data to objectively sustain his conjecture. Unfortunately, manipulating numbers to get the results one wishes to see is not ‘science’; but because Blanchard was required to publish his original data, others could see where he committed errors in his calculations; in fact, the data he published showed that transexuals did not only not fall neatly in two groups, but that there was no reasonable argument to divide the sample in separate groups — in other words, beyond the common label of ‘transexuals’, there was no statistical evidence that you could subdivide the sample in any amount of subgroups: each of them was an individual, not sharing (statistically) enough common traits with any other. Now we could argue that this only happened with Blanchard’s original small sample, and that a larger sample might show distinctive subgrouping; but this would be another question, and another completely different article. What Blanchard’s peers could only claim was that his published data did not validate his conjecture, and, as such, the only possible answer to Blanchard’s question would be: ‘with the sample provided, there is no statistical evidence that transexuals fall into two and only two different subgroups’.

As said, Blanchard is a serious scientist, so he admitted to the errors, and what he did was what every serious scientist does: he studied more, got more samples, and published more articles. The next papers he published did not contain blatant statistical errors. But they still raised some questions; and it was only years after publication that it was found out that he left out of his statistical analysis all people which did not ‘fit’ into his theory. That is not necessarily ‘cheating’: it’s common, in science, to get rid of statistical anomalies (you can call them edge cases if you wish, or singularities, whatever), because often these are just the result of some stupid errors which have nothing to do with the methodology that was used, and are just confusing the whole picture. We can assume that Blanchard was still being honest in his approach, and therefore he shrugged off those cases that did not ‘fit’ in his theory and discarded them. The problem is that when such people are included and the same calculation is applied — a calculation, mind you, designed by Blanchard himself to ‘prove’ his conjecture — the results, once again, run contrary to the conjecture: once more, they continue to prove the opposite of what Blanchard intended, namely, that there are not two different and distinct groups. The difference between them is not statistically significant, no matter how many people are added to the samples. It can only be made statistically significant if you discard all results that do not fit in the conjecture, but, as you can surely agree, this is not ‘good science’, and it means that Blanchard’s ‘theory’, from a purely scientific perspective (and not an ideological one), does not hold water.

There are two further issues here (and this is why Blanchard is considered to be debunked in several different ways). The first one is raising some doubts about Blanchard’s own data, and the way he extrapolated conclusions from it. The second one, of course, is to independently replicate his results, using his methodology, but using different samples, and see what kind of result is achieved.

None of those attempts managed to replicate Blanchard’s conclusions. In other words: when Blanchard’s methodology of determining unequivocally who is a ‘homosexual transexual’ and who is a ‘autogynephiliac transexual’ is applied to a different sample (i.e. a different group of transexuals, not the ones Blanchard has interviewed), then the results show that, once more, Blanchard’s conjecture is not proved. And, again, the reverse is always true: instead of splitting transexuals in two neat groups, the results always show that there are no statistically significant differences between them. And this has been replicated over and over again, using different samples, different sample sizes, in different countries, and so forth. Things like tweaking the age (thus formulating a slightly different conjecture, i.e. ‘is there a point in a MtF transexual’s life where he is either homosexual or autogynephiliac?’) to see if different age groups responded differently did fail as well. Because Blanchard’s sample came mostly from his own patients, others have attempted to either broaden the samples (people from different ethnicities, different social classes, and so forth) or even shorten them (limit to the kind of sample distribution used by Blanchard, but just using different people who happened to have similar profiles than Blanchard’s samples), all of them produced over and over the same result: no matter how you measure things, no matter what sample you use, Blanchard’s results cannot be replicated in any way, except, of course, by ‘cheating’, i.e. leaving out any members of the group who do not fit in the conjecture! But that proves nothing, of course, except that such a methodology is necessarily flawed.

But there is a second aspect as well, which is different in medical sciences, compared to other areas of scientific research. While Blanchard believed that transexuals divided themselves neatly in two groups — ‘homosexual transexuals’ and ‘autogynephiliac transexuals’ — he eagerly proved to his own satisfaction that both groups would benefit from transition to eliminate their symptoms of gender dysphoria. And this, indeed, is true. When someone is making the claim that ‘Blanchard was never debunked’ they very likely mean this particular point: that the only ‘cure’ or ‘treatment’ for both kinds of transexuals is the same, i.e., transition, and that transition, in either case, has an incredibly high rate of success. And this, indeed, has been validated over and over again, and it even follows from Blanchard’s data: it should not come as a susprise that all pseudo-groups of transexuals benefit from the same treatment, since those ‘groups’ are artificial anyway and do not have any statistical differences between themselves; in other words, it would be very surprising that any artificially created sub-grouping of transexuals would not benefit from transition! Indeed, if we even want to be kind to Blanchard, he actually proved (against his own will) that no matter how you subdivide transexuals among different groups, all will benefit from transition. This, again, is a corollary from Blanchard’s work, and Blanchardians can correctly claim that he was the first to figure this out and even prove it with his own data — even though he wanted to reach a different conclusion!

There is a catch, though. In medical science, if you have a range of similar symptoms for two conditions, and the same ‘cure’ or ‘treatment’ produces exactly the same results, then those ‘two conditions’ are just one and the same. This is how, in medical science, we often see many different illnesses ‘converging’ on a single one. For instance, take a look at the discussion about auto-immune diseases, which so often all have symptoms of fibromyalgia, irritable bowel syndrome, and clinical depression. Such patients are usually regarded as suffering from ‘different’ diseases. However, it was found out that anti-depressants would also reduce fibromyalgia and minimise the impact of irritable bowel syndrome (IBS); while at the same time, treating IBS with changes of the intestinal microbioma also made fibromyalgia and depression disappear. Thus, researchers are slowly concluding that all these ‘diseases’ are very likely the same one, just appearing to have completely different causes (fibromyalgia: pain, expressed by the nervous system which is oversensitive for some reason; IBS, problems originating in the intestine; clinical depression, mostly happening inside the mind, even if it responds well to some medication affecting certain neurotransmitters). In other words: when just looking at the origin of the symptoms, these seem to be completely different areas of medical science (and, indeed, they have been first reported independently from researchers looking at the issue from totally different perspective, knowledge, and scientific research); but these seem to be very common in most anti-immune diseases, appearing all together; and what seems to ‘treat’ one of them, treats the others as well, even though they do not seem to be related at all. Now, these are still hypothesis (‘do all these diseases have a common origin or not?’) which are being submitted to clinical tests, so there is no ‘unifying theory’ for all of them yet, we just know that they are not independent variables, but correlated ones: when one of those diseases appears, the others tend to manifest as well, to a higher or lower degree; and, conversely, when one of them is treated, the others tend to get treated as well. That shows there is a correlation between them. Are they perhaps just one and the same disease, just manifested differently, and treatable by different (apparently not related) means?

Back to Blanchard. What he is saying is that a certain population, divided neatly into two groups according to his conjecture, benefits from exactly the same treatment, and the end results (a successful transition, where gender dysphoria disappears, as well as other issues such as depression, anxiety, etc. caused by gender dysphoria) are precisely the same (or, scientifically speaking, there is no significant statistical difference between the results in one of the groups and in the other), then, according to the principles of medical science, this population has just one common origin (and not two, or three, or many). Inadvertently, because Blanchard’s results regarding successful ‘treatment’ of his ‘two’ types of transexuals showed, beyond a shadow of doubt (meaning: this result was proved over and over again, by different researchers, with different samples), that they reacted in the same way to the same treatment, he has invalidated his own conjecture without meaning to do so. In other words: yes, you can split the transexual population in as many groups as you wish (why just two?). All of them benefit from the same treatment. All of them get the same statistical results from transition. So there are not ‘many’ groups, or ‘many’ types of transexuality, there is just one. And, from a purely medical perspective, it’s worthless to ‘tag’ one kind of transexual as ‘type A’ or ‘type B’ or C or D etc. since all of them will benefit from transition in exactly the same way. Thus, from the point of view of medical research, there is no valid reason to postulate that there are ‘two different kinds’ of transexual people, since all kinds will respond to the same treatment in the same way. And, indeed, Blachard was the first to prove exactly that, which is ironic! In other words: in science, there is no ‘blame’ if you accidentally disprove yourself, which is what Blanchard did; you’re still a valid, honest scientist if you show results proving that you were wrong in your original conjecture — from the perspective of science, showing a ‘dead end’ in science is also good science, even if it’s not that glamorous as discovering something new.

And now to Lawrence. To be honest, besides Bailey’s book and some scattered articles/interviews, I don’t know his research work very well. Lawrence, by contrast, has a more interesting perspective, and that’s why I read a little more about her own work. She’s still a die-hard Blanchardian, of course, but she has perceived the weakness of Blanchard’s argument when it comes to actually trying to prove it with data. So Lawrence cleverly postulated that there are not precisely two types of transexuals, but rather a spectrum of transexuality variance between those two types. While this is a radical departure from Classic Blanchardism (as Rod argues, transexuals are either homosexual or autogynephiliac, there is no middle ground, all is black-or-white), it still keeps the rest of Blanchard’s argumentation intact; the beauty of it, of course, is that this explanation fits the data perfectly! In other words: Lawrence is still able to argue that many transexuals are ‘homosexual transexuals’ (and that’s true, even though that expression is not commonly accepted among scientists and the community, since it is meaningless; but let’s use it just for the sake of the argument), and that many are autogynephiliac; the data will certainly show that; but she has no choice but to recognise that all the others are not clearly separated among the two extremes but somewhere in-between. And this conclusion is validly drawn from the data.

One would therefore think that Blanchardism, with Lawrence’s subtle modification, would be un-debunkable that way — in fact, when good, honest scientists have a conjecture that does not explain the data, then they change the conjecture; that’s good science. The truth is that Lawrence’s argumentation is fallacious, and let me give you an analogy which will be immediately clear:

Suppose you’re a scientist exploring the following conjecture: people either like the colour green, or the colour red. You set up interviews with the question: ‘Do you prefer green or red?’ About half the people will prefer green, the other half red. Your conclusion: human beings split neatly between either liking the colour green or the colour red, just as predicted, so you have now a solid theory. Correct?

Of course not. Your peers will point out: there are more colours than green or red. You have not asked for those! But you persistently claim that people still prefer green or red, all other colours are irrelevant (and it’s also irrelevant what ‘green’ or ‘red’ means: what matters is the perception of ‘greenness’ or ‘redness’ and you argue that people naturally either prefer one or the other, and it matters little if they are pointing at a colour like, say, aqua or turquoise and label it ‘green’, or orange and pink and label it ‘red’). So you set up a new set of surveys, this time asking: ‘Do you prefer red/green/none of the above?’

Now here things become a bit more confusing in the data. What the results show is that a large proportion (possibly even the vast majority) of your respondents have answered ‘none of the above’. Many have answered red, of course, and many have answered green. You assume that the unexpected number of ‘none of the above’ is due to a badly formulated question, and so disregard those answers, claiming that those who are ‘undecided’ are the ones who preferred, say, ‘pink’, but did not answer ‘red’ because they were unsure if the same colour was meant. You publish therefore the results showing that the number of people answering ‘red’ is roughly equal to the number of people answering ‘green’, and this is actually the case with your data — and discard the rest of the answers, claiming that they come from a badly formulated question.

Your publisher refuses to take that answer. So you design a new test. This time, instead of a question, you put a swatch of colours that you consider green-ish — so this will include all blue-green and yellow-green shades, etc.; and a second swatch of red-ish colours, including oranges, pinks, and so forth. And the question, once more, is ‘Do you like the colours on the first group more, or the ones on the second group, or none of them?’

On this example, you have the choice of either liking one group of colours, the other group, or none at all (simply do not check either option)

The results are awesome for your conjecture. Almost half the people answered that they preferred the first group; almost all the other half answered they preferred the second group; and just a few scattered ‘anomalous’ kinds have answered they preferred none. You submit your paper again, do all the statistical calculations properly, even claim that the few people that answered ‘none of the above’ (i.e. didn’t check either group) did possibly misunderstand the question, or were colour-blind, or something or other, but they were so few that they do not interfere with the result — you can claim, with, say, 99.5% certainty, that human beings either prefer green (or green-ish) or red (or red-ish); you might even suggest that, to enhance the degree of confidence, further research could add more colours on the ‘red’ and ‘green’ swatches, and enlarge the number of participants, so as to diminish the noise from those few anomalous results, which would expect to diminish towards zero. Well, the maths seem to be solid, so that article is published.

But you see where the problem is: in fact, this test doesn’t prove anything. The more swatches of colour you place on either red/green category, the more likely someone will identify with one of them. So, someone who hates green but loves blue, for example, seeing that there are some blue-ish colours on the ‘green’ category, will answer ‘green’; while someone preferring yellow, seeing that there are some very light orange colours on the ‘red’ swatches, answers ‘red’. This just means that, given only two choices, but broadening the meaning of each of the two choices, you can neatly divide any group of people between the two. Eventually, someone could even design a test with all possible colours that humans can see and assign them to two arbitrary groups, and the results will be simply that humans, given only two choices, will pick one of them half the time.

So this is what Blanchard did. Of course, his peers called ‘foul’, because that’s just cheating. If Blanchard had created two transexual categories, one with those who like to set fires (‘pyromaniac transexuals’) and the other with those who don’t (‘non-pyromaniac transexuals’), he would come to the same conclusions: all transexuals would fit in either of those two categories (and a few would refuse to answer such a stupid question, but you could discard their rude answers from the data). But that means nothing. There is no difference in the essence of transexuality just because some like to set fires and the others do not; in other words: being pyromaniac or not is an independent variable with no correlation with transexuality itself. It’s exactly the same as trying to divide humans in ‘red-lovers’ and ‘green-lovers’ and somehow infer that there are two kinds of humans. It doesn’t work like that, this is just trying to correlate independent variables to each other and designing a phony test that will achieve that.

Now Lawrence was a bit more clever. She saw the fault with Blanchard’s analysis, so instead of asking yes/no questions regarding homosexuality and/or autogynephilia, she probably asked MtF transexuals something like this: ‘What is the degree of your homosexuality? Answer 1-5, where 1 means only drawn to men, 3 means bisexual, and 5 means only drawn to women. How often do you have autogynephiliac fantasies? Answer 1-5, where 1 means never, 3 means occasionally, and 5 means all the time’. Now, I have not seen Lawrence’s actual tests, but I can imagine such a test being given to participants. What would the results be?

Again, I can only speculate, but based on my personal experience (also known as ‘anecdotal evidence’ and not valid in science!), there would be many who answer ‘1’ for being attracted to men while simultaneously also replying ‘1’ to never having autogynephiliac fantasies; and, similarly, on the other extreme, we would certainly have lots of individuals admitting to autogynephiliac fantasies all the time but claiming to be only attracted to women. So these two extremes would, indeed, still fit into the Blanchardian model. The rest of all other transexuals would be a complex mix of both things.

So what did Lawrence do? Instead of assuming that ‘homosexuality’ and ‘autogynephilia’ are opposites of the same variable in transexuality, which was part of Blanchard’s conjecture and which he failed to prove, Lawrence admitted that they were two different variables, and went on to show that they are strongly inversely correlated, that is, when a transexual is ‘very strongly attracted to men’ it is more likely that they are ‘less prone to autogynephiliac fantasies’ and vice-versa. All that is in between can be ordered to show that the inverse correlation still holds (just with a lesser degree). What Lawrence can therefore claim is that all transexuals are, indeed, spread among either ‘homosexuals’ and ‘autogynephiliac’, but this distribution is not either/or, but rather a mix of both, while still remaining true in some cases — the very ones that Blanchard also found out — when at the same time she could provide an explanation for the data points that Blanchard had discarded: those were ‘undecided’ individuals, both homosexual to a degree and autogynephiliac to a degree, but, in general, the more they liked men, the less they had autogynephiliac fantasies. In essence, thus, Lawrence’s work still ‘proves’ Blanchardism, or some sort of ‘weaker Blanchardism’, and is more sound based on the data.

In this questionnaire, people can show a gradation of preferences. Note how it is still possible to say ‘no’ to both, and a ‘yes’ result to both would be nonsensical, since the point is to try to prove that the more you like green, the less you like red and vice-versa.

It is still faulty reasoning, of course. Let’s get back to our colour tests. Now, instead of asking people if they prefer the first or the second group, we ask, on a scale of 1-5, how much they like the colours in the first group, and how much they like the colours of the second group. This time, what we will see in the data is two peaks — those who rated ‘red-ish’ as 5 and ‘green-ish’ as 1, and those who rated things in the opposite way. But there will also be a lot of people in the middle. Perhaps even the majority will not immediately clear what they prefer most (i.e. everybody who likes violet, indigo, or pastel blue, creamy pink, whatever…. they might have rated both groups as ‘1’; while many people will be indifferent, i.e. rating both as ‘3’). But it might still be possible to reach two conclusions: that humans are divided mostly among those who prefer red and those who prefer green, and that even when they don’t, there is an inverse correlation among the rest of them: the more they prefer green, the less they prefer red, and vice-versa. Such a study (albeit still stupid and faulty!) would have a much higher chance of being published. because, mathematically, it’s more sound.

But in all these cases the issue is not really only in the math. It’s in the assumption that humans can fit in just two, and only two, categories. While this can happen in some trivial cases, e.g. ‘humans who had a higher education and humans who do not’, or ‘humans who suffer from the flu and those who do not’, it’s simply not provable in other realms — unless you cheat in the original questions. More precisely, if the classifications are vague enough to encompass a lot of quite different people, while giving little choice to give an answer that does not fit in either question, then we can ‘prove’ almost everything we want, and we can mathematically show in almost all cases that each category is inversely correlated to the other (in the cases of either/or categories, this will always be true to a degree). The problem is that humans simply may not fit in either category, or fit in both at the same time, and the tests designed by Blanchardians simply do not account for that.

Also, it’s easier to describe ‘redness’ or ‘greenness’; even so, we have some difficulties with it: while we can give the precise nanometer wavelength of ‘red’ and ‘green’ and design a test giving some variance around those precise limits, people’s perception of red and green vary, so it’s hard to know how a particular individual rates a colour as ‘red’ or ‘green’. The exact same shade (if it’s somewhere in the middle of red and green, in terms of wavelength!) can be classified by some as red, and by some as green. So, essentially, asking about a subjective perception of a given conceptual idea (colours are just names, and different cultures have different names for colours; for example, there are far more names for ‘red’ colours in the West, because the pigments for red colours were more easily available than those for green colours) means that the results cannot be evaluated in the simplistic matter that Blanchardians do; it won’t work that way. And even though ‘the degree of homosexuality’ can, to an extent, be asked (after all, Kinsley did that in the 1950s), ‘autogynephilia’ is a far more complex concept — we certainly know that it exists, but the issue is how ‘strong’ it is in each individual (since each individual will internally rate it differently…) and how much it is connected to either transexuality, some sort of sexual paraphilia, or merely a normal and healthy sexual fantasy.

When you pick up two concepts out of thin air, and wish to tag them to human beings, and try to ‘prove’ anything based on them, the results are usually very complex, and they rely much more in the kind of questionnaires handed out to participants, and how each participant is motivated to reply, than about the validity of those two concepts in the first place. There is a good reason why older ‘psychological profiling’ tests are now viewed with suspicion, since many come from social prejudice of the time they were made, and even if the researchers were genuinely interested in the question they formulated and the kind of answer they got, and even did the math correctly, the problem was that the original assumption was based on prejudice — not science! — and this influenced the way the questionnaire lead people to answer according to their prejudice, or, even more likely, according to their perception of what prejudice the researchers had. So we have layers and layers of abstract conceptions here, which are difficult to correlate to a concrete question and a concrete answer that satisfies the question.

This is the problem with the Blanchardian autogynephilia ‘theory’ (between scare quotes, because it would be only a theory if it had been proven beyond a doubt by Blanchard’s peers, replicating his tests and reaching the same conclusions — which they didn’t). It relies upon assumptions that are conceptually complex and that may make no sense at all. It also implies that such concepts are, indeed, related to transexuality, but although it is clear that they are present (everybody has a sexuality, even being asexual is a kind of sexuality; everybody has sexual fantasies, even very boring, vanilla ones; but there is a huge difference between what is merely a healthy fantasy, and what is a seriously disturbing paraphilia which prevents that person from performing functionally in society), it’s by no means obvious (and has not be ‘proven’ to be obvious) how they exactly influence what we call ‘transexuality’. And, of course, from the purely medical perspective (and not ‘pure scientific research’), coming to the conclusion that two different ‘kinds’ of transexuals have to be ‘treated’ the same way just disproves directly the initial assumptions, that is, that there were two categories in the first place. So from a clinical point of view, Blanchard is just saying: ‘no matter what categories or tags we stick to transexuals, we can only ease their gender dysphoria through transition’. We can all agree with that. Why bother to divide transexuals in two (or more) categories at all, since that categorisation does not lead to different ‘treatments’, nor even a better understanding of what transexuality actually is?

Does that mean that we should scratch Blanchard (and Blanchardian followers) out of the history of science, and forget all about them? Many would claim yes (the activists certainly would!), but I’m a bit more tolerant. I think that Blanchard has made two important contributions to transgender studies. First, he was the first to name ‘autogynephilia’. While this word is now too politically loaded to be useful any more, it’s quite true that some people assigned male at birth (and a handful assigned female at birth) do, indeed, have frequent erotic dreams as themselves as women (and some natal females also have frequent erotic dreams as themselves as men, although this group is substantially smaller). This was not apparent to, say, Benjamin or early researchers in clinical sexology. Recognising that such individuals actually exist was a big step, and an important one. The trouble is just relating autogynephilia to some kind of ‘special transexuality for older men who do not remotely look like women and are physically attracted to women besides having weird dreams’. But just having a name, a category, for psychologists to probe into their patient’s mind, was important: some kinds of so-called autogynephilia may be caused by traumatic experiences which can be addressed by therapists, for instance (and in that scenario these people are not even remotely transexual and would thus not benefit from transition — unlike what Blanchardians claim); recognising them is important.

And the second contribution is actually involuntary. While Blanchard wanted to claim the contrary, he actually proved that, no matter what ‘label’, ‘tag’, or ‘category’ we place transexuals in, they all benefit from transition in an equal way. That is actually a very important contribution, because it means that, no matter what the prejudice of the clinical sexologist is, and how they wish to ‘morally’ categorise transexuals, transexuals will nevertheless always benefit from transition — no matter how they look, how old they are, and what kind of sexual fantasies they entertain. This is actually a breakthrough in the sense that up to the 1970s it was thought that only very feminine-looking (i.e. androgynous) natal males attracted to males, and believing in a binary gender, would benefit from transition, excluding everybody else. After Blanchard, we learned that all transexuals benefit from transition, not just those the doctors think that they are ‘fit’ for transition. That was an important accomplishment, and in that particular conclusion Blanchard was right and continues to be right.

Unfortunately, it also disproves his own assumptions that there are only two kinds of transexuals, but that’s ruthless science for you.

Also published on Medium.