Graham first drew my attention to the article (link at end) about the 'predictive coding' theory of autism. In researching it a bit more, I am now very persuaded by the theory.
Unlike so many other 'unified theories' of autism, which seek to explain all aspects of the condition (often by way of a model of 'impairments'), this one builds on a broad account of brain function based upon perception, expectation and response. Put simply, we all build up models in our heads based on our experiences of things: how we perceive them. These models are then used to inform us about what to expect when we encounter these experiences again. The brain makes 'predictions' based on the model it has already constructed. As a simple example... we see a cat chasing after a piece of string. That perception, of how the cat will behave when a piece of string is pulled, is then built upon when we encounter the same situation involving other cats. So we construct the model in our heads: we know from prior experiences and expectations that all cats will most likely behave in a similar way. This enables us to 'predict' the probability of future outcomes when we encounter cats and present them with pieces of string.
But what happens when we pull a piece of string in front of a cat... and it ignores it? We are presented with a discrepancy. Our expectation, based on the model our brain has constructed, has been confounded. Now, most people will come to understand that such discrepancies are likely to occur, so are therefore able to cope with them. They can fit them into their model as a variable, if you like. The small details become less salient, and the brain shifts its focus to the bigger picture. With the autistic brain, however, such discrepancies - essentially, violations of expectation - are not so easily assimilated into the model. The focus is more likely to be on those small details rather than on broader patterns, so far more weight will be given to those discrepancies.
This is a very broad overview of the theory. The article explains it in far greater detail. But taken all in, it can be used to explain many things that are common with autism: minute attention to minor details (background noises, etc), preference for routine (predictable) activities, the pervasiveness of anxiety in unpredictable situations, and motor control deviations. It would also explain many of the problems we face in dealing with those most unpredictable things of all: human beings!
From the article:
'To predict what someone will do in a given context, you may need to make a guess based on what they or someone like them did under different circumstances. That is hard for anyone, but more so for people with autism. It’s very common, for example, for [people with autism] to get into social interactions and have difficulty taking what they’ve learned from situation A and bringing it to situation B.'
For me, this helps to explain, for instance, my reaction to my bullying colleague in my last job. How at first she empathised with me and encouraged me to trust her, leading me to 'predict' that she would always be someone I could have confidence in. And then the mess I was left in when her behaviour took a radical change. It made no sense to me. And because of that, I couldn't just 'let it go over my head', as my NT colleagues were suggesting. I couldn't just accept it and assimilate it into my mental 'model' of her - as some kind of extreme variable. It just completely threw my switches. Which is what I was trying to explain to my managers, and what they - as supposed autism experts - signally failed to grasp.
As the article rather succinctly puts it:
For a person with autism, the world never stops being surprising.
Quite! Now, here it is: