According to the theory of mind account of autism, the social difficulties experienced by people on the autism spectrum arise because they are unable to understand their own or other people's behaviour in terms of motivating mental states such as beliefs, desires, intentions, and so forth.
The prime evidence to support this account comes from studies showing that children with autism tend to perform badly on tests of false belief understanding. In the standard false belief task, a story is told or acted out whereby a character has a mistaken belief about the location of an object or the contents of a container. Children with autism on average perform worse than non-autistic kids on these kinds of tasks. They will tend to say, for instance, that the character will look for the object in the place where it really is rather than the place where the character thinks it is.
However, not all kids with autism fail. Indeed, by adulthood, many people with autism pass the tests fairly easily, although their social difficulties remain. Some of these individuals struggle on more advanced tests of theory of mind including tests of 'second-order' theory of mind (e.g., John thought that Sue thought that the icecream van was still in the park). It's not clear whether these tests are more difficult for them because they tap more sophisticated theory of mind skills or because the stories are just harder to comprehend or remember. In any case, there are still plenty of people with autism who can pass these tests too.
The prime evidence to support this account comes from studies showing that children with autism tend to perform badly on tests of false belief understanding. In the standard false belief task, a story is told or acted out whereby a character has a mistaken belief about the location of an object or the contents of a container. Children with autism on average perform worse than non-autistic kids on these kinds of tasks. They will tend to say, for instance, that the character will look for the object in the place where it really is rather than the place where the character thinks it is.
However, not all kids with autism fail. Indeed, by adulthood, many people with autism pass the tests fairly easily, although their social difficulties remain. Some of these individuals struggle on more advanced tests of theory of mind including tests of 'second-order' theory of mind (e.g., John thought that Sue thought that the icecream van was still in the park). It's not clear whether these tests are more difficult for them because they tap more sophisticated theory of mind skills or because the stories are just harder to comprehend or remember. In any case, there are still plenty of people with autism who can pass these tests too.
Advocates of the theory of mind account often play a kind of "Get Out of Jail Free" card, arguing that the people who passed the false belief test somehow did it in an abnormal way, perhaps by treating it as a logical reasoning problem rather than relying on more spontaneous intuitions. But, until now, there has been little evidence to back this assertion up.
This brings us to Senju et al.'s recent paper in Science. These authors attempted to bolster the theory of mind account by providing evidence that adults with Asperger syndrome have difficulties with theory of mind, even though they are able to pass traditional theory of mind tests. Asperger syndrome is related to autism and many people assume that it is really a mild form of autism. Indeed, Senju et al provide no evidence to show that their participants met criteria for Asperger syndrome rather than high-functioning autism.
This brings us to Senju et al.'s recent paper in Science. These authors attempted to bolster the theory of mind account by providing evidence that adults with Asperger syndrome have difficulties with theory of mind, even though they are able to pass traditional theory of mind tests. Asperger syndrome is related to autism and many people assume that it is really a mild form of autism. Indeed, Senju et al provide no evidence to show that their participants met criteria for Asperger syndrome rather than high-functioning autism.
The researchers monitored participants' eye-movements while they watched a series of video clips. In each clip a puppet placed an object in one of two boxes before leaving the scene briefly. A chime then sounded and the puppet returned to retrieve the object. By the fourth video clip, when they heard the chime, all the participants anticipated the puppet's return, looking towards the box containing the object.
The fifth trial was critical. The object was moved in the puppet's absence, so the puppet now had a false belief about its location. If a participant was tracking the puppet's beliefs then, when the chime was played, they should look towards the now empty box because the puppet did not know that the object had been moved from it. Otherwise, they should look towards the box where the object really was (as they had all done on the previous 'true belief' trial).
8 of the adults with Asperger's looked at the 'correct' box but the other 11 looked at the incorrect box. This compares with a group of non-autistic adults, of whom 13 looked at the correct box and only 4 looked at the incorrect box.
Senju et al. conclude on the basis of these findings that:
The fifth trial was critical. The object was moved in the puppet's absence, so the puppet now had a false belief about its location. If a participant was tracking the puppet's beliefs then, when the chime was played, they should look towards the now empty box because the puppet did not know that the object had been moved from it. Otherwise, they should look towards the box where the object really was (as they had all done on the previous 'true belief' trial).
8 of the adults with Asperger's looked at the 'correct' box but the other 11 looked at the incorrect box. This compares with a group of non-autistic adults, of whom 13 looked at the correct box and only 4 looked at the incorrect box.
Senju et al. conclude on the basis of these findings that:
- adults with Asperger syndrome do not spontaneously anticipate others’ actions
- individuals with Asperger syndrome have a persistent impairment in spontaneous mentalizing
- compensatory learning might explain the apparent paradox between success on explicit false belief tests and continued difficulty in everyday social interaction for individuals with Asperger syndrome
Senju et al's argument seems to be that, because the Asperger group was split roughly 50:50 between passers and failers, they were probably all guessing. It would be like each person tossing a coin and guessing heads or tails - you'd expect about half of them to be right just by chance, even though nobody actually has any idea what's going to happen. But this logic doesn't follow. On the previous trial, participants with Asperger's had unanimously looked towards the box containing the object. So why would they start guessing now? If their eye-movements were driven by their own knowledge of reality rather than the puppet's (false) belief then they should always look to the box containing the object. In other words, there's no reason to think that the 8 individuals in the Asperger group who looked in the correct location just 'got lucky'.
It's also worth remembering that 4 of the 17 non-autistic control participants also looked in the wrong location too. If we assume that these people were generally able to track the puppet's beliefs, this suggests that the test isn't completely reliable. In other words, even if you have a fully functioning theory of mind, you still only look in the correct location about a quarter of the time. So we can't even assume that all the 'failers' in the Asperger group lacked a theory of mind either.
To sum up, Senju's study suggests that there may well be some people with Asperger syndrome or autism who can pass standard theory of mind tests but don't spontaneously use theory of mind to anticipate other people's actions. This could perhaps explain their lingering social difficulties. However, this still leaves unaccounted for a large proportion of people with Asperger syndrome who pass standard theory of mind tests and can use theory of mind to anticipate other people's actions, but yet still face serious challenges in their everyday social interactions. This is not to diminish the importance of theory of mind. But it shows once again that we need to move on from one-size-fits-all theories of autism and Asperger syndrome.
Reference:
Senju, A., Southgate, V., White, S., and Frith, U. (2009). Mindblind Eyes: An Absence of Spontaneous Theory of Mind in Asperger Syndrome. Science, 325, 883 - 885.