anthony Wrote:How did we decide that this guy and Elliot Roger are manifestations of the same genetic condition?
This question hits the nail on the head. I cannot speak much to the social causes of autism, of which I'm convinced there are important ones and the OP already points in the right direction. But on the neuro-biological, genetic side, I can shed some light.
Post scriptum intro: I wrote much more than I initially intended, so I highlighted the most important sentences for the skimmers.
The first tripping wire in this discussion is of course the colloquial terminology. It may seem unimportant, but the fact that we've come to call slight social akwardness "autistic" really does something to our perception of the phenomenon. When discussing this, you should at least check your own terminology to try and be exact about what you mean. I would also refer to recent literature showing that much of the supposed increases in autism and ADHD in the new millenium is due to less strict diagnostic guidelines - this process cannot be separated from the public discussion around it.
The second thing is then much more general, pervasive and difficult to untangle. What everybody outside of psychiatric/neurological research needs to understand (and many within, unfortunately) is that these "diseases", menthol elf problems, are phenomenological collection bins, conceptual groupings of symptoms. The underlying physiological mechanism are diverse and very poorly understood. In plain English: Two people can be diagnosed with depression, or schizophrenia, or autism, and even share symptoms, while having little to nothing in common "under the hood". The brain (or even the whole body) is in many ways much more complicated than the behaviour it produces. The outputs are very limited compared to how much can go wrong.
Because it is important to me that I communicate this properly, I will create a comparison: When somebody has the flu, a doctor has also learned a "conceptual grouping of symptoms" that helps him recognise that, yes, this patient has the flu. Every patient with the flu will not have the all of the symptoms, and some will have almost none of them, and many will have symptoms from other diseases as well. In that way, you could say there is a "flu-spectrum". However, we understand the etiology of the flu pretty well. We can do a test for influenza virus, and determine the variant even. We know what the virus does to the body and can explain the symptoms in terms of those vectors of attack.
Now we compare this to autism. We know nada. There is no biomarker, no virus, bacterium, fungus, contaminant, nothing we can objectively test. There are symptom checklists. We say Elliot R. and the screaming retard both shared some symptoms like eye contact aversion and mysoginy, and later we determined that they both have a rare mutation in some gene that may or may not explain 0.003% of the variance in social behaviour. So we classify them both on the autism spectrum, and treat them according to convenience. That's it. That's the summary of fifty years of autism research (not being generous, but you get the point).
Why are we not progressing to a point where we can say "Serotonin-deficiency in the brain causes depression", "Vaccines cause autism", "TV causes schizophrenia"?
Because these are category mistakes. Some cases of the extremely varied behaviours that are currently all classified as autism could very well be due to vaccines*, or smartphones, or deleterious genetic mutations. But never all of them. By the maxim that "absence of evidence is (after doing loads of research) evidence of absence", we can even with some confidence say that not even a larger subset of these symptoms has a single common etiology.
These categories are simply incoherent in biomechanical terms, and rest on social intuition first and foremost.
So, what should we do then to progress this research?
Abandon the symptoms-based approach altogether. It's unlikely that we can convince doctors of this yet, but basic research needs to deal with research-domain specific approaches. The goal has to be an insight of this format:
"Cause X leads to abnormal Y leads to behaviour Z"
"Mutation in gene X disturbs development of neural pathway Y which changes social interaction Z".
"Glyphosate contamination of leafy greens leads to malabsorption of Vit.C in 4% of population which increases their average eye contact avoidance by 7.3%"
If that sounds unsatisfactory, that's the point. If there were a slam dunk for this, it would have been found. My whole point is, there can't be a slam dunk while we focus on classifications that are based on phenomenology instead of mechanical connections.
Fortunately, there is movement in this direction, and my prediction is that it will result in the disolution of psychiatry as we know it now, if successful. Instead of treating all depression with SSRIs, we will hopefully come to recognise different ways in which somebody can be damaged that lead to him presenting "depressed", and find interventions appropriate for each. Similarly, the nonsense about autism has to stop. I am not very well versed on it's likely causes, to I won't comment, but it's clear to me that there's a ton of different disorders, from autoimmune to social neglect, that have an influence on the behaviours that are now haphazardly categorised as autism.
Something that should be mentioned about Harlow is that he also
studied how to reverse the damage he caused. Anthony correctly identifies him as an archetypal scientist worthy of admiration.
*
(I don't think it's likely btw, it's more likely to be a compound effect of contaminants that includes some vaccines. There are weird auto-immune effects when the body has to deal with a bunch of intrusions that could separately be harmless)