Self-Diagnosis, Misinformation, and the Ontology of Autism

It's true that anyone can spread misinformation by saying it. However, what they can't do is spread misinformation by BEING it, since someone can't be other than what they are. Even when someone is acting, their motivation and ability to act determines how they act, so how they act is still a reflection of how they are. For example, even though the "goal" of masking is to appear non-autistic, it's actually a symptom of autism. Normal people don't pretend to be normal, and masking doesn't actually look normal either - it only "works" because people don't know to think "autism" when they see it. So as long as you see people as they are rather than just blindly going with what they say, then they are valuable source of information. Now, it's also true that this also applies to people who aren't diagnosed, so they also provide information about autism. The issue is verifying that this information describes autism requires a diagnosis. Without verification, meaningful information can't be distinguished from noise, so looking to self-diagnosed people is useless for learning about autism by example. To show why it's important to have such a source, I'll need to explain a few other concepts.

The next thing that needs to be discussed is the ontological problem of autism. Autism isn't "real" - what's real is autistic people. Obviously there is something that distinguishes autistic people from the general population so they can be identified as being autistic, but that thing doesn't exist as an abstract quality separate from the people who are autistic. It's like how you can have a big truck, but you can't just have "bigness" that doesn't describe anything, especially since a big truck is still smaller than small star. With this in mind, it becomes more clear why diagnosis is important. The fact of the matter is that psychologists often don't know what they are talking about when they are describing autism, but that doesn't matter when it comes to diagnosing as they use a certain process and that process identifies certain people, which is one way of defining autism. It's certainly not the best way of defining autism as their are real similarities between some people who end up being diagnosed as autistic and some people who aren't, but the key take away is that the same process should identify the same sort of people, and that is something that can be relied on. In some verifiable sense, people who diagnosed with autism know that they are similar to each other.

The reason it's important to identify people similar to oneself is due to introspection bias. The term bias is a bit misleading as it seems to imply that introspection isn't valid, but what it's really saying is that people see themselves differently then other people see them, and that difference is the bias. One cannot diagnosis autism in themselves not because they don't understand themselves, but because the diagnosis process was made based on observations, not introspection, so it would lead to different results. There are also other things that people want to know about themselves that they can't find out through introspection, so they need to look to others like them. By finding people other people like them, people can learn more about themselves.

Of course, autism is a spectrum with a wide range of presentations, so someone can't just look at some else with autism and conclude they are the same in any particular way. It's all autism though. One particular useful feature of seeing the diversity of the autism spectrum is disproving myths about autism in the form "everyone with autism" since it only takes seeing one person with autism to disprove it. For example, I've heard people claim that autistic people can't have jobs. A young autistic person may hear this and be anxious about their future. However, autistic people do have jobs, and having an example of specific autistic person with a job can serve as role model to this young autistic person. It gives them hope. It also helps all the people who aren't diagnosed as well, as it was from starting from people who actually were diagnosed that they discovered they may have been missed due to myths about autism.
Now, consider an undiagnosed autistic person with a job who previously thought they couldn't be autistic because they have a job. They then learn that autistic people actually can have jobs, so then they decide to "self-diagnosis" and later give themselves as an example of an autistic person with a job - I've seen people do stuff like this when people ask questions about autism. Someone then looks at this person presenting themselves as an autistic person with a job, points out they are self-diagnosed, and claims they are actually not autistic BECAUSE they don't have a job. It's a circular argument, but without a diagnosis the self-diagnosed person is unable to counter it without referring to someone who actually is diagnosed. Unfortunately, they only heard that autistic that autistic people can't have jobs from other people in the community, and don't actually know any diagnosed with autism who has a job, so they can't counter the claim. Maybe the reason they aren't able to get a diagnosis *is* because they have job. An onlooker then sees this exchange, and is left believing that autistic people can't have jobs after all. Of course it's an incorrect conclusion, but why should they believe otherwise based on what they had seen?

So yes, the real issue is misinformation, but you can't fight misinformation with memes like merely telling people that autistic people have jobs because you heard it elsewhere, you need a rigorous foundation. Myths are ALSO spread through memes. The difference is the truth stands up to rigorous testing, while myths fall apart. It's not like the autistic community is immune to myths either. Just in the last couple years I've been seeing this bizarre idea that female autism is completely different from male autism because females mask while males don't. This is nonsense, as it's all one autism, and people of all genders and sexes mask to varying degrees. I guess the reason for this myth is an overgeneralization of the fact that women are underdiagnosed (for a variety reasons), and that it's been found that women diagnosed with autism on average mask more then men do, but there is more going on than just that. As another example, I think I actually know more autistic women (who I assume are mostly AFAB) with monotone voices then I know autistic men (who likewise I assume are mostly AMAB) with such voices, it's just that it was *always* a myth that autistic people must speak in monotone (in fact, I saw a study that found that autistic people actually tend to have a wider range of tone when speaking then neurotypicals do). As for the women, I guess their unusual prosody just got attributed to anxiety or depression instead - misattribution of features due to the application of gendered schemes is as much of an obstacle to diagnosis as an actual difference in presentation.

Aside from being sexist, these myths about autism being gendered can bring back the myths that the truths they were based on helped fight. The underlying threat goes back to the ontological problem of autism. Without a rigorous foundation for what makes autism real, a skeptic can just look at the fact there are supposedly two different autisms, and conclude only one of them is the "real" autism". Specifically, they are going to conclude that all the myths about autism are real, that male autism is the real autism and the disorder is just inherently linked to sex, and that all these women aren't autistic, and then we're back where we started. However, there is a rigorous foundation, and it goes back to diagnosis. First, it's important to recognize how diagnosis works systemically as that explains how diagnosis can miss people even if it's a valid process. On a systemic level, diagnosis is actually divided into two processes, the less rigorous screening and the more rigorous process of assessment. Screening picks out people to assess, but by studying people who were diagnosed with autism through an assessment after failing screening it was proven that screening doesn't pick out all autistic people, and thus there are biases in the diagnostic process independent of the definition of autism. Next, there is the fact that autistic people don't present as autistic children, but all validated assessment tools were designed for children with autism (the ADOS is approved for assessing adults, but it still wasn't designed for them). This is proven by the simple fact that autistic children grow up to become autistic adults, and the differences between them can be observed. Since we've already confirmed that autistic children can be missed during screening, they can then fail assessment as an adult, but logically they are still autistic. The final essential piece of evidence is that AFAB people tend to be diagnosed later in life than AMAB people. This is enough to explain why more AMAB people are diagnosed then AFAB people since there is a temporal lag in the numbers (which is further compounded by the fact more children are being diagnosed now then in the past), but the fact autism is harder to diagnose in adults means that there isn't just a temporal lag - the women who were missed as children often stay missed. From there it's just a matter of applying statistics to try to estimate how many women are left undiagnosed, and that's all while sticking with the idea that autism is defined by the people diagnosed with it.

The bias is even greater if you try to solve the ontological problem of autism by redefining it terms of something that actually is "real". To fix the problem of women and girls being underdiagnosed, in the very least the screening process for children need to altered so it is no longer biased against girls, and tools for assessing adults needs to be developed, especially ones with women in mind as well as men. This can all be done rigorously by looking at the data of people who were diagnosed at other points in their life eg. identifying traits that autistic girls who were diagnosed as teenagers had back when they were children that differed from the neurotypical peers. That's hard to do so it's still a problem being worked on, but it's still not enough. As long as we assume there is something real behind autism, the diagnosis criteria can be just as biased as the screening process and assessment tools, and thus should also be revised, but it's less clear how exactly one would go about doing that without already having a foundation for autism that is not only rigorous, but "real". Some people have already started looking down these lines by looking at brain-scans and genes, since apparently biological markers are more real than psychological or behavioral ones. When it comes to the questions that actually matter though, I think psychology is much more relevant than neurology - I'm like someone because we think the same way, not necessarily because our brains superficially resemble each other. For answering these questions, introspection is a more valuable tool than an MRI scan.

This brings us back to self-diagnosis. We already established that it's not equivalent to actual diagnosis. However, say that instead of reading Wikipedia, someone reads first person autism-literature, and they find how they describe their own experiences match how other people describe their experiences. If we instead define autism in terms of how we think instead of how we behave, then wouldn't self-diagnosis actually be MORE valid than actual diagnosis? It would be, if there wasn't the fundamental issue that it isn't validated. The first issue is that it doesn't rest on a stable definition. Defining autistic people as people who relate to other autistic people is self-referential - substitute "autistic" with "neurotypical" and it's still true, but now it's referring to a mutually exclusive group of people, and the only difference is a word that is otherwise undefined. We still have to back to actual diagnosis in order to identify the people who are autistic, and not some other group. The second and more important issue is that the process hasn't been verified to lead to meaningful results. For example, if we go back to our young autistic person concerned about if they are going to ever get a job, then considering self-diagnosed people to also be autistic wouldn't be helpful to them if the self-diagnosed people have a very different prognosis then those who are diagnosed. It should be noted that psychologists are EXTREMELY skeptical to the validity of introspection, and will not accept any method that relies on introspection without rigorous testing to ensure it's validity. Until a process of self-diagnosis is devised that hold up to rigorous psychometric standards, then it cannot be considered valid.

I personally though am less skeptical to the validity of introspection, so I think most "self-diagnosed" people are in fact autistic, at least those who properly looked into first-person literature for autism and as well as disorders that can be confused for it so they are can identify that they probably have autism and not something else. In the very least, people who "self-diagnosis" often end up having their suspicious be validated when they get a diagnosis later, so it has potential for screening tool even without questioning the ontological nature of autism. Still, these are all anecdotes, not data, so it still has yet to be validated as a screening tool. The closest thing we have to that is the AQ, which has been validated as screening tool, but it isn't perfect - I only score in the 20s, which just made me more skeptical of either my own diagnosis or it's validity as a screening tool (meanwhile my friend who at the time was not diagnosed as autistic and did not suspect it in herself scored in the high 30s so that furthered my suspicions, though she did end up getting diagnosed seven years later). We've still got a long way to go before any form of self-diagnosis can be considered valid in the same way as an actual diagnosis.

In the meantime though, it works for people who are simply looking for a community of like-minded people. They are an important part of our community, and I think they will play an important role in redefining autism so it's not longer restricted to a pathological context. Until they get the confirmation I don't think they should got around telling people that they autistic, but they belong in our spaces, and should be made to feel welcome. I don't know how to ensure people will feel welcome also while also telling them they shouldn't do thing that other people in the community are allowed to do, but I've always valued honesty more than tact, and I've been told that's an autistic thing. Either way, we need to make sure the dialogue keeps going, and doesn't just get reduced to repeating information with nothing to back it up.